MY
wistad Oa
feu t!
ANY phot Dey Mw
ue ek weed SET
vs
¥ By Panera he weal vey ‘
on
Ls) Aer RSE
as
Se
ea { ‘ Sian ne
At
Ht
Ny
(ie
3 a sits RUGS pipe
wv is i
xt! Onn
/ ew,
pete af
+
Digitized by the Internet Archive in 2016 with funding from South African National Biodiversity Institute Libraries
https://archive.org/details/journalofsouthafi2unse
es
—
=
lll al
VOLUME XII. 1946.
THE JOURNAL OF
SOUTH AFRICAN BOTANY
PUBLISHED UNDER THE AUTHORITY OF THE TRUSTEES OF THE
NATIONAL BOTANIC GARDENS
OF SOUTH AFRICA
KIRSTENBOSCH, NEWLANDS CAPE PROVINCE
EDITOR R. H. COMPTON, M.A. (Cantab.), F.R.S.S.Af., Hon. F.R.H.S.
HAROLD PEARSON PROFESSOR OF BOTANY IN THE UNIVERSITY OF CAPE TOWN, DIRECTOR OF (PR RASS INR
4 f_ {= my 5 py :
| | |
Se eae
THE JOURNAL OF SOUTH AFRICAN BOTANY.
FF SEE ———————————————
VoLuME XII, 1946.
i Pia
CONTENTS. PAGE Some New SpEci=zs AND VARIETIES IN THE GENUS Hawortaia, V, By G. G. Smith (With Plates I and II) ., 00 0 aD a6 an 00 00 00 00 1 DrEz SAADKNOP EN EMBRIOLOGIE VAN BRABEIUM STELLATIFOLIUM. L. Deur Dr. P.G. Jordaan 15 THE DEVELOPMENTAL MorPHOLOGY OF THE POLLEN OF PROTEACEAE. By S. Garside .. 27 PrLantTaE Novak Arricana®. Series XXV. By Professor R. S. Adamson, M.A., D.Sc. 5 Rev. J. Gerstner, 0.8.B., Ph.D. ; and Captain (S.) T. M. Salter, R.N. (rets) ae ae 35 Series XXVI. By Professor R. H. Compton, M.A., and Captain (S) T. M. Salter, R.N. (Ret.) a yi ok fa ag a a si it a 81 A Dryineé CABINET FOR THE HERBARIUM. By G. G. Smith a ifs ie 26 43 Novres ON THE SoutH AFRICAN SAPOTACEAR. By Rev. J. Gerstner, O.S.B., Ph.D. a 47 RESERVE CARBOHYDRATES IN SouTH AFRICAN GRASSES, By Dr. H. Weinmann, M.Sc., Ph.D., and Miss L. Reinhold. (With Plate III) is BG Bs an j 57 REVIEWS AND ABSTRACTS : A. W. Sampson, Plant Succession on Burned Chaparral Lands (Dr. C. L. Wicht) .. U6) 8S. D. Garrett, Root Disease Fungi (Dr. E. S. Moore) .. ae vt a av 76 J.G. Horsfall, Fungicides and Their Action (Dr. J. E. van der Flank) .. oe ae 77 J. G. Gilman, Manual of the Soil Fungi (Professor W. J. Liiijeharms) .. ke as 78 C. E. ZoBell, Marine Microbiology (Dr. W. E. Isaac) .. as an a Bs 80 J. Hutchinson, A Botanist in Southern Africa (Professor R.S. Adamson) .. Bits 121 S. A. Wilde, Forest Soils and Forest Growth (Dr. C. L. Wicht) .. N: Bis ae 123 Some Norrs on THE GENUS HERMANNIA WiTH DESCRIPTIONS oF Fives NEw Species. By Captain (S) T. M. Salter, R.N. (Ret.) .. Se Be 06 aa a oh 95 A Note on UrGINEA pyeGMarA DuTHIE. By Professor R. S. Adamson, M.A., D.Sc... 105 ROELLA RETICULATA. By Professor R. S. Adamson, M.A., D.Sc... ee me 5.0 107 THE SourH Arrican QuININE TREE. By Dr. E. P. Phillips, M.A., D.Se. .. Ee a 111
THE GENERA CHILIANTHUS BuRCH AND BUDDLEJA LINN, By Dr. E. P. Phillips, M.A., D.Sc. 113
Drz ANATOMIE VAN DIE WoRTEL VAN RESTIO PALUDOSUS EN LEPTOCARPUS VIMINEUS.
Deur Dr..P. G. Jordaan .. an av aie Si 06 BO ars ie 115 Cart Peter THUNBERG. AN EARLY INVESTIGATOR or Capr Borany. By Miss M. C.
Karsten.
Part IV (with Plates IV—VII) .. dc be 6 3 te ae oo |
Part V (with Plate VIII) .. if O10 ave ae ot or at eG:
InDEX TO PLant Names, Vou. XII, 1946 .. Ap ars one oe Ms ae 191
; - . 3 a 8 1 ; 5 Us 5 1 i 1 ‘ * E ‘ 4 ' ; / i ; ' Ys :
JOURNAL. OF
SOUTH AFRICAN BOTANY
VOLUME XII.
y
SOME NEW SPECIES AND VARIETIES’ IN THE GENUS HAWORTHIA. V.
(With Plates I and 100.) By G. G. Surrn..
Haworthia Schuldtiana var. major G. G. Smith. (Liliaceae- Aloineae.) Sect. Retusae.
Rosetta acaulescens, diam. 8} em.
Folia circiter-30, ascendentia, longa 39 mm., lata 12—13 mm. - pars
>
retusa longa 18 mm., lata 10 mm., convexa, turgida, multis tuberculis .
prominentibus concoloribus albidis denticulatis ornata, pellucide viridis, lineisque 3(—7) albido-viridibus percursa, lineis 1—3 apicem prope attingentibus ; swbtus apicem versus tuberculis orbiculatis prominentibus albidis ornata.
Pedunculus simplex, diam. 21 mm., longus una cum racemo 30 em., pedicelli longi 24 mm. ; perianthium opace ravo-albidum, longum 19 mm. : ovarium longum 33 mm., diam.: 14} mm., viride ; stylus longus ‘1 mm., albido-viridis, flexus, capitatus.
Rosette acaulescent, up to 84 em. across.
Leaves up to 34, ascending, recurved, firm, 39 mm. long, 12—14 mm. broad below base of end-area, 9 mm. broad at base of leaf, 10} mm. thick at base of end-area, obovate, acute, end-awn 2—4 mm. long, persistent ;
KEY TO THE.:DRAWINGS.
_ F = Leaf face. B = Leaf back. LS = Longitudinal section. SM = Section near middle. SB = Section near base. ST = Section near tip. INF = Inflorescence. Del. M. C. L. = Drawn by Miss M. O & S = Ovary and Stamens. Courtenay-Latimer. C= Capsule
2 The Journal of South African Botany.
face below end-area convex, smooth below and with pellucid whitish mostly toothed spots near end-area, dark green to ‘brownish-green, dull ; end-area 18mm. long, 12—l14 mm. broad, convex, turgid, with many raised concolorous and whitish tubercles which are beset with white teeth, dark greenish pellucid, dull, with 3(—7) greenish-white + converging
Es:
lz va. MCL. 1M, oe. WC-L. Fie. 1. H. Schuldtiana var. major. H. retusa var. multilineata.
lines of which 1—3 nearly reach the tip; back convex, smooth below, beset with round raised whitish mostly toothed tubercles near tip, light green below, very dark green to dark brownish-green above, dull; keel 1—2, in the upper 3, + acute near tip, with ? mm. long whitish subulate teeth about 14 mm. apart; margins + acute below, with 1 mm. long white subulate teeth about 1 mm. apart, becoming very short towards base of leaf.
Peduncle simple, terete, 21 mm. diam., 30cm. long including the
Some New Species and Varieties in the Genus Haworthia V. 3
raceme, greyish-brown below; raceme 84cm. long, about 20 spirally arranged flowers and buds, 1 open at a time; pedicels 24 mm. long, 1—1} mm. elliptic, dark greyish-green ; sterile bracts 16, very large, 114 mm. long, the lowest one 2 cm. from base of peduncle ; fertile bracts 10} mm. long, deltoid acute-acuminate, white with a fine dark brown nerve ; perianth dull greyish-white, greyish-brown lined, 19 mm. long, the cylindrical-triangular sub-globose base 5mm. diam., constricted to 24 mm. above, ascending, curved, not stipitate ; wpper segments obtuse, the outer ones replicate towards the tip, long recurved, + spreading, face colour white with a broad greenish-brown nerve, the inner segment channelled, hardly recurved, much shorter, face colour white at tip, green below, with a medium-fine dark greenish-brown nerve ; lower segments obtuse, channelled, recurved about 75 degrees, the outer ones spreading, face colour white at tip, green below, the inner segment longer, face colour white with a medium fine greenish-brown nerve ; stamens 54 and 6 mm. long ; ovary 3? mm. long, 1} mm. diam., green ; style 1 mm. long, greenish-white, bent, capitate.
Habitat : Cape ‘Province ; Riversdale Dist. (Type G. G. Smith, 5370, (Dekenah 9) in East London Museum Herb.).
Distribution : not further known.
This variety is described from material collected by Mr. J. Dekenah on the Karroo side of Garcia Pass, and sent to the Author in Oct.,1943. The plants grow tightly packed among rocks and in rock crevices on a northern aspect. Compared with H. Schuldtiana v.P. the plant is larger, much more heavily toothed and tubercled, and has a long end-awn.
Haworthia retusa var. multilineata G. G. Smith. (Liliaceae-Aloineae. ) Sect. Retusae.
Rosetta acaulescens lata usque ad 8 cm.
Folia circiter 15, longa 49 mm., lata 29mm.; pars retusa longa 26 mm., lata 29 mm., deltoidea, acuta, lineis 16—22 longis brevibusque pallidissime viridibus percursa, apice seta longa 5mm. ; subtus concava, laevia ; carina non denticulata ; margines parte inferiore minutissime denticulati.
Pedunculus simplex, diam. 23? mm., longus una cum racemo 16 cm., basi ravo-fuscus ; pedicelli longi }mm., virides; perianthiwm opace album, longum 15} mm.; ovarium longum 3}mm., diam. 13 mm., viride ; stylus longus ? mm., opace albus, capitatus.
Rosette acaulescent, up to 8 cm. tall, not or slowly proliferous from the base. ;
4 The Journal of South African Botany.
Leaves 12—15, recurved, firm, 49 mm. long, 29 mm. broad and 17 mm. thick at the base of end area, end-awn about 5mm. long, persistent ; face below end-area flat at base, concave above, smooth, with fine longitu- dinal flecks, light greenish-brown + mottled, dull; end-area 26 mm. long, 29mm. broad, deltoid, acute, recurved at an angle of about 85 degrees, concave, turgid, beset with one or two raised concolorous tubercles, greenish-pellucid, the tip brownish-pellucid, -- shining, with about 16—22 long and short + parallel very light green lines, the middle one .reaching the tip, the others progressively shorter; back convex, smooth, with many longitudinal flecks, light green below, light reddish- brown then dark greenish-brown above to blackish brown at tip, dull, with about 8 broad lines alternating with fine ones; keel 1, acute near tip, naked; margins acute below, + acute along the end-area, with very minute teeth in the lower half of leaf.
Peduncle simple, terete, 2? mm. diam., 16 cm. long including the raceme, greyish-brown towards the base of peduncle ; raceme 5 cm.long, about 27 spirally arranged flowers and buds, 2 open simultaneously ; pedicels 4 mm. long (almost sessile), green ; sterile bracts about 33,12 mm. long, the lowest 3 cm. from base of peduncle : fertile bracts 8 mm. long, deltoid, acuminate-subulate, white with a fine brown nerve ; pervanth dull white, 153mm. long, the cylindrical-triangular base 4 mm. across, constricted to 3mm. above, ascending, not or hardly curved ; wpper segments narrow obtuse, recurved about 80 degrees, channelled, face colour of the two outer segments dull white with a fine greenish-brown nerve, + spreading, face colour of the inner one dull white at tip, greenish below, with a broad dull green nerve; lower segments narrow obtuse, channelled, recurved about 90 degrees, face colour of the two outer segments white at tip, green below with a broad dark green nerve, spread- ing, face colour of the inner one dull white with a fine green nerve ; stamens 5 and 6mm. long; ovary 34mm. long, 14 mm. diam., green ; style £ mm. long, dull white, bent, capitate.
Habitat - Cape Province ; Riversdale Dist. (Type, G. G. Smith 5383, (Dekenah 83) in the East London Museum Herb.).
Distribution : not further known.
This attractive variety was collected by Mr. J. Dekenah in Dec., 1942, 2 miles north of Riversdale growing on a very dry stony hill facing east, in the open and under small bush. Compared with the species this plant is larger, has longer, broader and darker green leaves which are more retused, there are many more face lines on the end-area, and the margins and keel are entire except for very minute teeth on the margins on the lower half of the leaf.
Some New Species and Varieties in the Genus Haworthia V. 5
Haworthia retusa var. solitaria G. G. Smith. (Liliaceae-Aloineae.) Sect. Retusae.
Rosetta diam. 8 cm., a basi non (vel aegre) proliferans.
Folia 15 circiter, longa 40 mm., lata 21 mm., seta extrema persistente longa 43 mm. ornata ; pars retusa longa 21 mm., convexa, opace pellucide viridis, aspera, tuberculis parvis concoloribus lineisque 9—15 supra ornata ; subtus convexa, laevia, claro-viridia, apice obscure caeruleo- viridia ; margines carinaque plurimum denticulati.
H. retusa var: solitaria. H. retusa var. densiflora.
Pedunculus simplex, diam. 4 mm., longus una cum racemo crebre floreato 14 em. ; pedicelli longi 2 mm., diam. 13 mm.; bracteae steriles 32 circiter ; perianthium albido-viride, longum 18mm. ; stamina longa 53 and 63 mm. ; ovarium longum 4mm., diam. 2 mm., viride : stylus longus I mm., capitatus.
Rosette acaulescent, 8 cm. diam., not (or slowly) proliferous from the base.
6 The Journal of South African Botany.
Leaves about 15, the young erect, the old ascending, recurved, firm, 40 mm. long, 21 mm. broad and 14} mm. thick at the base of end-area, with a distinct 44 mm. long whitish naked persistent end-awn ; face below end-area + flat below, concave above, smooth, with very pale papillae, light green, not shining ; end-area 21 mm. long, 22 mm. broad, convex, turgid, deltoid acute (—acuminate), rough with small con- colorous tubercles, dark pellucid green, + shining, retused at an angle of about 90 degrees, with 9—15 + parallel lines, the middle one reaching or nearly reaching the tip, the others progressively shorter ; back convex, smooth; light green, dark bluish-green at tip, dull; keel 1, in the upper 4, acute near tip, with many # mm. long teeth which are green below and pellucid near tip, ? mm. apart ; margins acute below, + rounded above, dark green above, light green below, the reflexed teeth green below and pellucid near tip.
Peduncle simple, terete, 4mm. diam., 14cm. long including the aggregate cylindric-conic raceme, brown towards the lower end ; raceme up to 53 cm. long, about 35 spirally arranged flowers and buds, 2 open simultaneously ; pedicels 2 mm. long, 1} mm. diam., dark green ;_ sterile bracts about 32, 15 mm. long, the lowest 1 cm. from base of peduncle ; fertile bracts 12 mm. long, deltoid, long acuminate, white with a fine dark brown nerve; perianth greenish-white, 18mm. long, the cylindrical- triangular base 5mm. across, constricted to 3mm. above, ascending, curved ; upper segments channelled, recurved at an angle of 90 degrees, face colour of the inner segment pinkish-green with a broad green nerve, broad, short, face colour of the 2 outer pinkish-white with a broad dark green nerve, narrower, acute, longer and + spreading; lower segments acuminate, recurved about 90 degrees, face colour of the inner segment pinkish-white with a fine green nerve, channelled, broad, face colour of the 2 outer white with a broad green nerve, narrower, replicate, spreading ;
stamens 54 and 64mm. long; ovary 4mm. long, 2mm. diam., green ; style 1 mm. long, bent, capitate.
Habitat : Cape Province; Riversdale District. (Type G. G. Smith 5373, (Dekenah 5) in the Kast London Museum Herbarium).
Distribution : not further known.
This variety is described from material sent to the author by Mr. J. Dekenah of Riversdale, in Oct., 1942. It occurs about 6 miles north of Riversdale on the Corrente River, growing amongst rocks and practically covered with turf. The plants grow. flat, with the retused surface level with the ground, and although found in dense patches, they are all single. Mr. Dekenah writes, ‘“‘ I have found no plants with suckers.” Compared with the species, this variety has a long persistent end-awn, the end-area is rough with small concolorous tubercles, has more face lines, the margins
Some New Species and Varieties in the Genus Haworthia JV. 7
and keel have many small teeth and the peduncle is much thicker and with a short dense raceme.
Haworthia retusa var. densiflora G. G. Smith. (Liliaceae-Aloineae.) Sect. Retusae.
Rosetta acaulescens, lata usque ad 9 cm.
Folia circiter 16, longa 45 mm., lata 23 mm. ; pars retusa longa 24 mm., lata 28} mm., tuberculis compluribus parvis prominentibus concoloribus ornata, lineisque circiter 11 longis brevisbusque claro-viridibus percursa ; subtus laevia, basim versus claro-viridia, apicem versus pulliora, opaca; seta extrema longa 6 mm., nuda, persistens ; carina marginesque denticulati.
Pedunculus diam. 3 mm., longus una cum racemo 123 em., basi claro-fuscus ; pedicelli longi | mm., obscure virides ; perianthium album, longum 17 mm. ; ovariwm longum 4 mm., diam. 1? mm., viride ; stylus longus ? mm., albus, capitatus.
Rosette acaulescent, up to 9 cm. across.
Leaves about 16, the young erect, the old spreading, recurved, firm, 45 mm. long, 23 mm. broad and 14 mm. thick at base of end-area ; face below end-area + flat at base, concave above, -- smooth or with a few small raised concolorous tubercles below base of end-area on some leaves, light green, dull; end-area 24 mm. long, 23 mm. wide, convex, + turgid, with a number of small raised concolorous tubercles and about 11 long and short light green + converging lines, the middle one nearly reaching the tip, the others progressively shorter, deltoid, acuminate, light pellucid green, shining, the older leaves retused at an angle of about 90 degrees ; back convex, smooth, light green below, darker green above, dull ; end- awn 6 mm. long, naked, persistent, reddish-brown at base, light brown at middle to dull white at tip ; keel central, 1 in the upper 4, + acute, with many very fine greenish-pellucid deltoid teeth in the upper 4; margins + acute, with many larger greenish-pellucid deltoid teeth in the lower half, entire above or with very small teeth.
Peduncle simple, terete, 3mm. diam., 12}cm. long including the aggregate cylindric-conic raceme, light brown towards the lower end ; raceme 44 cm. long, about 30 spirally arranged flowers and buds, 2 open simultaneously ; pedicels 1mm. long, 1—l} mm. elliptic, dark green ; sterile bracts about 29, 18 mm. long, the lowest 1 cm. from base of pe- duncle ; fertile bracts 9mm. long, deltoid, subulate, white with a fine reddish-brown nerve; perianth white, 17mm. long, the cylindrical- triangular base 5mm. across, constricted to 35 mm. above, ascending, straight, not stipitate; wpper segments recurved about 90 degrees, channelled, face colour of the inner one white with a broad green nerve,
8 The Journal of South African Botany. =
narrow obtuse, face colour of the 2 outer white with a fine greenish-brown nerve, + acute, + spreading: lower segments more recurved, face colour of the inner one white with a fine greenish-brown nerve, channelled, narrow obtuse, face colour of the two outer ones white with a broad green nerve, deeply channelled to replicate, acute, spreading ; stamens 5and 6mm. long; ovary 4mm. long, 1? mm. diam., green; style ? mm.
long, white, bent, capitate.
Habitat : Cape Province ; Riversdale Dist. (Type G. @. Smith, 5056 (Venter 106) in the Kast London Museum Herbarium).
Distribution : This variety was collected by Lt.-Col. H. Venter in the Riversdale Dist., and sent to the author in May, 1941. It is nearest to H. retusa var. solitaria G. G. Smith, but is a larger and taller growing plant, the leaves are somewhat less retused, and the end-awn is longer.
Haworthia mundula G.G. Smith. (Liliaceae-Aloineae.) Sect. Retusae
Rosetta acaulescens, a basi proliferans. :
Foha longa 38 mm., lata 13 mm.: pars retusa longa 17 mm., laevis vel paucis tuberculis parvis prominentibus concoloribus ornata, lineisque longis 1—3, brevibus 5 pallide viridibus percursa : margines dentibus longis } mm., + prominentibus albido-viridibus, retrorsis ornati.
Pedunculus simplex, diam. 1 mm., longus una cum racemo 16 em. ; pedicelli longi 2 mm., fusco-virides ; bracteae steriles 12, longae 6 mm. ; bracteae fertiles longae 4 mm. ; perianthium opace album, longum 17 mm. ; ovarium longum 3mm., diam. 14 mm., viride; stylus longus }mm., opace albus, inflexus, + capitatus.
Rosette acaulescent, 60 mm. diam., freely proliferous from the base.
Leaves about 16, the young erect, the old spreading, recurved, + incurved near tip, firm, 38 mm. long, 13 mm. broad and 10 mm. thick at base of end-area, 14mm. broad below, oblong, acute ; end-awn 3 mm. long, naked, mostly deciduous, brownish-white : face below end-area flat towards base, -- convex above, smooth, pale green below, dark green above, dull ; end-area acute, 17 mm. long, 13 mm. broad, convex, smooth or with a few small raised concolorous tubercles in 1—2 longitudinal rows, retused at an angle of about 70 degrees, greenish-pellucid, + shining, with 1—3 long and about 5 short pale green lines, the middle one nearly reaching the tip, the others becoming shorter ; back convex, — triangular near tip, smooth, beset with very minute whitish longitudinal flecks in the lower half, light green below, dark green above, dull, with about 12 longitudinal lines; keel 1, in the upper 4, acute, with } mm. long, acute, pellucid-green teeth ; margins acute, whitish pellucid below, with + prominent greenish-white 4mm. long retrorse acute-acuminate teeth which become smaller near the base of leaf.
UT
Some New Species and Varieties in the Genus Haworthia V. 9
Peduncle simple, terete, 1mm. diam., 16cm. long including the raceme, greenish-brown below; raceme 6 cm. long, about 8 spirally arranged flowers and buds, 1—2 open simultaneously ; pedicels 2 mm. long, ¢ by 1 mm. elliptic, greenish-brown ; sterile bracts 12, 6 mm. long, the lowest about I cm. from base of peduncle ; fertile bracts 4 mm. long, ovate, acuminate, white with a medium-fine dark brown nerve ; perianth dull white, 17 mm. long, the cylindrical-triangular base 3 mm. across, constricted to 2+ mm. above, erect + ascending, curved ; wpper segments obtuse, the outer ones + replicate, recurved, spreading, face colour dull
F. a MCL:
H. mundula. H. turgida var. pallidifolia.
pinkish-white with a medium-fine dark green nerve, the inner segment channelled, recurved, face colour dull pinkish-white towards tip, green below with a fine dark green nerve ; lower segments obtuse, the inner one very recurved, channelled, face colour dull white with a fine dark green nerve, the outer segments replicate, -+ revolute, spreading, face colour dull pinkish-white below, green above, with a fine dark green nerve ; stamens 5 and 6mm. long; ovary 3mm. long, 14 mm. diam., green ; style 4 mm. long, dull white, bent, + capitate.
Habitat : Cape Province: Bredasdorp Dist. (Type G. G. Smith, 5479 (Otzen 10) in East London Museum Herbarium.)
Distribution ; not further known.
10 The Journal of South African Botany.
This neat little plant is described from material collected by Mr. M. Otzen in Dec., 1942, some miles west of Bredasdorp, from which district it has been recorded by several collectors. It is nearest to H. mirabilis Haw. in its clear pellucid end-area, but in the new species the end-area is shorter and less pointed, there are no tubercles on back of leaves and the end-awn is shorter.
Haworthia turgida var. pallidifolia G. G. Smith. (Liliaceae-Aloineae.) Sect. Retusae. :
Foha 25 circiter, longa usque ad 27 mm., lata 12 mm., obovata, deltoidea, acuta ; parte retusa longa 1O—11 mm., lata 12 mm., convexa, laevia, lineis longis 3 brevibusque 3—4 percursa ; subtus convexa, laevia, claro-viridia ; carina marginesque integri.
Pedunculus diam. vix 1 mm., longus una cum racemo 20 em. ; pedicelli longi 12mm., diam. 1 mm., obscure virides; bracteae steriles 8—9, longae 4mm.; bracteae fertiles longae 54mm.; perianthium album, longum 1I6mm., basi cylindrico-triquetra diam. 3}mm.; ovarium longum 23 mm., diam. 1} mm., viride ; stylus longus }mm. __
Rosette acaulescent, 4 cm. diam., proliferous from the base and forming clusters.
Leaves about 25, the young erect, the old erect-spreading, + firm, up to 27 mm. long, 12 mm. broad, 7 mm. thick at base of end-area, obovate, deltoid-acute ; end-awn 1} mm. long, naked, persistent, pale green ; face below end-area + convex, smooth, with minute longitudinally oblong flecks, light green, dull; end-area 10—11 mm. long, 12 mm. broad, convex, smooth, light pellucid green, dull, with minute longitudinally oblong lighter to silvery flecks and 3 long lines, one of which reaches the tip, and 3—4 short lines extending only slightly into the pellucid part, the lines distinctly reticulate ; back convex, smooth, light green, dull, with many oblong pellucid spots and longitudinally oblong lighter flecks, and about 12 reticulate lines; keel obliquely arranged in the upper }, entire ; margins rounded above, + acute below, entire.
Peduncle simple, terete, barely 1 mm. diam., 20 cm. long including the few flowered raceme, light brown to light greenish-brown below ; raceme 54cm. long, about 12—14 spirally arranged flowers and buds, 1—2 open simultaneously ; pedicels 12mm. long, 1 mm. diam., dark green ; sterile bracts 8—9, 4 mm. long, the lowest 2 cm. from base of peduncle ; fertile bracts 54mm. long, deltoid, acuminate, white, with a medium- fine brown nerve; perianth white, 16 mm. long, the cylindrical-triangular base 3} mm. across, tapering to 2? mm. above, not stipitate, oblong-ovate, ascending-spreading, curved; wpper segments obtuse, channelled, face
Some New Species and Varieties in the Genus Haworthia V. 11
colour white with a fine green nerve, the 2 outer segments recurved, spreading, the inner one less recurved, shorter ; lower segments obtuse, channelled, face colour white near tip, light green below, with a broad darker green nerve, the outer ones recurved about 90 degrees, the inner one very recurved, longer ; stamens 44 and 4mm. long; ovary 24 mm. long, 1mm. diam., green; style 4mm. long, white, bent, capitate ; capsule 12 mm. long, 3—34 mm. diam., oblong.
Habitat : Cape Province: Riversdale Dist. (Type G. G. Smith 5714, (Dekenah 146), in Kast London Museum Herbarium.)
Distribution : not further known.
This distinctive variety is described from material collected by Mr. J. Dekenah of Riversdale, on the Valsch River, and sent to the author in May, 1944. It is easily grown and soon forms neat clusters. Compared with the species the leaves are longer, broader, thicker and more acute, and have a 14mm. long end-awn. The flecks on the end-area are more numerous, the lines on the back are more conspicuous, those on the face are fewer and less defined and are not as reticulate. In colour, this variety is a lighter green than any other plant so far recorded in the Retusae Section.
Haworthia asperiuscula var. patagiata G. G. Smith. (Liliaceae- Aloineae.) Sect. Trifariae.
Caulis foliatus erectus, longus usque ad 15 cm., latus una cum foliis 25 mm., a basi proliferans.
Folia ordinate trifaria, imbricata, divergentia, longa 20 mm., lata 13—15 mm.; supra concava, parte superiore minute scabrosa, claro- fulvido-viridia, apicem versus obscure viridia ; swbtus convexa, partibus duabus superioribus minute scabrosa, obscure viridia, opaca.
Pedunculus simplex, diam. 1 mm., longus una cum racemo 28 cm., basim versus fuscus ; pedicelli longi 7 mm., viride-fusci ; bracteae fertiles longae 3} mm. ; ovarium longum 3 mm., 14 mm. diam. ; stylus longus 2mm.
Leafy stem erect, up to 15 em. long, 25—28 mm. across including. the leaves, proliferous from the base and forming clusters.
Leaves regularly trifarious, imbricated, spreading, recurved, 20 mm. long, 13—15 mm. broad above the base, 44 mm. thick at the middle, the sheathing leaf base 64—74 mm. long on the side opposite to the lamina, ovate-deltoid, + abruptly acuminate, cuspidate ; face concave, triangular- concave towards the tip, minutely scabrous in the upper half with con- colorous papillae, often with a longitudinal concolorous raised line along- side the middle groove, pink at base, light yellowish-green becoming dark
12 The Journal of South African Botany.
green towards the tip, + shining below to dull above; back rounded, becoming triangular towards the tip, minutely scabrous in the upper 3, sandy, pink at base, dark green above, dull: keel prominent, straight and central in the upper 3, below turned sharply to one side in a long curve, green, horny, + shining; margins acute below, becoming broad
and flattened towards the tip of leaf, green, horny, shining.
Fig. 4, | Hi. asperiuscula var. patagiata. H. Nort eri.
Peduncle simple, terete, 1 mm. diam., 28 cm. long including the raceme, brown below ; raceme 123 cm. long, about -12 spirally arranged flowers and buds, 1 open at a time ; pedicels 7mm. long, barely 1 mm. diam., greenish-brown : sterile bracts 2, 54 mm. long, the lower one 5 cm. from base of peduncle: fertile bracts 33mm. long, deltoid-lanceolate, acuminate, white, with a fine brown nerve : perianth white, brown nerved,
Some New Species and Varieties in the Genus Haworthia V. 13
17mm. long, the cylindrical-hexagonal sub-globose base 3 mm. diam., only slightly constricted above, the base tapering to the pedicel, oblong, spreading, + curved ; upper segments obtuse, channelled, the inner one straight, tip incurved, face colour white at tip to green below with a broad greenish-brown nerve, the 2 outer segments short-recurved-
revolute, not spreading, face colour white with a broad greenish-brown nerve ; lower segments obtuse, the inner one channelled, revolute, face colour white with a greenish-brown nerve ; the 2 outer replicate, recurved about 90 degrees, spreading, face colour pure white at tip, greenish below, with a broad greenish-brown nerve ; stamens 6 and 74 mm. long ; ovary 3mm. long, 14mm. diam., bright green; style 2mm. long, + bent, capitate, yellowish-white.
Habitat : Cape Province ; Willowmore Dist. (Type, G.G. Smith, 2176 in Kast London Museum Herbarium.)
Distribution : not further known.
This variety was collected by the author in Oct., 1937, 15 miles N.E. of Willowmore. Compared with the species, the leaves are more pointed and not as scabrous, the margins are more horny and shining and the pedicels are shorter.
Haworthia Nortieri G. G. Smith. (Liliaceae - Aloineae.) Sect. Laetevirentes.
Rosetta acaulescens, diam. 64 cm.
Foha 43 circiter, longa 37 mm. lata supra mediam partem 12 mm., obovata, acuto-lanceolata; seta extrema longa 13 mm., denticulata ; supra convexa, notis parte superiore irregulariter dispositis oblongis pellucidis percursa, lineisque 7 subrubido-fuscis vel viridibus, reticulatis anastomosentibusque, basim versus rubido-viridibus, apicem versus subrubido-fuscis instructa ; swbtus convexa notis superiore parte oblongis pellucidis percursa ; margines carinaque dentibus pellucide albis ornati.
Pedunculus simplex, diam. una cum racemo sublaxe floreato 1} mm. ; pedicelli longi 2} mm., diam. | mm., fusci: perianthiwm album, longum 15mm.; ovarium longum 3mm., diam. 1}mm., claro-viride ; stylus longus 3mm.
Rosette acaulescent, 64 em. diam., not or very slowly proliferous from the base.
Leaves about 43, the young erect and + incurved, the old ascending and incurved, + firm, 37mm. long, 12mm. broad above the middle, 6mm. thick, obovate, acute-lanceolate ; end-awn 13mm. long, white, persistent, denticulate ; face convex, -- swollen above the middle, smooth, irregularly pellucid in the upper } in solitary and confluent longitudinally
14 The Journal of South African Botany.
oblong pellucid markings, this pellucid area about 18 mm. long, with 7 mauvish-brown or green reticulate and anastomosing lines, 2—3 of which reach the tip, reddish-green below, mauvish-brown above, dull; back convex, smooth, with solitary and confluent longitudinally oblong pellucid markings in the upper 4, and 11 indistinct darker (+ reticulate) and anastomosing lines which are reddish near the tip, the leaf light green below, mauvish-brown above, dull ; margins acute, with 14 mm. long pellucid-white acuminate teeth at middle of leaf, smaller at tip and becoming very small below ; keels 2 in the upper 4, the teeth the same colour as those on the margins but smaller.
Peduncle simple, terete, 14 mm. diam., including the sub-laxly 17- flowered raceme about 27 cm. long, the flowers and buds spirally arranged, light greyish-brown below ; raceme 10cm. long, 3 flowers open simul- taneously ; pedicels 24 mm. long, 1 mm. diam., brown; sterile bracts 9, 8mm. long, the lowest 1} cm. from base of peduncle ; fertile bracts 4mm. long, deltoid, acute, white with a fine reddish-brown nerve ; perianth white, 15 mm. long, the cylindrical-triangular base 4 mm. across, gradually constricted to 3} mm. above, not stipitate, ovate, ascending- spreading, + curved; upper perianth segments obtuse, channelled, re- curved, the inner one short, face colour golden-yellow and white tipped, and with a broad greenish-brown nerve, the 2 outer ones more recurved (about 90 degrees), -- spreading, face colour white with a medium-fine greenish-brown nerve ; lower segments obtuse, channelled, the inner one recurved-revolute, long, face colour white with a medium-fine greenish- brown nerve, the outer segments more recurved (about 90 degrees), spreading, face colour golden-yellow and white tipped, with a broad greenish-brown nerve ; stamens 6 and 5mm. long; ovary 3 mm. long, 1} mm. diam., light green ; style ? mm. long, dull white, bent, capitate.
Habitat: Cape Province ; Van Rhynsdorp Dist. (Type G. G. Smith 1676/a in East London Museum Herbarium.)
Distribution : not further known.
This interesting species was collected by the Author near Doorn River bridge, about 28 miles N.W. of Clanwilliam, in Jan., 1937, the locality having been given by Dr. Nortier, of Clanwilliam, after whom I have much pleasure in naming the plant. The leaf markings and the colour of the flowers are quite distinctive and the plant is thus easily recognised.
“pLoyisuap "IVA DSNJaL * AT DULDILIOS “LRA DSNQIL * ET “DIDIUYUYINUL “IVA DSNGQAL * ET LO(DUL “IBA DUDYPINYIS! * EL ‘TT fivtg
"La MON * ET “pyorboyod “ea pynosniiadsp “HY ‘pyofupynd *1er vprbing > Ay “DINPUNW “FT ‘Tl tiv
DIE SAADKNOP EN EMBRIOLOGIE VAN BRA- BHIUM STELLATIFOLIUM, L.
Drur P. G. JoRDAAN.
Die materiaal wat vir die studie van die wilde amandel (Brabeiwm stellatifolium) gebruik is, is langs Eersterivier en die meulsloot (Stellen- bosch) versamel, waar die plante in hul natuurlike staat voorkom.
Die plante blom in Desember. Die blomme en vrugte ontwikkel vinnig. Teen die end van Desember is byna al die blomme tot niet en teen die end van Februarie is al die vrugte volwasse.
1. Diz VooR-ONTWIKKELING VAN DIE SAADKNOP.
In elke vrughok ontstaan twee saadknoppe gelyktydig langs mekaar op dieselfde hoogte op die buiknaat. Die saadknoppe ontstaan feitlik in die top van die vrughok. [Fig. 1.]
Op ’n vroeé stadium groei die aanleg van die saadknop in die rigting van die blombodem en kry sodoende ‘nlangwerpige en geboé vorm. Hierna word die saadknopvliese aangele. Hers ontstaan die binnevlies as ’n kraag om die basis van die aanleg [fig. 2] en daarna ontstaan die buite- vlies as “n tweede kraag tussen die binnevlies en die placenta. [Vel.
Beide vliese is by hul ontstaan twee sellae in deursnee. Wanneer die makrospoor-moedersel uitgedifferensieer is, is die kerntepel vier sellae hoog en word die moedersel van die makrospore aan elke sy deur drie tot vier lae nucellusselle omring. As die moedersel van die makrospore gevorm is, kan ’n dekweefsel onderskei word. [Fig. 4.]
Wanneer die blomme oopgaan, is die saadknop -30 mm. lank en het die binnevlies verby die top van die nucellus gegroei en ‘n kort endostoomkanaal gevorm. Die binnevlies is nou drie sellae in deursnee.
2. Diz VOLWASSE SAADKNOP.
Wanneer die saadknoppe volwasse is [fig. 7], is hulle 440 mk. in lengte. In elke vrughok is twee normale, feitlik hangende saadknoppe wat aan die buiknaat [fig. 8] en bo in die vrughok vas is. Daar is geen raphe nie en “n funiculus is nie te onderskei nie. Hoewel die chalaza effens krom
is, kan die saadknop as ortotroop beskrywe word.
16 The Journal of South African Botany.
Die nucellus is -17 mm. lank en is aan sy basis nou heelwat wyer (-12 mm.) as voor die blomme oopgegaan het. In die kerntepel, wat ongeveer vier sellae hoog is, is n dekweefsel beswaarlik te onderskei. In die kerntepel sowel as om die sye en die basis van die kiemsak bestaan die nucellus uit isodiametriese of veelkantige selle. Om die kiemsak is die nucellus 4—6 sellae in deursnee. [Fig. 9.] In die basale sye van die nucellus is die selle reghoekig. Behalwe die een of meer uitgeteerde en (of) half-verworde selle teen die kiemsak, is al die selle plasmaryk en teelvermoénd.
Die endostoomkanaal is ongeveer 60 mk. lank en die deel van die binnevlies wat dit omring, is vier sellae in deursnee en effens dikker as die res van die vlies. Die binneylies is origens drie sellae dik en die selle is almal eenders. [Fig.9.] Die buitevles is twee sellae [fig. 9] in deursnee en bereik nie die top van die binnevlies nie.
Die chalaza bestaan uit plamaryke selle en grens direk teen die nael. *n Vaatbundel tree deur die nael direk in die chalaza. Die vaatbundel, waarin trachelede gedifferensieer is, loop dood ‘n end verwyder van die basis van die nucellus sonder om te vertak. Die chalaza is massief in vergelyking met die res van die saadknop en is feitlik net so lank soos die nucellus. Die saadknop bevat geen basaalweefsel nie en geen grens kan tussen die chalaza en die nucellus onderskei word nie.
3. Diz Voor-ONTWIKKELING VAN DIE KIEMSAK.
Net na die binnevlies ontstaan het, is die makrospoor-moedersel in die nucellus te onderskei. [Fig. 2.] Die sel vergroot, word meer promi- nent [fig. 3 en4] en wanneer die blomme hul volwasse lengte feitlik bereik het, verdeel dit en vorm ‘n liniére reeks van vier baie klem makrospore. [Fig. 5.] Een van die spore, vermoedelik die chalazale een, gee oorsprong aan die kiemsak. Die makrospoor-moedersel is die enigste sel in die nucellus wat sitologies verskil van die aangrensende selle.
Uit die primére kiemsakkern ontwikkel ag kerns wat, gewoonlik voordat die blomme oopgaan, georiénteer is soos in die volwasse kiemsak. ‘n Kenmerk van die kiemsak is dat sy ontwikkeling vanaf die volwasse makrospoor-moederselstadium tot die volwasse stadium baie vinnig verloop.
4. Diz VotwassE KiEmsak.
In die vars, oop blomme is die kiemsak plasmaryk, 100—130 mk. lank en 36—48 mk. breed op sy wydste. Die kiemsak is chalazaalwaarts en mikropielwaarts toegespits. [Fig. 6 en 7.]
Die chalazale punt van die kiemsak word ingeneem deur drie plas- maryke, eenkernige, antipodale selle met gemeenskaplike dwarswande.
Die Saadknop en Embriologie van Brabeium stellatifolium, L. 17
Die selle lé soms in ’n vertikale reeks. [Fig.6.] Meestal lé een sel in die uiterste punt en lé twee teen hom. [Fig. 11.] Namate die blomme ouer word, trek die protoplasma weg van die selwande en word deursigtiger.
Die twee plasmaryke polare kerns lé styf teen mekaar net bokant die middel van die kiemsak. [Fig. 6 en 7.] Na die blomme ’n rukkie oop is, is die polare kerns meestal half versmelt. [Fig. 9.]
In die toppunt van die kiemsak lé die drie plaasmaryke selle van die eler-appraat naastendy op dieselfde hoogte en styf teen mekaar. [Fig. 6 en 10.] Die synergidae word meer langwerpig namate die saadknop ouer word. Die skerp toppunte van die ou synergidae is plasmaryker as die res van die sel.
5. Dre KiEmwit.
Wanneer die vrug begin te ontwikkel, neem die saadknop vinnig toe in grostte. Gepaard met die aanwas van die saadknop gaan ’n buiten- gewone toename in lengte van die kiemsak. As die kiem 62 mk. lank is, is die kiemsak 1-6 mm. lank en as die kiem 2:7 mm. lank, is die kiem- sak 10 mm. lank. [Fig. 17.] Die kiemsak baan sy weg deur die nucellus tot in die chalaza wat deur sekondére groei in lengte en breedte toeneem. Die onderpunt van die kiemsak groei nie eweredig nie met die gevolg dat die kiemsak uitstulpings (vertakkings) ontwikkel. [Fig. 14 en 17.] Die chalazale deel van die kiemwit-bevattende deel van die kiemsak is gewoon- lik die wydste en hierdie deel is meestal met ’n nou hals verbind met die deel om die kiem. In jong na-bevrugtingstadia lé die kiemsak los van die omringende weefsels en kan dit met min moeite uitgedissekteer word.
Die kiemwit vorm eers ‘n lagie met vry kerns teen die wand van die kiemsak. [Fig. 12 en 13.] Wanneer die saadknop 4 mm. en die kiemsak 3 mm. lank is en die kiem aktief begin te ontwikkel, vorm die kernkiemwit wat teen die kiem grens en waar dit die kiemsak van kant tot kant vul, selwande. Op hierdie wyse ontstaan duidelik gedifferensieerde kiem- witselle. Aanvanklik is die selle plasmaryk en meestal bevat elk esel net een kern. [Fig. 14 en 15.]
Namate die kiem verleng, word die kiemwit teenaan die kiem geab- sorbeer en word nuwe kiemwitselle chalazaalwaarts gevorm. Die absorbsie en aanwas van die selkiemwit hou aanvanklik tred met mekaar en op geen tydstip is daar ‘n groot volume selkiemwit aanwesig nie. Kiemwit word nooit tussen die kiemblare gevind nie. [Fig. 17.]
As die saadknop 12 mm. en die kiem 6 mm. lank is, bevat die kiemsak ‘n klein hoeveelheid half-uitgeteerde selkiemwit. Die basale helfte van die kiemsak is dan nog uitgeklee met “n wandbeleg van kernkiemwit. Voordat die kiem sy volwasse lengte bereik, is die selkiemwit en kern- kiemwit, behalwe vir ’‘n vormlose vliesie hier-en-daar, geabsorbeer.
18 The Journal of South African Botany. 6. Dre Kiem.
Die stuifmeelbuis groei deur die endostoomkanaal en baan °n weg deur die dekweefsel. [Fig. 12.]
Die kiemsak is meer as 700 mk. lank en bevat al heelwat kernkiemwit voordat die bevrugte eier [fig. 12] begin te verdeel. Geen kiemdraer kon onderskei word nie. [Vel. fig. 13.]
Namate die kiemselle verméerder, word die kiem langwerpig en toe- gespits in die rigting van die mikropiel. Hers as die kiem ongeveer 130 mk. lank is, word dit gedifferensieer in “n dermatogeen en ’n proto- meristeem. Die dermatogeen word nie oor die skerp punt van die kiem gevorm nie. [Vel. fig. 15.] Die kiem vul die mikropilére deel van die kiemsak van kant tot kant.
Die kiemblare word eers aangelé nadat selkiemwit teen die kiem gevorm het en wanneer die. kiem meer as 300 mk. lank is. [Fig. 15.] Teen dié tyd is die kiemsak al ruim 3-7 mm. lank. Die kiemblare ont- staan deurdat in twee teenoorgestelde sye van die wye deel van die kiem groei vinniger geskied as in die res van die kiem. Wanneer die aanleg van die kiemblare onderskei kan word, kan die stam-aanleg gewoonlik ook onderskei word. [Fig. 15.] Reeds by hulle ontstaan is die aanlegte van die kiemblare en stam bedek met ’n dermatogeen wat aaneenlopend is met dié van die res van die kiem.
Die kiemblare ontwikkel vinnig en is spoedig langer as die kiem- wortel. Op ’n vroeé stadium word die protomeristeem van die kiemblare gedifferensieer tot ‘n pleroom en peribleem. - Die pleroom word uitge- differensieer as “‘n onvertakte desmogeenbundel vanaf die pleroom van die kiemwortel. Later vermeerder die desmogeentakke in die kiemblaar.
Die kiemblare is spoedig ook heelwat wyer as die kiemwortel. As die kiem ongeveer 2-5 mm. lank is, begin die selle in die basis van die twee sye van elke kiemblaar teen die kiemwortel aktief te groei en elke kiemblaar vorm sodoende twee lobbe wat afwaarts om die kiemwortel groei. [Vgl. fig. 19—25.] In die volwasse kiem strek hierdie lobbe teen die sye van die kiemwortel soms verby die punt van die wortelmus. [Vel. fig. 27.]
Wanneer die aanleg van ’n kiemblaar te onderskei is, is “n differen- siasie in die protomeristeem van die kiem waar te neem. Op die mediaan van die kiem, naby die spits punt, is die inisiaalgroep duidelik te onder- skei. [Fig. 15.] Chalazaalwaarts van die inisiaalgroep word die pleroom uitgedifferensieer. Die differensiasie begin teen die inisiaalgroep en word in die rigting van die kiemblare voortgesit. Die pleroom en die binnenste lae van die peribleem loop uit op die inisiaalgroep.
Byna onmiddellik na sy ontstaan is die pleroom van die kiemwortel
Die Saadknop en Embriologie van Brabeium stellatifolium, L. 19
gedifferensieer in “n endisteem wat teen die inisiaalgroep grens, en “n desmogeensilinder. [Fig. 17, 20 en 21.] Ongeveer halfpad tussen die inisiaalgroep en die spits van die pluimpie verval die desmogeensilinder eers in twee, daarna in vier [fig. 22] en aan die basis van die pluimpie [fig. 23] in ’n hele aantal desmogeenbundels. Sommige van hierdie desmogeenbundels is kontinu met dié van die kiemblare.
Wanneer die inisiaalgroep uitgedifferensieer het, is die wortelmus ook afgebaken. [Fig. 15.] Die selle in die punt van die wortelmus is meer deurskynend as die aangrensende selle. Voordat die kiemblare baie ver ontwikkel het, kan ‘n kaliptrogeen in die basis van die wortelmus onderskei - ‘word. [Fig.17.] Die kaliptrogeen word uit die buitenste lae van die peribleem, en tot ‘n geringe mate uit die inisaalgroep, gevorm. Die wortelmus is ’n prominente deel van die kiemwortel.
Nadat die kiemblare ’n ent ontwikkel het, vorm die stam-aanleg ’n boggel wat dan duidelik as die pluimpie uit te ken is. [Fig. 17.] Die pluimpie bly kort en in die vlak ewewydig met die plat kante van die kiemblare is dit die wydste. [Fig. 25.] Die pluimpie ontwikkel voor ontkieming nie verder as die protomeristeemfase nie.
In die volwasse saad is die kiem naastenby 3 cm. lank en ongeveer een-derde van die lengte van die kiem lé in die chalaza. Die kiemblare is hard en hul lé so styf teen mekaar dat hul beswaarlik van mekaar te skei is. Die lobbe van die kiemblare lé ook so styf teen mekaar dat ook hulle met moeite afsonderlik te onderskei is.
Setmeel is die belangrikste reserwestof in die kiem. ‘n Groot hoeveelheid eiwit word ook gebére.
7. Drm NUCELLUS NA BEVRUGTING.
Na bevrugting gaan een saadknop altyd tot niet. Dikwels teer albei saadknoppe uit.
In ’n saadknop wat na bevrugting normaal ontwikkel, bly die nucellus lank meristematies. In die sye word die selwande reélmatig aangelé en verdeel die selle hoofsaaklik periklinies. In die basis van die nucellus is die verdelings in alle rigtings. Die nucellusselle vergroot en hul toename in lengte is veral opvallend. Terwyl die nucellusselle groter word, bly die selle van die chalaza aanvanklik klein met die gevolg dat kort na bevrug- ting die nucellus duidelik van die chalaza onderskei kan word. Soms is ’n grens tussen die chalaza en nucellus te onderskei voordat die selle in die basis van die nucellus verleng.
Wanneer die kiemsak die chalaza bereik, en namate hy sy weg deur die chalaza baan, verleng die chalazale selle in die nabyheid van die kiemsak. Hierdie langwerpige chalazale selle verskil nie van die nucel-
20 The Journal of South African Botany.
lusselle nie met die gevolg dat, wanneer die onderpunt van die kiemsak eers in die chalaza is, geen grens tussen laasgenoemde en die nucellus te onderskei is nie.
Die boonste helfte van die nucellus bly teelvaardig totdat die kiemblare aangelé word. Op hierdie stadium bestaan hier nog enkele lae plas- - maryke nucellusselle. [Vgl. fig. 16.] As die kiemblare ontstaan, is die basis van die nucellus nog aktief verdelend. Die basis van die nucellus behou sy meristematiese toestand totdat die kiem sy volwasse grootte feitlik bereik het.
Wanneer die kiem 5-2 mm. lank is, het die nucellus om die mikro- pilére helfte van die kiem heeltemal verdwyn. Verder chalazaalwaarts bestaan nog nucellus wat in die rigting van die chalaza in omvang toe- neem. In die volwasse saad bly hoogstens ‘n vormlose vlies van die nucellus oor. [Veg]. fig. 26.]
Die kerntepel bly bestaan lank nadat die nucellus om die mikropilére deel van die kiem geabsorbeer is.
9. Diz SAADHUID.
Namate die saadknop na bevrugting ontwikkel, vul dit die hele saadknop en lé die buitevlies mettertyd so styf teen die vrugwand dat dit beswaarlik van laasgenoemde te onderskei is.
Behalwe in die basis bly die buitevlies op die meeste plekke twee sellae in deursnee. [Vgl. fig.-16.] In na-bevrugtingstadia bevat die buitevlies byna altyd kleurstowwe wat of in die buitenste of in die binnenste sellaag of in al die lae voorkom. Vandat die kiem ongeveer 3mm. lank is, is die selle van die binnenste laag gewoonlik meer plas- maryk [fig. 16] en mettertyd is hul inhoud bruin gekleurd. In die volwasse saad bestaan die laag gewoonlik nog en is daar geen wesenlike verskil tussen sy selle en dié van die vrugwand nie. ‘n Tweede sellaag van die buitevlies bly ook soms bestaan. [Fig. 26.]
Die getal sellae in die binnevlies vermeerder nie na bevrugting nie, behalwe heel in die basis. [Vegl. fig. 14 en 17.] Na die kiem ongeveer 2 mm. lank is, begin die selle in die buitenste laag te verleng en gaan dan tot niet. As die genoemde laag verword het, gebeur dieselfde met die middelste laag. [Vgl. fig. 16.] Later gaan die binnenste laag ook tot niet en bly net “n vormlose vlies oor wat saam met die oorblyfsels (indien enige) van die nucellus en kiemwit die binnehuid vorm. [Fig. 26]. In die binnehuid is geen kutikula nie. -“n Swak ontwikkelde kutikula is soms in die jong na-bevrugtingstadia in die binneste sellaag van die binnevlies te onderskei.
Na bevrugting bly die chalaza meristematies totdat die kiem sy
Die Saadknop en Embriologie van Brabeium stellatifolium, L. 21
volwasse lengte bereik. Wanneer die kiemsak die chalaza bereik en sy weg daardeur baan, verloor die selle in die nabyheid van die kiemsaak hul teelaktiwiteit, vergroot en dié wat teen die kiemsak grens, word geabsorbeer. Die teelaktiwiteit word sodoende beperk tot die sye van die chalaza.
Namate die saadknop verleng, vertak die vaatbundel en hierdie takke verloop in die buitenste deel van die chalaza tot by die basis van die binnevlies. Gewoonlik het baie selle aan die binnekant van die vaat- bundelwyk ’n rooierige inhoud, terwyl dié aan die buitekant van die vaatbundelwyk ’n bruinerige inhoud bevat. [Fig. 14 en 17.]
As die saad volwasse is, is die chalaza ongeveer 1-5 cm. lank en bestaan die vaatbundelwyk en die bruingekleurde deel in die buitekant van die chalaza nog. Tussen die vaatbundelwyk en die kiemblare bestaan nog enkele lae onreélmatige en uitgeteerde parenchymatiese selle.
SUMMARY.
The structure and the development of the ovule, embryo-sac, embryo and seed of Brabeium stellatifolium, L., the only species of the genus Brabeium, are described.
The ovary contains two sessile, orthotropous ovules on the ventral suture. The ovule is without a raphe and has two integuments. In the full-grown ovule the nucellus consists of less than seven layers of cells around the embryo-sac. After fertilisation the nucellus and chalaza remain meristematic and the secondary cells thus formed are absorbed.
The testa is membranous, without a crystal layer and no cuticula develops from the inner integument.
The embryo-sac develops normally and contains eight nuclei or cells. The part of the embryo-sac which lies in the chalaza after fertilisation, is branched. Cellular endosperm is never formed in the tips of the branches. The ripe seed is exendospermous.
The two large, hard cotyledons have basal lobes and the endistem of the radicle is clearly differentiated up to the initial group. The calyp- trogen is a product of the external layers of the periblem. The embryo stores starch and proteins.
25
74,
The Journal of South African Botany.
VERDUIDELIKING VAN DIE VERKORTINGS WAT AS BYSKRIFTE BY DIE TEKENINGE GEBRUIK WORD.
a.pl. a.s. a.sk. bb. bee. bin.h. b.kb. bl.
bs. buit.h. ch., ch.s. d.
des. dw.
e.
en. end.k. ev.
m. mer. m.m,.
aanleg van pluimpie.
antipodale sel.
aanleg van saadknop.
blomdekblaar.
bevrugte eler.
binnehuid.
basis van kiemblaar.
blomsteel.
blomskub.
buitehuid.
chalaza.
dermatogeen.
desmogeen.
dekweefsel.
eier.
endisteem.
endostoomkanaal.
binnevlis.
flap van kiemblaar.
chalazale makrospoor.
inisiaalgroep. ‘
kiem.
kiemblaar.
kaliptrogeen.
kernkiemwit.
kiemsak.
kerntepel.
kiemwit.
een van die lobbe aan die onderpunt van die kiem- sak.
makrospoor.
kernmeristeem.
mikropilére makrospoor.
m.ms. nue.
makrospoor-moedersel.
nucellus.
*n deel van die saadknop wat aan die buitekant van die vaatbundels voorkom en wat selle met ‘n bruin inhoud bevat.
onderpunt van kiemsa’:.
pleroom.
peribleem.
polare kerns.
pluimpie.
protomeristeem.
raphe.
synergida.
stuifmeelbuis.
sekondére kiemsakkern.
selkiemwit.
top van kiemsak.
buitevlies.
vrugblaar.
vaatbundel.
vrughok.
vrugwand.
kiemwortel.
wortelmus.
placenta.
‘n deel van die saadknop wat aan die binnekant van die vaatbundels yvoorkom en wat selle met *n rooi in- houd bevat.
die binnenste basis van die binnevlies.
f D Mil ogo OF poo Y O 7 ‘ e Certs ae 0 OR icntets
fee 7,
L=7 A
oJ a
wes
Fig. 6 Fig. 7
Fic. 1. *n Lengtesnit deur die vrugbeginsel en blombodem van ‘n blomknop. Die blom is ongeveer 1-3 mm. lank. bs. = die skottelvormige blomskub. 110. Fic. 2. ’n Jong saadknop waarin die buitevlies nog nie begin ontwikkel het nie. Die blom is ongeveer 2-5 mm. lank. 270.
Fic. 3. ’n Jong saadknop. Die blom is ongeveer 5mm. lank. x 270. Die nucellus met die makrospoor-moedersel van fig. 3. 550.
. Vier makrospore. Die protoplasma van die middelste twee is aan die verword. Die blom is ongeveer 7-5mm. lank. x 1100.
Fie. 6. ’n Volwasse kiemsak. 550.
Fic. 7. ’n Volwasse saadknop. » 110.
Fie. Fic.
Fic. Fie.
Fie. Fig. Fic.
The Journal of South African Botany.
Fig. 12 Fig. 18. Fig. 14.
8. ‘n Dwarssnit deur *n volwasse vrugbeginsel, deur die naels van die twee saadknope. 110.
9. “n Dwarssnit deur ’n deel van ’n volwasse saadknop. Die twee polare kerns
het gedeeltelik versmelt om die sekondére kiemsakkern (sk.) te vorm. 550.
10. ’n Dwarssnit deur die toppunt van ’n kiemsak. 550.
11. Die onderpunt van ’n volwasse kiemsak. Die protoplasma van die antipodale selle is besig om af te rond (of te verword ?). x 1000.
12. Die top van ’n saadknop met ’n bevrugte eier (b.e.) en ’n stuifmeelbuis (sb.). Vrugte 3-8 mm. lank. x 180.
13. Die toppunt van ’n kiemsak met ’n viersellige kiem. Die sel x van die kiem 1é¢ reg bo die vierde kiemsel. x 550.
14. “n Saadknop in ’n vrug wat 10 mm. lank is. x. = “n onverteerde deel van die chalaza tussen die kiemsaklobbe. 22.
Die Saadknop en Embriologie van Brabeium stellatifolium, L.
i) Ou
skw.
a-pl.
tv.
Fig. 16.
Fie. 15. *n Jong kiem in ’n vrug wat 13mm. lank is. Die pleroom (p.) is nog
ongedifferensieer. x 220. Fic. 16. ’n Deel van *n saadknop om ’n kiem in ’n vrug wat ongeveer 1-7 cm. lank is. xX. = van die verworde nucellusselle om die kiem. 360.
Fie. 17. *n Saadknop in ’n vrug wat 2cm. lank is. 18.
26 The Journal of South African Botany.
Fig. 28. Fig. 26 -
Figure 18 tot 25 is tekeninge van dwarssnitte van ’n kiem wat ongeveer 6 mm. lank was. Die dermatogeen is in geen van die tekeninge ingeteken nie.
Fie. 18. ’n Dwarssnit van die wortelmus. 18.
Fie. 19. ’n Dwarssnit van die kiemwortel deur die inisiaalgroep. 18.
Fig. 20. “n Dwarssnit van die kiemwortel naby die inisiaalgroep. 18.
Fie. 21. “n Dwarssnit van die kiemwortel agter die rand van die wortelmus. 18.
. . . & . . . .
Fie. 22. ’n Dwarssnit van die kiemwortel onderkant die basisse van die kiemblare en ongeveer half-pad tussen die inisiaalgroep en die spits van die pluimpie. 18.
Fie. 23. °n Dwarssnit deur die kiem by die basisse van die kiemblare. 18.
Fie. 24. “n Dwarssnit deur die basisse van die kiemblare. X18.
Fig. 25. “n Dwarssnit van die kiem deur die pluimpie. x 18.
Fie. 26. Die saadhuid en ’n deel van die vrugwand van ’n byna volwasse vrug. Die selle van die binnenste selJaag (b.tv.) van die buitehuid en die meeste selle van die vrugwand bevat ’n bruin inhoud. x 220.
Fie. 27. ’n Deel van ’n Jengtesnit deur die sy van ’n kiem. x 14.
THE DEVELOPMENTAL MORPHOLOGY OF THE POLLEN OF PROTEACEAE.
By S. GaARSIDE.
Over a long period of years, the pollen grains of Gymnosperms and Angiosperms have been studied entirely from a morphological stand- point, most of the researches into their development and physiology having been made comparatively recently. A comprehensive historical review, more especially of the morphological work, has been given by Wodehouse (1935, pp. 16-100).
Most of the earlier workers were concerned with purely descriptive accounts of the mature grains, or with their comparative morphology, and it remained for Wodehouse (1935, 1936) to put forward a general theory of the evolutionary relationships of pollen forms in Gymnosperms, Monocotyledons and Dicotyledons.
Whilst this clever theory has many ramifications, its more Recor ial features are that the monocolpate or one-furrowed grain which is so prevalent in Monocotyledons and a few of the primitive Dicotyledons, is the direct homologue of the monocolpate grain found in many Gymno- sperms, whilst the tricolpate, or three-furrowed grain so prevalent as the basic form in Dicotyledons is a new development, the three furrows in this case being meridional, at right angles to the spore equator, and alternating with the triradiate marking of the inner or proximal face of the spore when still in the tetrad. (Fig. 15.)
In the one-furrowed grain of Monocotyledons, the furrow is said to be formed on what is the distal or outer face of the spore in the original spore tetrad. In pollen without furrows, but with one germ-pore only (as in grasses) or with three germ-pores (as in some Dicotyledons) the pores are regarded as modifications of the furrows and occupy similar positions in each case.
According to this theory then, the great evolutionary advance made by the Dicotyledons is in the development of three equatorial furrows or pores, resulting in the possibility of pleurosiphonic germination, that is of several possible places of exit for the pollen-tube, situated on the equator, as opposed to basisiphonic germination, or germination from one pore only (or furrow) situated on the outer or distal face of the grain.
Wodehouse has quoted very few researches into the development of the mature pollen grain from the spore tetrad stage, especially researches in which the relative positions of the furrows or pores have actually been
28 The Journal of South African Botany.
observed, and indeed this could scarcely have been expected of him, as very little really precise knowledge of this kind exists.
He has therefore had to rely chiefly upon evidence obtained from the positions of furrows or pores in such grains as retain the tetrad grouping until maturity, e.g. Drimys with one pore on the distal face of each grain of the tetrad, and Erica with pores in six pairs. (Fig. 15.)
Cytologists usually lose interest in the pollen grain during the period between the completion of tetrad division and germination, consequently very few workers trouble to figure the changes undergone by the cell walls of the grains during their maturation, although in most cases adequate material for the purpose must have been available.
Thus a great deal of information of use to the pollen morphologist is not forthcoming, and lack of researches of this nature is doubtless the reason why Wodehouse has been forced to make so many assumptions (e.g. Wodehouse, 1935, pp. 334, 336 and 338) in support of his theory.
Doubt has already been cast upon the assumption that all monocolpate grains are homologous, e.g. the monocolpate grains of Asimia angustifolia A. Gray (Anonaceae) have been shown by Golub to have the furrow on what is the proximal or inner face of the spore when in the tetrad (Bailey and Nast, 1943) and it is now obvious that investigation of almost every case will be necessary before exact homologies can be elucidated.
The present work was undertaken in order to determine the relative positions of the germ-pores in the tetrad resulting from the division of the spore mother cell in Proteaceae ; only the gross developmental features of the cells concerned were examined, cytological details being neglected.
From an extensive series of measurements of numerous genera and. species of South African Proteaceae, the writer found that Leucadendron argenteum K.Br. has the largest mature pollen grain (45u to 61u in dia- meter) and this plant was therefore chosen for investigation as it was likely to have a large spore tetrad. In addition, it has three large circular germ-pores (11-5u to 12u in diameter) one at each angle of the flattened triangular grain. (Fig. 8.)
Tetrad division stages in L. argentewm are to be found in the anthers in June, when the young spherical knob-like spike of the male inflores- cence is only about I cm. in diameter. Usually a developmental series can be found in a single inflorescence, as the lower flowers of the spike are considerably more developed than the upper ones.
Material preserved in 75% alcohol was found adequate for all purposes ; from it the spore tetrads can be dissected out entire and moved about and rotated under the microscope to obtain various views of the same
Ik Clilwo = IO fhe
Fras. 1—10. Leucadendron argenteum R.Br. 1. Spore mother cells. 2, 3, 4, and 5.
Tetrahedral spore tetrads. 6. Young microspore immediately after liberation from tetrad. 7. Young pollen grain, showing differentiation of exposporium. 8 and 9. Pollen grains showing early stages of germination. 10. Germinating pollen grain with pollen-tube. Fras. 11—13. Embothrium coccineum Forst. 11 and 12. Decussate tetrads from two points of view. 13. An almost mature pollen-grain. The scale is for all the figures.
30 The Journal of South African Botany.
tetrad. Microtome sections were not made, and no staining was necessary, though light staining with acetic methyl-green was of value. All the drawings have been made to the same scale so that the relative sizes of the various stages can be seen at a glance. The germination studies were of course made on fresh material.
From the younger anthers of the above-mentioned material, it is easy to isolate groups of spore mother cells (Fig. 1) when each cell is seen to be surrounded by a broad, clear gelatinous zone. In anthers in which tetrad division has taken place (Figs. 2 and 3) the spore mother cells have separated and the gelatinous wall becomes rounded off (Figs. 3, 4 and 5) and the tetrads are completely isolated. The young spores which compose the tetrad always have the tetrahedral arrangement and no deviation from this was observed. Various views of tetrads are shown in Figs. 3, 4 and 5; Fig. 5 is of special interest as at first glance it would appear to be an example of the decussate arrangement, but by revolving such a tetrad under the cover-glass, a view similar to that in Fig. 4 will appear. It is, in fact, only a somewhat unfamiliar view of the tetrahedral arrangement.
The clear gelatinous wall surrounding the spore tetrad now dissolves, and the spores readily separate (Fig. 6) without other change than that of rapid enlargement and thickening of the spore wall. When the exospore and endospore are differentiated (Fig. 7) it can easily be observed that the germ-pores are present on the angular tips of the spore (Fig. 7), i.e., they are on the spore equator and terminate the lines of the tri-radiate marking, which does not, of course, persist in the mature spore. Fig. 14 gives a diagram of the spores in the tetrad to illustrate the position of the pores in a spore in relation to the tri-radiate marking, and also to show the grouping of the pores in four groups of three, one group of three not being visible in the figure.
Fig. 14 should be compared with Fig. 15, which represents a tetrad of Erica, in which the germ-pores are seen to be alternating with the tri-radiate marking, and are grouped in six groups of two, an arrangement which Wodehouse appears to have considered to be typical of all tricolpate and three-pored grains in Dicotyledons.
It is clear therefore, that most of the tricolpate and three-pored pollens will require investigation before we know their real nature, and following the type of nomenclature used for embryo-sac types, we may call them after the genera or families in which they were first discovered.
The type shown in Fig. 15 has been established for Epilobium by a study ofits development by Luerson (quoted by Wodehouse, 1935, p. 91) and might be called the “ Epilobium type,” whilst the development which follows Fig. 14 could be called the “ Proteaceous type.” Such a
The Developmental Morphology of the Pollen of Proteaceae. 31
terminology will only be of value until a wide survey of developmental types has been made, when we may expect the types to fall into some logical sequence, and it will then no longer be necessary.
In the writer’s opinion, the proteaceous type may represent a very primitive spore type in Dicotyledons, as the grain has become very little altered from its original tetrahedral shape given to it in the tetrad division, and in this respect resembles the spores of many liverworts and of Sphagnum and some ferns. The question then arises as to the germ- pore in the proteaceous type being more primitive than the fold or colpa frequently found in spherical pollens. The fold or colpa is, as Wodehouse has clearly indicated, primarily an organ of size adjustment to the water
Fria. 14. Tetrahedral pollen tetrad of Proteaceae. Fia. 15. Pollen tetrad of Erica. Both figures are diagrammatic.
content of the grain, and appears to be a decided advance on the simpler pollen of Proteaceae which has no such obvious specialisation. In any event, it is clear that the fold in many pollens is, because of its position, not the homologue of the pore in proteaceous pollen.
It is noteworthy that Thompson in his brief treatment of pollen grains (Thompson, 1942) remarks that the three folds “are formed (I take it) along lines of weakness at the edges of the tetrahedron,” and it may be assumed that they would be more likely to form in grains in which an early thickening of the exosporium (preserving the tetrahedral form) does not occur. However, in making any assumptions regarding evolu- tion, we must bear in mind Thompson’s warning (loc. cit., p. 613) that resemblances in spores are “a matter of physical and mathematical symmetry, and carry no proof of near relationship or common ancestry.”
32 The Journal of South African Botany.
GERMINATION.
As the writer’s object was to observe germination, and not to investi- gate the physiological conditions necessary for it, only a few pollen cultures were made.
Germination was slow in all cases, and comparatively few grains in each culture showed the first stages of germination, and fewer still formed a long pollen tube. 50, 20, 10 and 5% cane sugar solutions in water were used, the lower concentrations giving the best results.
No doubt the addition of ripe stigmas, yeast or traces of mineral salts might have- greatly stimulated germination, but these experiments remain for future workers.
From the morphological viewpoint, it is to be noted that not more than one tube was ever obtained from one grain, and that there is no ‘* dehiscence ~ or splitting away of the exospore on germination, only occasionally a slight tearing or cracking of the margins of the pore from which a tube has emerged, is to be observed. (Figs. 9 and 10.)
However, a very remarkable feature was the extreme thickness of the cellulose wall of the pollen tube (Fig. 10) which may be up to 5u thick ; a feature which has also been noticed by the present writer in other proteaceous pollen tubes, e.g., Grevillea oleoides Sieb., where the tube is very irregularly thickened. Tassi (1898) has also figured the pollen tube of Stenocarpus as having fairly thick walls.
In concluding this account of the three-pored pollen of Leucadendron argenteum, we must make a brief reference to the work of Brough (1933), who has investigated the development of the microspores of Grevillea robusta Cunn. After describing the miscospores as of “ a precise tetrahedral form in the nearly mature flower bud ” (loc. cit., p. 51), he says that the ‘“ thickening of the exine is not uniform however, and is distributed in such a manner that four unthickened areas are evident at the points of the tetrahedron,’ and also says that ‘“‘the consequent increase in internal pressure is no doubt the cause of the protrusion of the intine at the four corners of the tetrahedron.” to suppose that there are here four germ-pores, and the writer has there-
These expressions undoubtedly lead one
fore re-examined the pollen of G. robusta, and finds it to be of the normal three-pored type. Brough’s error is probably due to a lapsus calami, as his figures of the microspores all show three pores only.
BreorRosE POLLEN.
We must now briefly consider the case of proteaceous pollen in which the grain is ovoid or sausage-shaped, and in the latter case often somewhat arcuate, with one pore at each end.
The Developmental Morphology of the Pollen of Proteaceae. Be
These peculiar two-pored grains are known for the genera Banksia, Dryandra and Embothrium, in which this form of pollen is a generic character in each case; no examples of Proteaceae are yet known in which both two-pored and three-pored pollens occur in one genus or in one species (dimorphic pollen).
In the material examined, Hmbothriwm coccineum Forst., the dividing spore mother cells invariably show the decussate arrangement of the young spores in the tetrad (Figs. 11 and 12), and it is clear that the two germ-pores must arise one at each end of the ovoid cells, as the latter undergo very little change of shape when increasing in size to become the mature ovoid pollen grains. (Fig. 13.)
Wodehouse (1932) in a short description, accompanied by a beautiful figure of the two-pored pollen grain of Banksia Candolleana Meisn. remarks, “‘ It appears that this grotesque form is firmly established in the family, and is not due merely to some individual or specific anomaly in the arrangement of the pollen cells in their tetrads as the form suggests. Its origin in relation to tetrad formation certainly deserves further study.” Now that such a study has been made and it is possible to compare it with the development of three-pored pollen, it appears that the form of the grain in biporose ‘pollen is much more closely bound up with the arrangement of the spores in the tetrad than Wodehouse had supposed it to be.
Although Embothrium is the only genus so far studied, it appears probable that all biporose pollen of Proteaceae will have the decussate arrangement of the spores in the tetrad, as opposed to the tetrahedral grouping found to precede the formation of triporose grains.
As a sequel to this paper, the writer has in preparation a detailed account of the pollen morphology of many of the genera and species of South African Proteaceae.
My grateful thanks are due to Baron Aberconway, C.B.E., who allowed me to collect developmental stages of Embothrium from the plants growing in the open on his estate at Bodnant, Tal-y-Cafn, North Wales, and to Prof. R. H. Compton, who provided developmental stages and
fresh material of Leucadendron.
34 The Journal of South African Botany.
BIBLIOGRAPHY. Battey, J. W. and Nast, C. J. (1943). The Comparative Morphology of the Winter- aceae. Jour. Arn. Arboret., Vol. 24, p. 341 and Plate IT, Fig. 14.
Brovuexu, P. (1933). The Life-History of Grevillea robusta Cunn. Proc. Linn. Soc., N.-S. Wales, Vol. 58, p. 52, Figs. 48 and 56.
Tasst, F. (1898). Le Proteaceae in specie dello Stenocarpus smuatus Endl. Bull. Lab. ed Orto Bot. Siena, Anno I, Fasc. 2-3. Tay. ILI, Figs. 13 to 17.
Tuomepson, D’Arcy W. (1942). Growth and Form, Ed. II, pp. 630-631.
WopesHouse, R. P. (1932). Tertiary Pollen, I. Pollen of living representatives of the Green River Flora, Bull. Torr. Bot. Club, Vol. 59, p. 336 and Fig. 21.
WoveExowssE, R. P. (1935). Pollen Grains.
PE ae P. (1936). The Evolution of Pollen Grains, Bot. Review, Vol. II, pp- 67-84.
JOURNAL OF SOUTH AFRICAN BOTANY
VOLUME XII.
PLANTAE NOVAE AFRICANAE.
“ Ex Africa semper aliquid novi.”— Pliny.
SERIES XXYV. By Proressor R. 8. ApAmson, Rev. J. GERSTNER and Captain (S.) T. M. Satter, R.N. (Ret.).
Hypertelis trachysperma Adamson sp. nov. (Aizoaceae.)
Herba annua succulenta glabra purpurascens. Caules prostrati, usque ad 15 cm. longi. Folia teretia vel subtrigona, carnosa, saepe glauca, 2—6 em. longa, stipulis acuminatis. Pedunculi alterni, 3—4 cm. longi, foliis subaequilongi, 3—8-floriferi, pedicelis 3—9 mm. longis, -apice incrassatis. T'epala carnosa, 2 mm. longa. Stamina 5, tepalis breviora, antheris rotundatis, filamentis tenuissimis. Stigmata 4 vel 5, saepissime 5, sessilia. Capsula subsphaerica, tepalis aequilonga. Semina reniformia, tuberculata.
At the margins of vleis on muddy soil. Fl. Feb.—Apl.
Hab. Cape Province. Cape Division: Zeekoe Vlei, Walgate 687 (Type, in Herb. Bolus): Kommetje Adamson (s.n.).
A larger plant than H. arenicola Sond., less glaucous and usually purplish, with larger leaves, stouter and longer peduncles, 5 stamens, 4 or 5 carpels, and tuberculate seeds. The stems are firmly pressed to the ground. The leaves on the main stem are subopposite and usually wither at the tip at flowering time, those on the branches are much shorter. The perianth segments are very thick,
MARY GUNN LIBRARY NATIONAL BOTANICAL PRIVATE BAG X 1014 PRETORIA 0001 REPUBLIC OF SOUTH AFRICA
36 The Journal of South African Botany.
Mezleria stenosiphon Adamson sp. nov. (Campanulaceae.)
Perennis decumbens. Folia obovata vel spathulata obtusa vel subacuta apice remote dentata, inferiora conferta breviter petiolata, Superiora sparsa sessilia angustiora. Flores rosei corymbosi. Lobi corollae aequilongi patentes tandem reflexi, tubus cylindricus long- issimus sepalis multoties longior. Stamina erecta, exserta, antheris posterioribus glabris, anterioribus minoribus pilis albis et seta unica longa terminatis. Ovarium obconicum.
nie, IL Bupeicns trachysperma. 1. Shoot, nat. size. 2. Young inflorescence x 5 3. Flower, perianth removed x 5. 4. L.S. flower x 5 5. Ri 5. Del. M. Walgate. a ee:
Stems 10—30 em. long, ridged, glabrous, slightly woody at the base. Leaves 6—12 mm. long or rarely longer, 4—6 mm. wide. Racemes simple or compound, corymbose. Pedicels slender, 4—8 mm. long. Corolla rose-pink, the tube cylindrical, 1—1-5 em. long, split to the base, the lobes equal, spreading, finally reflexed. Anthers dark blue or blue-black, ‘ glabrous, the anterior smaller and tipped with a tuft of white hairs antl a single longer bristle, with minute membranous scales between the hair tufts. Style with a prominent ring below the small stigma.
Humus soils on the higher mountains, FI. Jan.—Apl.
Plantae Novae Africanae. B37
Hab. 8.W. Cape: Upper Wellington Sneeuwkop, 5,500 ft. Adamson 3605 (type in Bolus Herb.); Mts. N. of Fransch Hoek Pass Pillans 6738 (as Isolobus) ; Sondereinde Mts. (A. Bolus) Guthrie 4570; K. of Kogelberg, Stokoe (s.n.).
In collections this has been assigned to [solobus or even to Laurentia ; from either it is at once separated by the corolla and stamens.
The fresh material examined had flowers of a bright rose-pink ; Pillans notes the flowers as “ purple’, Guthrie as “ blue’. In the dried state there is no difference between these specimens.
Prismatocarpus debilis Adamson sp. nov. (Campanulaceae.)
Perennis. Caules tenues, ramosi, decumbentes vel prostrati, hispidi. Folia subglabra, inferiora opposita, ovata ovato-elliptica vel etiam subrotunda, obtusa vel acuta, brevissime petiolata, remote dentata, superiora alterna, angustiora, minora. Flores solitarii vel pauci, distantes, pedunculis longis tenuibusque. Corolla campanulata, sepalis linearibus longior, tubo lobis duplo longiora. Fructus sulcatus, glaber.
A very slender diffuse perennial in loose tangles. Stems 10—20 cm. 4 mm. wide, the upper
long. Lower leaves opposite, 4—6 mm. long, 2-5 alternate, smaller and more distant. Peduncles very slender, with 1 or less often 2—4 distant flowers. Corolla white or pale blue, 5—7 mm. long, the lobes about half as long as the tube. Fruit 1-2—1-8 cm. long.
Sheltered rock crevices on the upper parts of the mountains. FI. Dec.—Feb.
Hab. S.W. Cape: Genadendal Mountains: at 3,000 ft., Bolus 7305 (Type, in Bolus Herb.); at 4,800 ft., Schlechter 9882; at 3,000 ft., Schlechter 10298 ; Stokoe (s.n.). Wellington Mts. ; Observation Peak at 4,000 ft. Hsterhuysen 1712; Upper Sneeuwkop at 4,500 ft., Adamson 3606.
This species was first collected in 1885 and recognised as distinct by Bolus in 1887 when he attached the name to specimens. No description was, however, published. His suggested name is that adopted and the first collected specimens are the type.
Specimens from the Genadendal mountains have leaves about twice as long as wide, whereas those from the Wellington mountains have wider leaves, about as wide as lone.
There are in Herb. Bolus two specimens, Schlechter 9975 from Michell’s Pass, and Bolus 8337 from Ceres, which resemble P. debilis but are of stouter more rigid plants with a more branched inflorescence. They are dlants from lower altitudes, 2,000 ft. and 1,700 ft. respectively. Further collecting and field observation is needed before these plants can be definitely associated with or separated from P. debilis.
38 The Journal of South African Botany.
Croton steenkampiana Gerstner sp. nov. (Euphorbiaceae.)
Frutex 1—3 m. altus. Rami dense lepidoto-stellati, exstipulati. Folia cordata ¢. 5 em —10 cm. longa et 3 cm.—6 cm. lata, integra, dis- colorissima, supra viridia et disparsite stellata, infra albis capillis stipitatis et stellatis tomentosa et basi glandulis 2 stipitatis ornata. Flores 6—15
in racemo monoecio. Pedicelli 3 mm.—5 mm. longi. Sepala triangularia 1-5 mm. x 3 mm. Petala nulla. Ovariwm globosum, stellato-tomen- tosum. Styli 5 mm. longi, tres bis bifurcati. Fructus capsula trilocu- laris, seminibus tribus atris nitidis.
lane:, D. Gian steenkumpiana. 1. Flower. Fic. 3. Melhania suluensis. 1. Flower ing branch x 4. 2. Female flower, nat. 2nd leaf. 2. Stamen. 3. Calyx and epicalyx. size. 3. Stipitate clare hair x 10, 4: Genitalia. 1, 3 and 4, nat. size: 2, x 6. 4. Young inflorescence. 5. Base of leaf Del. F.J. Gerstner.
from below, showing two glands. Del. F.
J. Gerstner.
Hab. Zululand, District Hlabisa, at False Bay Bird Sanctuary. Gerstner 6128 et 6129. (National Herbarium et Kirstenbosch Herbarium.) It is named after Dr. L. Steenkamp, M.P. of V ryheid, in thankful acknowledgement of the interest and assistance which he gave me in my botanical research during the last two years.
Plantae Novae Africanae. 39
The Zulu name is “ iHubeshane ’’, a name which is also used around uBombo for Croton subgratissumus. It is a shrub 1—3 m. high. The branches, leaves and ovary are covered with white stellate and stipitate hairs. The cordate leaves especially are densely tomentose. and white beneath with these stellate hairs. The base of the leaves is furnished with the stipitate glands usually present in this genus and are 3- or 5- nerved. The racemes are monoecious and have 6—15 small flowers, the few female ones being always basal. The male flowers were unfortunately all dried up in consequence of the drought during the last two years. The female ones have no corolla. The five calyx-lobes are triangular and hairy. The 3 stigmas are twice bilobed. The fruit is a trilocular capsule with three black shiny seeds.
Melhania suluensis Gerstner sp. nov. (Sterculiaceae.)
Frutex 50—120 cm. altus, erectus, foliosus ubique tomentosus seu pilosus et multibrachiatus. Folia numerosa, alternata, stipulata, lanceolata, c. 22 mm. longa et 12 mm. lata, dentata, 3—-5-nervata, con- colorata infra superque, tomentosa, nervis inconspicuis immersis, petiolo 5—10 mm. longo. Flores solitarii, pedicelli c. 20 mm. longi. Hpicalyx segmentis 3 lanceolatis. Calicis segmenta 5, c. 7 mm. longa, lanceolata. Corolla quinque foliis c. 10—18 mm. longis et latis, ovatis, citro-luteis ornata. Stamina 5 fertilia et 5 infertilia. Stylus c. 12 mm. longus, stigma 5-brachiata ornatus. Capsula 5-locularis intra glabra, extra pilosa noma glabrescens. Semina 5, atra.
Hab. Zululand. Very frequent around the Dhlangubo Store half-way between Eshowe and Empangeni: found throughout the Umblatuzi Bushveld. Gerstner 2888. (Durban Herbarium and Bolus Herbarium.) Flowered 15/9/36. This muJtibranched and rather woody shrublet is usually about 1 m. high and has very attractive lemon-yellow single flowers. The Zulus call it, as they do in the case of Sida and all malvaceous flowers, ““ uVemvane ’’, the “ Butterfly ’’, on account of its quickly-fading nature. The serrate leaves are about 20 mm. long and 12 mm. broad, tomentose and greyish-green on both sides. The veinlets are sunk in the velvet surface and therefore scarcely visible. The petiole is about 5—10 mm. long. The flowers are usually single, the pedicels about 20 mm. long. The epicalyx has three lanceolate lobes, the calyx 5 similar ones. The 5 corolla leaves are ovate and about 10—18 mm. long and broad. The fruit, a 5-locular capsule, has 5 black seeds and is glabrous inside and hairy outside.
Grubbia gracilis Salter sp. nov. (Grubbiaciae.) G. rosmarinifoliae aftinis, sed ita differt :—Fruticulus minor, gracilior, 30—50 em. altus. Folia anguste obcuneata, 4—6 mm. longa, valde
40 The Journal of South African Botany.
revoluta, basi subsagittata, lobis obtusis petiolo fere in longitudine aequantibus. Bracteae quarta pars bilobatae. Filamenta 0-5—O-7 mm. longa, subsigmoidea, basin versus ampliata, minute puberula.
A slender diffusely branched shrublet, 30—50 cm. high, the stem terete, striate, with tufts of hairs at the nodes. Branches downy, the ultimate branchlets hirsute. Leaves opposite, patent, at length reflexed, narrow-obcuneate to almost linear, but always tapering to the apex, 4—6 mm. long, 0:-6—2 mm. broad, scabrid and pubescent when young but soon glabrescent above, strongly revolute, tomentose beneath, sub- sagittate at the base, the lobes obtuse, nearly as long as the petiole : petiole about 1 mm. long, hirsute. Cymules sessile, axillary, about 2 mm. long, 3- or rarely 2-flowered. Bracts }-bilobed, keeled, mem- branous, chestnut-brown, 1-4—1-6 mm. long, 2—2-4 mm. broad, the
DG ow bt
Fie. 4. Rafnia ericifolia Salter. 1. Vexillum x 3. 2. Carina x 3. 3. Calyx, flattened x 3. 4. Ala x 3. 5. Ovary and style x 4. (Salter 6567.) Del. T. M. Salter.
lobes very obtuse. Flowers closely pressed together. Perianth segments oval, convex, about 1 mm. long, incurved at the apex, the margins thickened, with 2 small swellings at the base : dorsal hairs white, about half as long as the segments. Milaments slightly sigmoid, swollen at the base, minutely puberulous: anthers 0:3 mm. long. Ovary 1- chambered, hirsute, with 2 pendulous ovules. Style filiform, 0-4 mm. long, enveloped in a dense tuft of disc-hairs. Fruits all three connate and inseparable in one row, forming a transversely oval hard dry 3- chambered syncarp, crowned with the three small, densely hirsute discs and the persistent styles.
Hab. Cape Province. Cape Peninsula: in marshes on Muizenberg Mt. Plateau, and near Muizenberg Reservoir. Alt. 1,300—1,500 ft. Salter 3594 (type in Bolus Herbarium), 2948 ; Compton 16430 (nearly in fruit). Flowers September. (Co-types distributed to Kew and British Museum.)
Plantae Novae Africanae. 4]
Of the four hitherto recognised species of Grubbia, G. tomentosa (Thunb.) Harms (G@. stricta A.DC.) alone shows distinctive floral characters, and in the remainder the differences, with the exception of slight variation in the bracts of the cymule, are almost entirely in the vegetative characters. G. gracilis is a much smaller and more slender species than its nearest affinities, G. rosmarimifolia Berg. and G. pinifolia Sond., from which it can be at once recognised by the narrow-obcuneate leaves with a sub- sagittate base. The syncarp in these species is remarkable, for the three 1-chambered ovaries of each cymule fuse together in a row in fruiting, into what appears to be, in the middle lateral section, an inseparable 3-chambered capsule.
Rafnia ericifolia Salter sp. nov. (Leguminosae-Papilionatae) §Ku- Rafnia.
Fruticulus parvus glaber, ad 20 cm. altus, ramis numerosis adscendenti- bus vel diffusis, habitu tholiformi. Folia sessilia linearia, plerumque 5—8 mm. longa, 0-6 mm. lata. Flores pauci laterales uniflori flavi, plerumque ramulorum apices versus positi. Pedunculi semi-reflexi. Calycis lobi, lanceolati, tubo obconico fere in longitudine aequantes, posterior angustior. Petala in longitudine fere aequantia : vexilli lamina ovata acuta carinata : carina rostrata. Ovariwm 4-ovulatum. Legumen planum, ad 2-3 cm. longum, 4-5 mm. latum, reflexum, subglaucum.
A small, rather compact glabrous shrublet, dome-shaped in habit, up to about 20 cm. high, with numerous ascending or spreading branches. Leaves ascending to erect, imbricate, sessile, linear, obtuse, mostly 5—8 mm. long, about 0-6 mm. broad, the median vein prominent above : margins sometimes slightly revolute. Flowers few, lateral, mostly near the apex of the branchlets. Peduncles 1-flowered, semi-reflexed, swollen at the apex. Sract filiform, scarcely 1 mm. long: bracteoles minute, 0-4 mm. long, near the base of the peduncle. Calyx about 7 mm. long, the tube subobconic : lobes as long as the tube, more or less lanceolate, the posterior narrower than the remainder. Petals yellow, subequal in length: vexillum about 1 cm. long, the lamina ovate, acute, keeled, 5 times as long as the curved claw: lamina of the alae obliquely lanceolate : carina rostrate, subacute, somewhat incurved, sparsely pilose. Anthers alternately globose and lanceolate, the latter sterile. Ovary about 5 mm. long, 1 mm. broad, 4 ovuled: style 6-5 mm. long, inflexed, the stigma minute. Pod about 2-3 cm. long, 4-5 mm. broad, flat, reflexed, somewhat glaucous.
Hab. Cape Province. Cape Division: near Hercules Pillar, Salter 6567 (type in Bolus Herbarium) and Salter 4346. Flowering Jan.—Mch.
42 The Journal of South African Botany.
Also observed (not in flower) south of Paarl Berg. (Co-types distributed to Kew and the British Museum.)
This species has no near affinity: it can be at once distinguished by its dwarf habit and small ericoid leaves. Apparently it never flowers’ very profusely.
A DRYING CABINET FOR THE HERBARIUM. By G. G. Smita.
Drying plants for herbarium purposes is a slow and laborious process at the best of times, especially with succulents. To speed up my work in connection with the drying of the large number of Haworthias and other succulents I am handling, I designed and put into use a small artificially heated drying cabinet, the results of which are proving most
satisfactory, not only in the time and labour saved, but also in the im- proved appearance of the dried material.
44 The Journal of South African Botany.
An important feature of this method of drying is that the plants do not become brittle, nor have they the withered appearance usually found in plants dried in the ordinary way, without artificial heat.
Owing to the fact that the air inside this cabinet is not mechanically circulated, a low temperature element is all that is necessary, and with the permanently open air-holes at the bottom of the cabinet and the controlled ones near the top, one is able to control the heat inside the cabinet, and there is no danger of overheating.
The best drying temperature is about 120° to 150° F., at about the middle of the cabinet, and while I have not fitted a thermometer to this cabinet, it would be quite a simple matter to do so.
The cabinet consists of a frame of 14 inches by 14 inches of well seasoned oregon pine or other suitable wood. This frame is covered with any suitable board (such as a compressed fibre board used in the ceiling of a house) on the inside and outside so as to form an air-space between the linings for insulation purposes. The width inside the inner linings which is the width inside the cabinet must be + inch more than the length of the drying presses that will be used in the cabinet, and the depth (from back to front) inside about 1 inch wider.
The bottom and top are closed in with timber about 4 inch thick and the height inside the cabinet should be about 40 inches. Along each side are fixed five evenly-spaced rails of 14 inches by ? inch of suitable wood for carrying the drying presses.
The door frame is made of 14 inches by # inch oregon pine with an intermediate horizontal rail and is also covered on both sides to form an air space.
The heating element can be easily made up by any electrical engineer- ing firm, and should be of a type where the resistance wires rest on top of the porcelain so as not to throw any heat in a downward direction. It is advisable to have the resistance wires longer than necessary to give a low temperature, which temperature can then be raised as desired by the shortening of the resistance wires. The element must be securely fixed to the bottom of the cabinet with a space of at least 1 inch between the inside bottom of the cabinet and the under side of the element porcelain. The bottom board has about eight 4-inch holes in a row immediately under the element so that air drawn in here must travel past the element.
The element is covered with a sheet metal hood to prevent the direct heat from the element impinging on the underside of the lower drying press. This hood is best made in the form of an inverted L, with lugs on the lower end for screwing on to the bottom board of the cabinet. The space between the top of the hood and the element must be at least
A Drying Cabinet for the Herbarium. 45
13 inches, and the top or horizontal portion of the hood about 6 inches wide, with the open side towards the back of the cabinet.
Near the top of one of the sides of the cabinet two ?-inch holes at about 14 inches centres are drilled through the inner and outer lining and a sleeve fitted to each hole so as to seal off the air space between the linings. The sleeves can be made of thick paper, glued to form a cylinder long enough to pass through the two linings, and may be glued in position with the outer ends flush with the outside of the cabinet. These holes are covered by a board about 5 inches long by 12? inches wide. and screwed near one end to hinge up and down and thus act as a damper for controlling the flow of air through the cabinet.
Jn operation the presses are placed in the cabinet, the lowest one as far back as possible, the next one as far forward as possible so as to be up against the door when closed. This arrangement of the presses enables the hot air to circulate along the under side of each press. It is only natural that the press nearest the element will receive more heat than the others, and it is therefore advisable if more than one press is being used, to change their respective positions in the racks, and even to turn them over occasionally. :
For higher temperatures in the cabinet, the damper must be almost closed and opened up for lower temperatures. The cabinet will also be found very useful for drying the papers. For this purpose they may be loosely arranged on one half of a drying press placed towards the top of the cabinet.
NOTES ON THE SOUTH AFRICAN SAPOTACEAE By Rev. J. Gerstner, O.S.B., Ph.D.
The following is a revised classification of the South African Sapo- taceae, based on that of Engler.
Tribe 1. Palaquieae. Petals without appendages, not two or three times as many as the calyx lobes. Sub-Tribe 1. Sideroxylineae. Stamens 5 alternating with 5 staminodes. Genus 1. Sideroxylon L. S. inerme L. Sub-Tribe 2. Chrysophyllineae. Stamens 5. Genus 2. Chrysophyllum L. C. natalense Sond. C. viridifolium Wood & Franks. C. magalismontanum Sond. Tribe 2. Mimusopeae. Petals with appendages. Segments of corolla tube twice or thrice as many as calyx lobes. Sub-Tribe 3. Multistaminales. Stamens twice or thrice as many as calyx lobes. No staminodes. Genus 3. Labourdonnaisia Boj. L. discolor Sond. Sub-Tribe 4. Staminodales. Stamens as many as calyx lobes and alternating with the same number of staminodes. Genus 4. Mimusops L. § Hexalobatae. Six calyx lobes, six stamens and six staminodes. M. concolor E. Mey. M. mochisia Baker. § Octolobatae M. caffra BH. Mey. M. marginata N. EB. Br. M. obovata Sond. M. Zeyheri Sond. M. Henriquesiana Sim. M. Schinzii Engl. M. Kirkii Bkr. f.
1. Sideroxylon imerme L., white milkwood, wit Melkhout: the Standard Zulu Name should be uMakwelefinqane ; other more local and ambiguous names are umHlahla, umBobe, and amaSetole. The Xosa name is umQuashu. It flowers in April and has ripe fruits in July. The black berries contain a very viscous juice and are therefore usually not eaten. The pulp of this berry is purplish green. The seeds, usually 1 or 2, are black. It is a frequent evergreen tree of the open bush of the lowveld and yields a good timber. The native doctors use a cupful of the rootbark as an enema to produce a drastic perspiration.
* In the accompanying illustrations the drawings of flowering branches are half natural size (linear).
48 The Journal of South African Botany.
2. Chrysophyllum natalense Sond., the Natal sweet plum, is found as a small tree or as a kind of undershrub in the mistbelt-forests of the East Coast. The red drupes are very delicious. The excellent timber is used for poles. The Zulu name is umTongwane. This name is used also for another tree Oncoba spinosa Forsk. of the dry bushveld. The Zulus rightly distinguish, therefore, between umTongwane wehlathi, i.e. of the forest, and umTongwane wehlanze, i.e. of the open bushveld.
Fie. 1. Sideroxylon inerme. 1. Branchlet with leaf and fruit. 2. Berry showing the position of the seed. 3. Seed. 4. Ovary in vertical section. 5. Part of the corolla tube expanded. 6. The calyx expanded. 7. Ovary with style. 8. Flower diagram.
Fic. 2. Chrysophyllum natalense. 1. Branchlet with leaves and flower. 2. Hair enlarged. 3. Vertical section through the ovary. 4. Corolla tube expanded. 5. Flower. 6. Flower diagram. 7. Fruit.
3. Chrysophyllum viridifolium Wood & Franks.
This is quite frequent in the closed mist belt forests of the first terrace. In the Eshowe forest in particular it is the king of the trees, having usually a straight trunk of 100 ft. or more and the shape of an old Pinus pinaster. The berries have the shape of a little apple with a yellow skin (exocarp), a yellowish-whitish pulp (mesocarp) and five brown seeds. The pulp
Notes on the South African Sapotaceae. 49
is saturated with a sticky white latex, which makes the fruits not very palatable although they are sweet and edible. They can be found ripe and fallen off from the tree in February and March. This is also the best time to recognize this tree in the forest. The royal timber is very much appreciated. Its Zulu name is umG(w)inya or amaSetole amakulu ehlathi.
4. Chrysophyllum magalismontanum Sond. is widely distributed in the Transvaal and is called Stamvrugte. In Natal it occurs only on the Transvaal border along and between the middle Mkuzi and Pongola Rivers. The Zulus call it there amaNumbele, the Shangaan umBovan- hlathi, and all Sothos Motlhatswa. It occurs often in shrub form. The fruits are edible. Synonym: Chrysophyllum Wilmsw Engl.
5. Labourdonnaisia discolor Sond. (= Mimusops discolor Hartog). Genus prius restauratum et completum :
Arbor media (diam. cire. 60 cm.), altitudine 10—20 m., cum latice albo et cortice cinereo. In trunco vetere magis in longitudinem quam in latitudinem fissa, in brachiis parvis cortice levi et cicatricoso. Forma eius similis Piri sed amplius ramosa. Folia alterna, obovata, oblonga, petiolo subterete (I—1-2 cm.). Laminarum longitudo circ. 6-5 em., latitudo cire. 2-5 cm. in loco non nimis sicco. In loco sicco lamina minora. Lamina coriacea, supra viridis, subter tomentosa nitens sicut argentum, basi cuneata, apice recurvato, apiculato noma emarginato, costa subtus prominente tomentosa, nervis lateralibus immersis. Flores infra folia ramulorum, pedicello 5—10 mm., calice duplici trilobato (4 x 4+ cm.). Lobi externi calicis lati, ovati, apice acuminato, subter rubiginosi et verrucosi, supra lutei, subglabri. Lobi interni calicis angustiores lanceo- lati, subter et supra flavi et pilosi. Perianthii segmenta in flore perfecto interiora 6 lanceolata-lineata, simplicia, exteriora 6, latiora, bilobata lobis acuminatis, Forma perianthii ,variat considerabiliter in flore imperfecto et tempore adverso. Itaque flores alii habent tantum 6 lobos simplices noma apiculatos, noma tridentatos, noma trilobatos, faciliter marcescentes et complicantes. Lobi omnes flavi, virescentes et pilosi. Stamina 12, 6 opposita et 6 alternantia lobis interioribus. Antherae caducissimae, luteae, cordatae. Filamenta albida. Stylus virescens ad apicem, paululum exsertus, stigmate sexlobato. Ovarium pilosum, sexloculare. Fructus bacca sphaerica cum pistillo persistente, exocarp) rufo, sarcocarpo succoso, edulis, sapore jucundissimo.
Labourdonnaisia discolor is a medium-sized tree, growing 10—20 m. high and having a diameter of about 60 cm. It has a white latex. The grey bark is longitudinally fissured. On young branches the bark is more smooth and full of leaf-scars. The shape of this tree is that of a Pear tree, but more branched. The leaves are alternate, obovate-oblong. The petiole
50 The Journal of South African Botany.
is subterete and 1—1-2 cm. long. The blade is about 6-3 cm. long and about 2-5 em. broad, green above, silvery-pubescent below. The midrib is prominent below, the lateral veins immersed, the base is cuneate, the apex recurved and acuminate or emarginate. The little fascicles of flowerlets are in the axils of the leaves. The pedicel is 5—10 mm. long. The double calyx is trilobate. The single lobe is 4 x } cm. in size. The calyx-lobes are fulvo-verrucose below, yellowish and smoother above. In a perfect flower the corolla tube has six lanceolate-lineate inner lobes and six larger deeply bilobate outer lobes. They are all greenish-yellow and hairy and vary considerably in form and size if the flowers are not perfectly developed, some being then simple, some tridentate and some trilobate. They wither and wrinkle very easily. There are 12 stamens, 6 opposite and 6 alternating with the inner corolla-lobes. The cordate- shaped anthers very easily fall off and are of brownish-yellow colour. The filaments are lighter coloured. The style is a little exerted. The stigma is minutely 6-lobed and the ovary is 6-locular. The berry is red and globose with the persistent style at the top. The pulp is juicy and has a good flavour. It would give a very good jam. The timber is excellent and termite-proof.
As this genus has not all the features of the genus Mimusops, differing in the double number of stamens, the lack of staminodes, the bilobate outside lobes of the corolla-tube, the variability of the perianth-lobes generally in imperfect flowers and the fixed and very reduced corolla- tube, the restoration of its first genus-name “ Labourdonnaisia ”’ seems to me necessary. It is usually found on the Ecca-series-sandstone in the transition-region of closed to open bush, most frequently in the Hlabisa District, but all through the coastal districts of Natal right north to the Mangusi Forest of the Ingwavuma District. It is called by the Zulus “umNweba,” ie., “the white shining thing,’ because the bark of the umNweba twigs shows, if peeled off, a kind of white wool, probably
Fie. 3. Chrysophyllum viridifolium. 1. Branchlet with flowerlets. 2. Flower bud. 3. Horizontal section through the fruit, diagrammatic. 4. Fruit. 5. Perianth- tube expanded. 6. Vertical section through fruit. 7. Flower-diagram. 8. Verti- cal section through flower.
Fic. 4. Chrysophyllum magalismontanum. 1. Part of an older branch with flowers. 2. Younger branch with leaves. 3. Ovary with part of calyx. 4. Corolla-tube expanded. 5. Flower-diagram. 6. Stamen.
Fic. 5. Labourdonnaisia discolor. 1. Flowering branchlet. 2 & 3. Stamens. 4. Fruit. 5. Part of the corolla-tube expanded and seen from outside. 6. Corolla- tube expanded and seen from inside. 7. Flower-diagram. 8—11. Four different variations of the corolla-lobes. 12. Vertical section through a flower-bud, diagrammatic.
Fie. 6. Mimusops concolor. 1. Flowering branch. 2. Stamen with two staminodes. 3. Flower enlarged. 4. Corolla tube seen from top. 5. Different varieties of staminodes. 6. Flower without corolla-tube. 7. Fruit. 9. Flower diagram.
Notes on the South African Sapotaceae.
ep
Um ss rn CTO I
5.Staminodes
Fics. 3—6.
52 The Journal of South African Botany.
the coagulated guttapercha. In using this native name we must remember that there are three kinds of umNweba: umNweba olwandhle, the umNweba of the sea, Mimusops caffra; umNweba wentaba, the um- Nweba of the hills, Labourdonnaisia discolor ; and umNweba wehlathi, the umNweba of the Ngome forest, Gymnosporia acuminata Szysz.
It flowers usually in September and October. The fruits are ripe in December or later, and are the most delicious wild fruits which I have ever tasted in Africa. In 1938 I found the flowers of this species not fully developed, but in October 1943 I received from the same place perfect flowers, which I have described above.
6. Mimusops concolor Harv.; the Zulu Standard name of this tree is ““amaSetole amhlope.” It is a tree or often a shrub of the dry bushveld preferring Sandstone formations. The Tonga name is umNqambo. This termite-proof tree yields very good poles and excellent timber. The shape is usually that of a Pear tree.
7. Mimusops mochisia Baker, a tree of the nothern bushveld, especially round Ndumo and the Nyamini-pan. It grows about 30 ft. high, has a whitish rugose bark and the shape of an apple or even a willow. The Tonga name is also umNqambo. I found in summertime no fruits or flowers on it. Its very crooked branching is typical. The leaves are deciduous and dull like Mimusops concolor, but larger. According to Sim the yellow fruit is of the size of a bullace contaming four seeds. May be only a tropical form of Mimusops concolor.
8. Mimusops caffra EX. Mey. Its Zulu name is uMakayi, i.e. the duiker’s hair brush, on account of the similarity of the opening flower to the tuft of hair on the head of the duiker. It is an excellent timber. Another more general and ambiguous name is the Zulu name “~“umTunzi,”’ the Shade tree. Chief John Dunn administered justice to the natives under this tree at his residency at the Umlalazi lagoon. Hence the township there is now called emTunzini. The fruit is edible and well-flavoured. The exocarp is orange coloured, the mesocarp contains a starchy sweet _ pulp. This tree is often dominant in the sand dune bush of the coast, where it is usually the staple food of the monkeys, which spread this tree not digesting the hard black seeds. In the forest at the Inyoni River
Fic. 7. Mimusops caffra. 1. Twig with leaves and flower. 2. Stamen. 3. Corolla- tube expanded. 4. Vertical section through flower. 5. Fruit. 6. Seed. 7. Flower diagram.
Fic. 8. Mimusops marginata. 1. Branch top with leaves and flowers collected at 1/9/43 at Nongoma. 2. Flower diagram.
Fic. 9. Mimusops marginata. 1. Fruit. 2. Stamen seen from outside. 3. Stame2r seen from inside with horizontal section. 4. Whole flower in vertical section.
Fie. 10. Mimusops obovata. 1. Twig with flowers and leaves. 2 & 3. Stamens seen from inside and outside. 4. Corolla tube expanded. 5. Flower. 6. Flower diagram. 7. Fruit in vertical section.
Notes on the South African Sapotaceae. 53
54 Fhe Journal of South African Botany.
mouth 75% of the trees are Mimusops caffra. At Sordwana Bay it gets even the direct spray of the surf. The afforestation of the many thousands of acres of shifting sands along the Zululand coast seems to me not impossible if the soil is prepared in a wet season by planting first sugar cane and then between the cane rows planting out the trees when about 3 ft. high in cheap containers. A mixed forest of Mimusops caffra E. Mey., Brachylaena discolor DC., Olea woodiana Knobl. and Casuarina equisetifolia, would give an evergreen bush and yield excellent timber. Monkeys, bucks and birds would soon create an association of undershrubs as well, if they get a chance.
9. Mimusops marginata N.E.Br., a forest tree and royal timber, which grows in mist-belt forests and in closed river bush. Its Zulu name is umPumpulo and the fruits, which are not eaten by man, are called amaPumpulo.
10. Mimusops obovata Sond., the red Milkwood, Rooi Melkhout, is a tree of the closed mistbelt forests, but found as well in the bushveld along the more closed bushes especially in kloofs. Its tasty red fruits are edible and very much appreciated if ripe. The royal timber is termite- proof. The Standard Zulu name is “ amaSetole ambomvu,” the Swazi name is umPushane, another more general Zulu name is umTunzi, Sim gives as Natal name isiPandane. It flowers from August to October and is fertilized by humble bees. I found a half a dozen of these humble bees on one shrub. It is the most frequent Mimusops of the country, growing along the Riverbushes near the sea in all forests and on good soil as well as on pure rocks, and it is not surprising that it varies very much in appearance of leaves and way of flowering. In shady places of the forests we find tender leaves and few flowers, on rocks in sunshine more coriaceous leaves and plenty of flowers. No wonder that there are quite a number of synonyms: Mimusops Rudatisi Engl. and Krause ; Mimusops oleifolia, a poor bushveld form; Mimusops Woodi Engl., probably a mistbelt form; probably also Mimusops blantyreana, a more tropical form.
11. Mimusops Zeyheri Sond., the Moepel, which all Sothos call moPudu and the Vendas Mubululu. It is a royal timber of the western Transvaal and Betchuanaland bushveld. The leaves are on both sides dull greyish-green. It flowers in January. The fruits are edible and very tasty and vary very much in size and shape from sub-globose to egg- shaped. They are ripe in November. .
12. Mimusops Henriquesiana Sim. I recorded this at False Bay without flowers or fruits. It has more copious Guttapercha-latex than other species.
13. Mimusops Schinzw Engl. may occur as well in northern Zululand,
Notes on the South African Sapotaceae. 55
but has not yet been found. Cf. Henkel, woody plants of Natal and
Zululand.
14. Mimusops Kirku Bkr. f. was recorded from Magoeba’s Kloof, North Transvaal. Cf. Gerstner 5881, National Herbarium.
15-18. There are four other species of Mimusops of Southern Rhodesia, which possibly may occur in the Limpopo-valley of Northern Transvaal : M. decorifolia S. Moore. M. Monroi 8. Moore. M. spiculosa Hutch. & Corb. M umbraculigera Hutch. & Corb.
RESERVE CARBOHYDRATES IN SOUTH AFRICAN GRASSES.
By Hans WEINMANN AND LEONORA REINHOLD. (Department of Botany, University of the Witwatersrand, Johannesburg.) (With Prats IIT.)
Figures in parentheses, see References at end of this article.
The significance of carbohydrate reserves in certain South African grasses has been discussed in various previous papers (12-15). The results of sugar and starch determinations led to the conclusion that these materials are the more important types of reserve carbohydrates in the species investigated. In recent years considerable attention has been focussed on the occurrence of fructosans in grasses and other monocotyledons (1, 4). Sullivan and Sprague (11) point out that fructo- sans may often have been reported as starch and dextrin, and consider that fructosan, rather than starch, is important as a reserve carbohydrate in grasses. As a contribution to the ultimate settlement of these con- troversies, the writers have now carried out detailed carbohydrate analyses on the roots, rhizomes and shoots of a number of important indigenous South African grasses, involving the determination of reducing and non-reducing sugars, fructosan, dextrin and starch in these materials. Carbohydrate analyses of various exotic grasses were included.
MATERIAL.
Roots and Rhizomes.—The species from which root and rhizome samples were taken are listed in Table I, together with data regarding localities and times of sampling. Wherever possible, the samples were taken from protected plots, and during autumn or winter, so as to ensure a high content of reserves. A number of plants of each species were dug up to a depth of at least four to six inches. The bulk of the adhering soil was removed by washing or shaking, and as soon as possible the material was taken to the laboratory, where it was killed by autoclaving at 5 lb. pressure for five minutes. The samples were then dried at a temperature of 50° to 60° C., cut up, ground and stored in air-tight bottles for chemical analysis.
58 The Journal of South African Botany.
TaBLeE I.
Roots and Rhizome Materials Investigated.
Species.
Sampling Date.
Locality.
Brachiaria serrata Elyonurus argenteus Eragrostis chalcantha Harpechloa falx Microchloa caffra Monocymbium ceresviforme Trachypogon plumosus Tristachya hispida
8th August, 1942 19th June, 1942
7th August, 1942
8th August, 1942 19th June, 1942 21st June, 1942 16th June, 1942
‘18th June, 1942
Natural veld; plots pro- tected from grazing for four seasons ; old herb- age removed by cutting every year in mid-
winter.*
Digitaria tricholaenoides
10th, August, 1942
Moderately grazed plots.*
Cynodon dactylon, Friel’s Selection
do., Wild Kweek grass
13th July, 1943
25th Sept., 1944
Turf plot, protected for two years.*
Intensively grazed Cyno- don pasture.*
Hyparrhenia hirta Themeda triandra
6th July, 1945 6th July, 1945
Protected plot, burnt every alternate year in winter since 1931.7
Agrostis tenuis Arrhenatherum elatius Lolwum perenne Phalaris arundinacea
21st March, 1945 19th March, 1945 19th March, 1945 21st March, 1945
6 to 10-year-old nursery plots.*
Chloris gayana Paspalum dilatatum Pennisetum clandestinum
10th July, 1945 10th July, 1945 10th July, 1945
Plot protected for seasons. *
two
* Botanical Research Station, Frankenwald, near Johannesburg.
+ Crescent Creek Experimental Plots, Johannesburg.
Shoots.—During the season 1943-44 herbage samples of four species of grasses were collected in one of the experimental plots of the Witwaters- rand University at Crescent Creek, Johannesburg. This plot had been burned every alternate year in June since 1931, but remained protected from any other interference. Cook (3), examining the plot in 1937, found that its vegetation had not changed appreciably since the first burn.
For the present investigation Brachiaria serrata, Hyparrhenia hirta, Trachypogon plumosus and Themeda triandra were chosen as test materials.
Reserve Carbohydrates in South African Grasses. 59
Herbage samples of these species were collected on 24th November, 1943, 15th February, 1944, 18th April, 1944, and 28th June, 1944. The last burn to which the plot was subjected before these samples were taken took place on 30th June, 1943. Information concerning the growth stage of the individual species on the sampling dates is given in Table IT.
TaBLe IT.
Phenological Observations.
Sampling 24th Nov., 15th Feb., 18th April, 28th June, Date. 1943. 1944. 1944. 1944. Brachiaria Leaves and | Endof flowering | Allstems and | All aerial serrata stems ; in- period ; in- most leaves parts dead. florescences florescences dead formed lost Hyparrhenia Leaves only | Flowering stage | Seeds being Do. hirta formed formed ; yel- lowing Themeda Inflorescences | End of flowering | Leaves partly Do. triandra formed period ; in- brown
florescences still present
Trachypogon Leaves ; stems | End of flower- | Leaves partly Do.
plumosus being formed ng period; in- brown florescences lost
The samples were always taken in the morning between nine and ten o’clock, and were killed immediately by autoclaving. The matérial was then dried and ground in the same way as the root and rhizome materials.
ANALYTICAL METHODS.
Sugars.—Sugars were extracted from 2-5-gram samples of the air-dry material by heating with 100 ml. of 95 per cent. alcohol for two hours. Some calcium carbonate was added to prevent acid hydrolysis during extraction. The alcohol was removed from the filtrate by distillation under reduced pressure, the remaining syrup taken up with water, cleared with neutral lead acetate, made to 100 ml. volume, and de-leaded with potassium oxalate. A 25-ml. aliquot was hydrolyzed by heating for half
60 ~ The Journal of South African Botany.
an hour on a boiling water bath with 1-25 ml. of 25 per cent. hydrochloric acid. The reducing power of the sugar extracts was determined by a previously described semi-micro method (16). The latter was also used to estimate the reducing power of the other carbohydrate extracts prepared in this investigation.
Fructosans and Dextrins—The residue from the sugar extraction was dried at 50 to 60° C., and transferred to a 500-ml. Erlenmeyer flask. 100 ml. of distilled water and a few drops of toluene were then added, and the mixture was agitated for two hours on a shaking apparatus. The liquid was filtered under suction through a dry filter on a Buchner funnel, without washing: 50 ml. of the aqueous extract were pipetted into a 250-ml. Erlenmeyer flask, and to these were added 20 ml. thymol- saturated acetic acid-sodium acetate buffer solution, pH 4-45, and 5 ml. of dialyzed takadiastase solution (corresponding to 25 mg. of original takadiastase, Parke, Davis & Co., Detroit, U.S.A.). The flasks were then tightly stoppered and incubated for 44 hours at 37 to 38° C. for the digestion of dextrins. At the end of the digestion period 1-1 ml. of 25 per cent. hydrochloric acid were run into the digests, and the flasks heated for half an hour on a covered bath of boiling water to hydrolyze fructosans (the liquid in the flasks reaching a temperature of approxi- mately 75° C.). The hydrolysates were cooled down, three-quarters of the acid was neutralised with a pre-determined quantity of strong sodium hydroxide solution, the liquid was transferred to a 100-ml. flask, cleared with neutral lead acetate and made to volume. The filtered liquid was then neutralized with solid sodium carbonate to a pH of 5-0 to 5-5, the pH being checked by a Beckman pH meter, after which the solution was de-leaded with solid potassium oxalate.
For the determination of fructosans a modification of the colorimetric method developed by Roe (10) and Hubbard and Loomis (6) for blood and urine analyses was used. The method is based on the fact that fructose when heated with a solution of resorcinol in the presence of concentrated hydrochloric acid gives a red colour, the intensity of which, within certain limits, is proportional to the fructose concentration. One ml. of a 1 per cent. solution of resorcinol in 95 per cent. alcohol was added to 1 ml. of the cleared and de-leaded filtrate in a graduated test-tube, followed by 3 ml. of 30 per cent. hydrochloric acid. After being shaken to ensure adequate mixing, the tubes were heated for exactly eight minutes in a water-bath maintained at 80° C., and then rapidly cooled in running water. The resulting coloured solutions were made up to the volume of 5 ml. with 95 per cent. alcohol, after which they were compared in a Dubosque colorimeter with a standard fructose solution which had been treated simultaneously with the filtrates by
Reserve Carbohydrates in South African Grasses. 61
the same procedure. The method was tested on pure fructose solutions over a range of 0-02 to 0-4 mg. per ml., and found to give results accurate to well within 5 per cent. of the expected value. Concentrations below 0-02 mg. per ml. gave a colour that was still detectable, but not measur- able in the Dubosque colorimeter. Superior sensitivity would pro- bably be achieved with a photo-electric colorimeter.
Interference due to the presence of glucose was investigated in a series of experiments. It was found that concentrations lower than 1 mg. glucose per ml. did not produce a measurable colour. The fructose equivalent of higher concentrations of glucose (concentrations up to 50 mg. per ml. were tested) tended to vary slightly. The average figure for the apparent fructose value of 1 mg. glucose was 0-006 mg. Wherever discrepancies between total sugar estimations and fructose determina- tions indicate that more than 1 mg. glucose is present per ml., therefore, corrections for glucose should be made. These were unnecessary in the present investigation as the glucose concentration never approached 1 mg. per ml.
In order to estimate the possible interference by chromogens other than reducing sugars, McRary and Slattery (8) in a recent paper recom- mend fermentation of the hydrolyzed extracts previous to treatment with resorcinol. In the present investigation several extracts were subjected to rapid fermentation with baker’s yeast (17). The fact that colourless solutions were subsequently obtained on heating with resor- cinol indicated, firstly, that all the chromogens present in the solutions were fermentable, and, secondly, that no complex polysaccharides capable of being hydrolyzed to chromogens under the strongly acid conditions obtaining for colour development were present.
It should be noted that some extracts were observed to yield deep- red solutions when heated with concentrated hydrochloric acid even in the absence of resorcinol. The presence of alcohol, however, inhibited this colour formation, which thus did not interfere with the estimations.
From the results of the determinations of fructose and of the total reducing power, the fructosan and dextrin content of the samples was calculated, taking into account the weight increase of these substances upon hydrolysis as well as the lower reducing power of fructose as compared with that of glucose (16).
Starch.—The residue from the water extraction was thoroughly washed, and, following the usual gelatinization treatment, was digested with 3 ml. saliva for 20 hours at 37 to 38° C.; 50 ml. of the cleared and de-leaded digest were hydrolyzed with 4 ml. of 25 per cent. hydrochloric acid in an autoclave at 15 1b. pressure for one hour. Starch was calculated by multiplying the glucose value by the factor 0-9.
62 _ The Journal of South African Botany.
Ash and Dry Matter.—Dry matter determinations were carried out on all samples, and in roots and rhizomes the ash content was deter- mined in addition. The analytical results were reported as percentages of the dry matter in the case of the shoots, and as percentages of the ash-free (combustible) dry matter in the case of the roots and rhizomes. Errors due to variations in the amounts of sand and soil adhering to the underground parts were thus eliminated.
Discussion OF RESULTS. I. RESERVE CARBOHYDRATES IN UNDERGROUND Parts.
The results of the carbohydrate determinations in roots and rhizomes are given in Table III. As will be seen from these figures, reducing sugars. occurred only in small amounts in most materials. Exceptions were the roots of Hlyonurus argenteus (1-04 per cent.) and Chloris gayana (2-24 per cent.), and the roots and rhizomes of Pennisetum clandestinum (4-95 and 2-81 per cent. respectively). Non-reducing sugars, on the other hand, were found in appreciable amounts in the majority of the materials investigated. In 12 out of the 14 root samples of indigenous South African species, non-reducing sugars occurred in larger amounts than did the other types of carbohydrates. Amongst the exotic species this was the case only in Paspalum dilatatum and Pennisetum clandestinum, both species being indigenous to tropical countries. While considerable quantities of non-reducing sugars were also recorded in most of the rhizomes, it was only in.those of Digitaria tricholaenoides that these compounds were present in excess over the other reserve carbohydrates.
No significant amounts of fructosans were found in the roots and rhizomes of any of the indigenous South African grasses, the highest concentration being reached in Chloris gayana with 0-70 per cent. Fructo- sans were, however, abundant in the roots of all the exotic grasses in- vestigated. In the roots of four of the exotic species studied fructosans constituted the principal carbohydrate. These species were Agrostis tenuis, Arrhenatherum elatius, Lolium perenne and Phalaris arundinacea, all indigenous to cool temperate climates. The occurrence of fructosans in these, or closely allied, species has been reported by previous workers (1, 4, 11). Particularly large amounts of fructosans were found in the rhizomes of Agrostis tenuis and Phalaris arundinacea (viz., 17 and 31 per cent. respectively).
Dextrins were generally found to be either absent or occurring only in very small quantities. It may be not without significance that the presence of somewhat larger amounts of dextrins was usually associated with a high starch content (roots of Cynodon dactylon, wild Kweek grass,
Reserve Carbohydrates in South African Grasses. 63
TasueE III.
Reserve Carbohydrates in the Underground Parts of Grasses.
Expressed as percentages of the combustible dry matter.*
Species. Red. Non-red.} Fructo- | Dex- | Starch. | Total. Sugars. | Sugars. sans. trins.
Indigenous South African Grasses.
Roots.
Brachiaria serrata ae 0-15 6-30 0-00 0-00 0-13 6-58 Chloris gayana .. a 2-24 4-69 0-70 0-09 0:03 7-75 Cynodon dactylon : é
Friel’s Selection as 0-06 1-65 0-00 0-06 1:99 3:76
Wild Kweek Grass f .. 0-28 6:05 0-34 0-42 5:86 | 12-95 Digitaria trichol. le 0-48 6-20 0-26 0-18 0-18 7-30 Elyonurus argenteus D0 1-04 6-90 0-34 0-34 0-28 8-90 Eragrostis chalcantha .. 0-08 2-68 0-00 0-60 0:13 | 2-89 Harpechloa falx .. Sus 0-19 2-50 0-20 0-28 i o1N65 4-32 Hyparrhenia hirta Se 0-85 10-50 0-57 0-00 O29) 1) 1227 Microchloa caffra sts 0-03 1-90 0-00 0-00 0-13 2-06 Monocymbium ceres. .. 0-21 4-51 0-00 0-00 0:14 | 4-86 Themeda triandra ae 0-19 8-75 0-41 0-00 0-04 | 9-39 Trachypogon plumos. .. 0-08 3-76 0-00 0-00 0-14 3°98 Tristachya hispida aS 0-01 1-26 0-32 0-44 7-73 9-76
Rhizomes.
Cynodon dactylon :
Friel’s Selection a 0-12 2-50 0-55 0-87 16-57 Wild Kweek Grass; .. 0:23 3-90 0-20 0:72 15-33 Digitaria trichol. Ae 0-26 0-95 0-00 0:28 1-90 Exotic Grasses. Roots. Agrostis tenuis .. Ae 0-14 0-43 2-61 0-00 0:48] 3-66 Arrhenatherum elatius .. 0-23 0-85 1:92 0-36 0-67 4-03 Lolium perenne .. a 0-03 0-33 3-03 0-19 0-25 3:83 Paspalum dilatatum NE 0:46 | 11-38 1-24 0-05 Mos |) i627} Pennisetum clandest. ne 4-95 7-41 1-13 0-23 0:19} 13-91 Phalaris arundinacea .. 0-22 1-61 3:44 0-00 0-24 5-51 Rhizomes. Agrostis tenuis. .. ae 0-61 0-64 16-75 0-14 0-95 | 19-09 Pennisetum clandest. Ae 2-81 3-89 0-69 1-19 4-78 | 13-36 Phalaris arundinacea .. 0-78 3-16 31-15 0-54 0:63 | 36-26
* The most abundant carbohydrate group in each species is given in bold type. + Analysed by Miss E. Goldsmith. Starch determined by digestion with yakadiastase.
64 The Journal of South African Botany.
and of Tristachya hispida ; rhizomes of Cynodon dactylon and of Penni- setum clandestinum). The dextrin content exceeded 1 per cent. only in the rhizomes of Pennisetum clandestinum.
Though the analytical methods used, involving digestion with saliva, indicated the presence of starch in the roots and rhizomes of all species, only a few contained starch in large amounts and as the principal carbo- hydrate. The highest percentages of starch were recorded in the rhizomes of Cynodon dactylon (viz., 10:3 and 12-5 per cent.). A similarly high starch content, namely 13-4 per cent., has been recorded in Cynodon dactylon rhizomes by De Cugnac (4), whereas Julander, a recent American worker, considers that not starch but “ glucosans ” (i.e., dextrins) are the important form of carbohydrate reserve in this grass (7). Another instance of a high starch concentration is provided by the roots of Tristachya hispida (7-7 per cent.). This finding is in agreement with previous results obtained by the senior author (13, 15).
In addition to the chemical analyses, qualitative starch tests were carried out on the tissues of ten of the native species. Sections of leaves, stems, roots and rhizomes were cut, treated with iodine solution, and examined microscopically. The results are summarised in Table IV.
TABLE IV.
Results of Qualitative Starch Tests in Tissues of South African Grasses.
Species. Leaves. Stems. Roots. | Rhizomes. Brachiaria serrata + (1) (2) — —— Cynodon dactylon . . : + (1) (2) | +++ (3) | ++ (6) | +++ (4)
' Digitaria tricholaenoides . . + (1) (2) ++ (3) + (6) ++ (3) Elyonurus argenteus + (1) — — Hragrostis chaleantha + (1) ++ (3) + (6) Harpechloa falx + (1) (2) — ++ (6) Hyparrhenia hirta + (1) (2) — -- Themeda triandra + (2) + (5) — Trachypogon plumosus + (1) (2) — = Tristachya hispida + (1) (2) +--+ (3) | +-+ (6) Key: — = Starch absent. -+- = Starch present in small amounts. ++ = Starch present in moderate amounts. +-+-+ = Starch present in large amounts. (1) Starch present in outer bundle sheath. (2) Starch present in leaf sheath, at junction with node, between each fibro-vascular bundle and inner epidermis.
(3) Starch present in parenchyma of cortex and ground tissue.
(4) Starch present in parenchyma of ground tissue.
(5) Starch present in parenchyma of ground tissue at node only.
(6) Starch present in pith parenchyma.
Reserve Carbohydrates in South African Grasses. 65
Starch was detectable in all the species, not only in the subterranean
organs, but also in the aerial parts. Where starch was present in the leaf blades, it invariably occurred in the chloroplasts of the outer bundle sheath. The colour obtained here with iodine solution was usually, however, not the typical blue, but rather a black-brown, even when the chlorophyll had been extracted with alcohol previous to testing. It is nevertheless considered that the stain was due to small amounts of starch. Typical blue starch grains were found in the leaf sheaths at the junction with the node, even in species where the iodine test gave doubtful or negative results in the leaf blades. The accumulation of starch in the plastids of the bundle sheath has been observed by Rhoades and Carvalho (9) in maize and sorghum. _ The results of the qualitative starch tests on the roots and rhizomes largely confirmed those of the chemical analysis. In the three species where quantitative analysis revealed the presence of larger amounts of starch (viz., Cynodon dactylon, Harpechloa falx and Tristachya hispida) distinct accumulations of typically-stained starch grains were found in the tissues. Positive tests were also given by the roots of Digitaria tricholaenoides and Eragrostis chalcantha. The rhizomes of Cynodon dactylon were strikingly rich in starch (see Plate III).
That the results of the chemical analysis indicated very small amounts of starch in a number of species where no starch could be detected micro- scopically, may be due to the hydrolysis by saliva of substances other than starch (5) or else to incomplete extraction of sugars or fructosans, a small fraction of which may have remained in the residue used for the starch determination. This was most probably also the case in the roots and rhizomes of Agrostis tenuis, Arrhenatherum elatius, Lolium perenne and Phalaris arundinacea, where no starch grains could be detected, though chemical analysis indicated small quantities of starch.
According to De Cugnac (4), grasses-can be divided into two groups, namely, those which accumulate in their vegetative organs sucrose with or without starch (‘* Graminées sacchariféres ”’), and those which store fructosan, usually together with sucrose (“‘ Graminées lévuliféres ”’). De Cugnac’s tentative suggestion that the latter group essentially com- prises grasses native to cool temperate climates, while the “ Graminées sacchariféres ” are mostly adapted to warm regions, seems to be confirmed
by the results of the present investigation. In none of the indigenous South African grasses studied did fructosans occur in appreciable amounts. Most of the grasses contained non-reducing sugars, presumably in the form of sucrose, as their principal carbohydrate. Some contained high percentages of starch, the latter carbohydrate actually constituting the main carbohydrate reserve in several cases. General statements offered
66 The Journal of South African Botany.
by some workers asserting the negligible importance of starch as a reserve carbohydrate in the vegetative organs of grasses are thus shown to be not permissible, since the type of carbohydrate reserve is apparently largely a characteristic of the species concerned.
II. SEASONAL CARBOHYDRATE TRENDS IN THE SHOOTS OF Four South AFrrican HIGHVELD GRASSES.
The results of the herbage analyses are indicated in Table V and Figure 1. The shoots of all four species examined contained reducing
TABLE V.
Seasonal Carbohydrate Trends in the Shoots of Four South African Highveld Grasses-
Expressed as percentage of the dry matter.
Sampling Reducing Non- Total Total Date. Sugars. reducing .| Sugars. Starch. available Sugars. Carbohydrate.* Brachiaria serrata. 24/11/1943 0 -42 1-23 1 -65 0 -69 2-34 15/ 2/1944 0 -43 0 -99 1 -42 0 -30 1-72 18/ 4/1944 1-29 0-75 2 -04 0 -23 2-27 28/ 6/1944 0-39 0-09 0 -48 0 -22 0-70
24/11/1943 0 -68 1 -68 2 -36 0 -70 3-06 15/ 2/1944 0 -59 1 -20 1-79 0-09 1-88 18/ 4/1944 1 -64 1-98 3 +62 0-04 3 66 28/ 6/1944 2-26 0-32 2-58 0-06 2 -64 Themeda triandra. 24/11/1943 1 -26 2-22 3-48 0 -64 4-12 15/ 2/1944 0-45 2-08 2 -53 0-19 2-72 18/ 4/1944 1-67 1-98 3°65 0-07 3-72 28/ 6/1944 1-05 0 -25 1-30 0-13 1-43 Trachypogon plumosus.
24/11/1943 0 -66 1-37 2-03 0 -42 2 -45 15/ 2/1944 0-17 0 -98 1-15 0 -26 1-41 18/ 4/1944 0-74 0-97 I oil 0-13 1-84 28/ 6/1944 0- 0 . 1:
* Total available carbohydrate = combined sugars and starch.
Reserve Carbohydrates in South African Grasses. 67
SUGARS & STARCH
STARCH
NON-REDUCING SUGARS REDUGINGESUGARS meerenere eee eeecee
BRACHIARIA SERRATA
HYPARRHENIA HIRTA
THEMEDA TRIANDRA
TRACHYPOGON PLUMOSUS
24.11.43.
Fig. 1. Seasonal carbohydrate trends in the shoots of four South African Highveld grasses.
68 The Journal of South African Botany.
sugars, non-reducing sugars and starch, whereas fructosans and dextrins could not be detected in measurable quantities. There was much variation in the relative proportions of these substances, but in all instances sugars were present in larger quantities than starch. Total sugars ranged from 0-48 to 3-65 per cent., and starch from 0-04 to 0-70 per cent.
Total sugars decreased in the four species in early summer (1.e., between 24th November and 15th February) and rose again in late summer and early autumn. In three of the four species the maximum percentage of total sugars was recorded on 18th April, 1944. Starch, on the other hand, occurred in maximum amounts in all species in spring (24th November), and decreased distinctly with the advancing season, though in three species somewhat higher amounts were found in winter (28th June) than in autumn (18th April). Total sugars, as well as the combined percentages of sugars and starch, showed a distinct fall during late autumn and winter (i.e., from 18th April to 28th June).
It is known that the sugar and starch content of leaves is subject to considerable diurnal variations, largely depending on meteorological conditions influencing the rates of photosynthesis and respiration. Carbo- hydrates elaborated in the Jeaves are used to build up new tissues or are temporarily stored in the stems. While meteorological conditions pre- vailing on the sampling dates may have influenced the results to some extent, it is considered that the recorded changes in carbohydrates largely represent seasonal variations, since entire shoots were analysed, and not only the leaves. As considerable growth took place in all four species after the first sampling date, involving the formation of stems and in- florescences, the decrease in the percentages of sugars and starch between the first and second sampling date does not necessarily imply a decrease in amount. On the contrary, it is highly probable that these substances increased in actual amount between 24th November and 15th February. The decrease in the percentages of sugars and starch during this period is probably due to the fact that dry matter increased more rapidly in amount than did these carbohydrates. By 15th February the flowering period was over in three of the species, while Hyparrhenia hirta was still in the flowering stage. The flowering period marks the end of rapid expansion of the aerial parts, and is followed by the period of maturation. It is significant that during the latter stage (i.e., between the second and third sampling date) total sugars as well as combined sugars and starch increased appreciably in percentage in all four species. When condi- tions are favourable for photosynthesis, and no further growth is taking place, carbohydrates are elaborated in excess after flowering, and conse- quently increase in percentage.
When the third sample was taken (18th April) typical autumn
Reserve Carbohydrates in South African Grasses. 69
weather had already set in, though no frosts had occurred. The process of ageing and gradual death of the aerial portions during late autumn was associated with the loss of an appreciable portion of the accumulated sugar. Some of this sugar may have been utilised in the aerial parts in respiration or lost by decomposition, while part of it was probably translocated to the roots. As shown in earlier investigations (13), the sugar content in the roots of Highveld grasses usually increases from February onwards, i.e., soon after flowering. At this stage the excess of carbohydrates elaborated by the plant is so great that these com- pounds increase in percentage (and actual amount) in the aerial parts though appreciabie portions are translocated to the roots at the same time. The removal of carbohydrates from shoots to roots is, however, not a complete one, as considerable amounts of sugars and starch were found in the herbage collected in winter (28th June). These ranged from approximately 30 to 70 per cent. of the amounts present in autumn (18th April). Bunting (2), who determined the sugar content of the stem bases of three grass species harvested at Frankenwald between 18th July and 6th August, 1937, obtained the following values :—
Total Sugars as Percentage of Combustible Dry Matter.
Species. Range. Average. Tristachya hispida .. hie me: ie 0-89 — 9-95 2-91 Trachypogon plumosus =i Re 20 1 -28 —7-30 a} oli! Brachiaria serrata .. oi tis ai 1 -62 —5-01 2 -48
It may be assumed that the carbohydrates which are temporarily stored in the stem bases can be utilised for new growth, at least as long as these plant parts are alive. Such utilisation may occur, for instance, after mowing. Whether the residue of sugars and starch in old stems can be utilised for spring growth in the following season, is, however, not known.
Table VI and Figures 2 and 3 show the relative proportions of reducing sugars and starch. Reducing sugars, expressed as percentage of total sugars, increased in the shoots of Brachiaria serrata and Hyparrhenia hirta right through the season, and in Themeda triandra and Trachypogon plumosus at least from February onwards. Thus with the advancing season reducing sugars increased at the expense of non-reducing sugars, so that in winter the former exceeded the latter in actual amount (Fig. 1 and Table V). These changes can be interpreted as being due to increased
70 The Journal of South African Botany.
TABLE VI.
Relative Proportions of Reducing Sugars and Starch in the Shoots of Four South African Highveld Grasses.
Sampling Brachiaria
Date.
Hyparrhenia Themeda Trachypogon
serrata. hirta. triandra. plumosus.
Reducing Sugars.
Expressed as percentage of total sugars.
24/11/1943 25-5 28-9 36 -2 32-5
15/ 2/1944 30-3 33-0 17-8 14-8
18/ 4/1944 63 -2 45-3 45-7 43-3
28/ 6/1944 81-3 87-6 80-8 68-3 Starch.
Expressed as percentage of total available carbohydrate.* 24/11/1943 29-5 22-9 15-5 17-2 15/ 2/1944 17-5 4-8 v0) 18-5 18/ 4/1944 10-1 iL ell 159) 7-1 28/ 6/1944 31-4 2-3 §) oll 29 -2
* Total available carbohydrate = combined sugars and starch.
hydrolysis of sucrose, and to a more rapid rate of translocation of sucrose, as compared with that of hexoses.
Starch, expressed as percentage of total available carbohydrate,* decreased in three of the four species between 24th November and 18th April, and in T'rachypogon plumosus between 15th February and 18th April. This indicates that starch synthesis was most active during the early stages of growth and gave way to hydrolysis of starch with ageing and maturation. In winter, however, there was an increase in the relative proportion of starch. Why the winter starch content of the dry matter in three of the four species exceeded that in autumn (Table V and Fig. 1), is difficult to explain. Though it is possible that during late autumn and early winter some starch was formed from sugars or by the break-down of more complex polysaccharides, it is more likely that these fluctuations were due to sampling or other errors, particularly since the percentages of starch were in most instances very small.
*“* Total available carbohydrate’, in this case, means combined amounts of sugars and starch, since fructosans and dextrins were absent.
Reserve Carbohydrates in South African Grasses. 71
BRACHIARIA SERRATA HYPARRHENIA HIRTA THEMEDA TRIANDRA TRACHYPOGON PLUMOSUS
24.11.43.
Fic. 2. Reducing sugars in the shoots of four South African Highveld grasses (expressed as percentage of total sugars).
BRACHIARIA SERRATA HYPARRHENIA HIRTA THEMEDA TRIANDRA TRACHYPOGON PLUMOSUS ......-......----
24.11.43.
Fic. 3. Starch in the shoots of four South African Highveld grasses (expressed as percentage of total available carbohydrate).
72 The Journal of South African Botany. SUMMARY.
The results of carbohydrate determinations on the roots and rhizomes of thirteen indigenous South African grasses and of six exotic grasses are reported and discussed. Reducing and non-reducing sugars, fructo- sans, dextrins and starch were estimated in this material.
In the majority of the South African species investigated, non-reducing sugars formed the principal reserve carbohydrate, while in some starch occurred in largest amounts. Results of qualitative tests for starch in sections confirmed the presence of this carbohydrate in both aerial and underground parts of a number of species.
None of the South African species studied was found to contain appreciable amounts of fructosans or dextrins. The presence of con- siderable quantities of fructosans was confirmed in four of the exotic grasses in which fructosans had previously been reported.
Seasonal carbohydrate trends were investigated in the shoots of four indigenous South African grasses. The shoots of all four species were found to contain reducing sugars, non-reducing sugars and starch, but no fructosans or dextrins ; sugars were present in larger amounts than starch.
Total sugars, as well as combined sugars and starch, increased in the shoots of these grasses after flowering, and decreased during maturation, i.e., in autumn and early winter. The results suggested that carbo- hydrates are elaborated in the leaves in excess after flowering, and are consequently translocated to the roots, though considerable portions of sugars and starch remain in the aerial parts during winter.
A new technique of estimating fructosans and dextrins is described.
ACKNOWLEDGMENTS.
The writers wish to express their sincere thanks to Professor John Phillips, Head of the Department of Botany, University of the Wit- watersrand, for having provided the facilities which made this research possible.
Thanks are also due to Miss E. Goldsmith and Miss N. Katz for their co-operation, and to Mr. A. Salbany for the photographs.
Reserve Carbohydrates in South African Grasses. 73
REFERENCES.
(1) AncuBoup, H. K.: “ Fructosans in the Monocotyledons.” A review. The New Phytologist 39: 185-219. 1940.
(2) Buntine, A.: “The Effect of Combined Clipping and Chemical Treatments on the Root Reserves and Development of some Highveld Grasses.” Unpublished M.Sc. Thesis, University of the Witwatersrand, 1938.
(3) Coox, L.: “A Report on the Veld-burning Experiments at Crescent Creek, Milner Park, Johannesburg.”’ Unpublished B.Sc. Hons. Thesis, University of the Witwatersrand, 1938.
(4) De Cuenac, A.: “ Recherches sur les Glucides des Graminées.”” Annales des Sciences Naturelles, 10e séries, 13: 1-129, 1931.
(5) Denny, F. E.: “Improvements in Methods of Determining Starch in Plant Tissues.”” Contributions from Boyce Thompson Institute, 6: 129-146, 1934.
(6) Huspparp, R. S., and Loomis, T. A.: “The Determination of Inulin.” Jour. Biol. Chem. 145: 641-645, 1942.
(7) JULANDER, O.: “ Drought Resistance in Range and Pasture Grasses.’ Plant Physiology 20: 573-599, 1945.
(8) McRary, W. L., and Suarrery, M. C.: “ The Colorimetric Determination of
Fructosan in Plant Material.” Jour. Biol. Chem. 157: 161-167, 1945.
(9) Ruoapss, M. M., and CarvatHo, A.: “‘ The Function and Structure of the Parenchyma Sheath Plastids of the Maize Leaf.’ Bull. Torrey Bot. Club 71: 335-346, 1944. (Ref. Biol. Abstr. 18, No. 20461, 1944.)
(10) Roz, J. H.: “A Colorimetric Method for the Determination of Fructose in Blood and Urine.” Journ. Biol. Chem. 107: 15-22, 1934.
(11) Sunttvan, J. T., and Spracus, V. G.: “*‘ Composition of the Roots and Stubble of Perennial Ryegrass following Partial Defoliation.”” Plant Physiology 18: 656-670, 1943.
(12) Wertnmann, H.: “Storage of Root Reserves in Rhodes Grass.” Plant Physiology 15: 467-484, 1940.
(13) —————————— “Seasonal Chemical Changes in the Roots of some South African Highveld Grasses.”” Jour. §.A. Bot. 6: 131-145, 1940.
(14) “ Effects of Defoliation Intensity and Fertilizer Treatment on Transvaal Highveld. Emp. Jour. Exp. Agr. 11: 113-124, 1943.
(15) “Root Reserves of South African Highveld Grasses in
Relation to Fertilizing and Frequency of Clipping.” Jour. §.A. Bot. 10: 37-54, 1944.
(16) ““Semi-micro Estimation of Reducing Sugars.” Plant Physiology 19: 148-156, 1944.
(17) Yemno, E. W.: “ The Respiration of Barley Plants: I. Methods for the Deter- mination of Carbohydrates in Leaves.”” Proce. Roy. Soe. London, Series B, 117: 483-504, 1935.
— jes, *
Micro-photograph of transverse section of rhizome of Cynodon dactylon, stained with iodine solution (magnification 92 x). The secticn was so cut as to decrease in thickness from left to right. The intact parenchymatous cells of the ground tissue on the left are densely packed with starch grains, in the right-hand part
many starch grains have fallen out so that individual starch grains can be seen in the cells.
Micro-photograph of transverse section of rhizome of Cynodon dactylon, stained with iodine solution (magnification 510 x ). Showing individual starch grains in parenchymatous cells of the ground tissue.
Pratt III,
REVIEWS AND ABSTRACTS.
A. W. Samrson: Plant Succession on Burned Chaparral Lands in Northern California. University of California Bull. 685, 1944. 144 pp., 46 figs., 17 tabs.
This bulletin will be of particular interest to those who are concerned about veld-burning in the sclerophyll scrub vegetation of the South- western Cape, because many of the problems associated with burning of chaparral vegetation in Northern California are analogous to those encountered here. The author is Professor of Forestry at the University of California and Plant Ecologist at the California Agricultural Experi- ment Station. He is an authority on the management of pasture lands, and has conducted studies of succession on burned chaparral areas over a period of fifteen years.
The bulletin is introduced by a brief deseription of ecological charac- teristics of chaparral vegetation followed by a historical résumé of burn- ing practices. The chaparral, like the sclerophyll scrub of the Cape, was burnt by indigenous races before Kuropean settlement. Sampson considers, however, that the burning was not sufficiently extensive or frequent to have a marked effect on the vegetation. Burning apparently reached maximum proportions through the activities of the white settler, in the eighties, but has since somewhat decreased through enforcement of improved legislation. The State Board of Forestry does not attempt complete protection of all lands under its control. Three vegetation zones are recognised. Zone 1 consists of the more valuable timber and water-shed lands, and also includes state parks and monuments, as well as other such areas subject to serious economic loss by fire. The lands in Zone 2 are less important in the matter of fire protection. They form a “ buffer’ zone. Zone 3 includes grassy valleys and agricultural lands, which are chiefly of local interest.
From a study of the literature Sampson concludes that no single criterion may safely be used to predict the outcome of burning. “~ An unusually dry or wet season preceding or following burning may un- predictably affect the plant population of burned areas. The seed of some inconspicuous plant which has accumulated in the soil for many years may germinate in large numbers under the suddenly changed conditions brought about by burning, and thus result in the conspicuous presence of that particular species after the fire. Only as a result of
76 The Journal of South African Botany.
careful study in a particular area, taking into account all possible biotic, climatic and topographic factors, can rational prediction of plant succes- sion be made.” This conclusion applies equally well to the sclerophyll scrub of the Cape.
The main portion of the bulletin is devoted to the results of sample plot and other studies of plant succession on burns. Many of these results will be of considerable interest to Cape botanists and numerous interesting comparisons between the two vegetational regions could be made. Conclusions, which appear to support observations made in sclerophyll vegetation at the Cape, are: that burning at infrequent intervals may, and often does, extend the chaparral cover; that the heat generated by a fire, activates the seeds of some chaparral species, which apparently le dormant for long periods under the unburned ground cover; that the root crowns of sprouting brush species are re- markably adjusted to endure severe and frequent burning, so that vigor- ous sprouting takes place after a fire. An outstanding feature of the succession on burns in the Cape sclerophyll scrub not mentioned by Sampson, and which is therefore apparently not so striking in the chaparral, is the profuse flowering of numerous geophytic plants.
Sampson groups the chaparral species into trees and shrubs (sprouting and non-sprouting) ; broad-leaved herbs; and grasses. It would, how- ever, have been an improvement had the second group been divided into life-form classes, such as therophytes and geophytes. It would probably also have been more satisfactory for readers not intimately acquainted with the Californian flora, had the Latin names been used in the text instead of the popular ones.
The publication is readable and is recommended to all who are in- terested in the ecology of sclerophyll vegetation, and more particularly in the effects of fire.
C. L. Wicur.
S. D. Garrerr: Root Disease Fungi. 177 pp. Waltham, Mass., the Chronica Botanica Co., 1944. $4-50. Johannesburg, Central News Agency, Ltd.
The fungi which cause root diseases are members of many systemic groups and are alike only in sharing as their common habitat the soil and the roots which occupy it. The volume under review is in no sense a textbook of these fungi, but concerns itself with their way of life and their ecological relations to the soil and its other inhabitants. By
Reviews and Abstracts. 17
limiting himself mainly to the consideration of a relatively small number of economically important fungi, the author has been able to discuss in detail the varied and complex factors which decide whether root or fungus shall survive.
To the plant pathologist, this new volume from Rothamsted is a valuable progress-report of widely-scattered investigations ranging from 1858 to 1942; a bibliography of some 400 titles is appended. To the cultivator of plants, its six concluding chapters offer a review of practical measures of disease-control in crop-plants of the field, plantation and glasshouse. To both, it is a milestone along one of the many highways leading to a fuller understanding of the soil from which humanity draws its main sustenance.
EK. S. Moors.
James G. Horsrati: Fungicides and Their Action. Waltham, Mass., the Chronica Botanica Co. $5.00. Johannesburg, Central News Agency, Ltd.
Plant pathologists owe much to the Chronica Botanica Co. First it gave us Bawden’s ~ Plant Viruses and Virus Diseases,” then Garrett’s “Root Disease Fungi,” and now Horsfall’s “ Fungicides’, all written by masters of their subjects. Dr. Horsfall is a particularly happy choice. He has been intimately connected with, and has taken an active part in, that great advance of knowledge of fungicides which has occurred during the last decade or two. His own contributions run through almost the whole field of fungicides: the action of copper and sulphur compounds, inoculum potentials, field assays, problems of deposition, coverage and weathering, dosage-response relationships, antagonisms, synergism, and phytotoxicity. In his book one expects, and one gets, a standard of discussion which only authority and familiarity with the subject can give. Particularly valuable is the author’s ability to select from the mass of facts in the modern literature those which are specially significant and likely to start new lines of progress. One might instance the way he picks out the possibilities of using basic nitrogen compounds as antidotes for the toxins of vascular diseases and root rots. ‘Time may show some of the judgments faulty ; but the book is never dull, and Dr. Horsfall’s way of approach will be a stimulus to research in the whole field he covers.
The weak point of the book is the chemistry. Formulae are repeated correctly, but one gets the impression that the author is out of his element
The Journal of South African Botany.
oa
when it comes to the chemical side. One shudders, for example, to see chlorobenzene and dichloramine T in the same context. To pursue the same subject for a further example, dichloramine T is used as an illustra- tion of the notion that chlorination increases the fungicidal properties of organic substances which are fungicidal to begin with, whereas in fact dichloramine T shares its potency with other chloramines, even those derived from inert organic compounds like methylamine bemg fungicidal and the inorganic chloramines being as potent as, and more widely used than, the organic. One does not wish to harp on this weakness, for despite it the book is very good, and one only does so in the hope that there will be an improvement when the book reaches a second edition, which it richly deserves. The print is readable and the volume well got up.
J. E. VAN DER PLANK.
JospeH C. Gruman: A Manual of the Soil Fungi. Ames, lowa; The Collegiate Press, 1945. xiv 392 pp. 135 figs. Price $5.
When the first comprehensive treatise on soil fungi by Niethammer appeared (1937), one could not but help feeling disappointed at the in- completeness and inaccuracy of the work. Gilman’s book, though less pretentious, certainly marks an advance in the systematic treatment of the soil fungi. The author is well aware of the restrictions of his Manual: he has not attempted to give many original data or descriptions, but wanted to provide “a tool in the hands of investigators that will enable them to identify the soil fungi which they may encounter in the course of their work.” This sympathetic and modest attitude of the author makes it well-nigh impossible to criticize the book which, in general, is well produced. It certainly will find its users, not only among soil mycologists, but also among others who have to identify fungi and the considerable use which its predecessor, Gilman and Abbott's Summary (1927), has found, will guarantee a wide distribution of the Manual.
Within the framework of the restrictions indicated by the author, there are a few omissions that must be mentioned. In particular the German and Russian literature has not always been considered fully and important publications like those of Duché and Heim (1931) and of Campbell on the Scottish Mucorales have been ignored. This results in the omission of a few typical soil fungi and in a certain incompleteness in the otherwise very extensive geographical records.
Reviews and Abstracts. 79
A much more serious criticism can be made about the illustrations. For each genus there is a small line drawing (never exceeding 2 by 24 inch), usually redrawn from some more or less “ classical” (though by no means always good!) illustration. In several cases the illustrations have suffered from redrawing. I cannot always approve of the author’s choice and I certainly cannot see why, at the present stage of mycology, hundred-year-old pictures from Corda and Bonorden should be copied again and again, sometimes even in a still more mutilated form than in Lindau’s works. In several cases the source is given wrongly or in- correctly. Another serious objection is that the species illustrated is never mentioned either in the captions or in the text. This is a dangerous principle as, strictly speaking, it is only possible to depict specimens or at most species and certainly not genera since in many cases the genera of Hyphomycetes represent very heterogeneous assemblies of species which will have to undergo considerable rearrangement in the future. It is therefore impossible to typify a genus in any other way than by a definite species. Moreover, a modern taxonomic treatise, even if it be a compilation, should not refrain from attempting to indicate generic types !
Though accepting fully the limitations of the book as expressed by the author, the question must be raised in how far systematic mycology and soil microbiology in general are served by compilations like this. For several decades systematic mycology has been in the grip of the Fries-Saccardo-Lindau tradition of shackling, artificial classification and compilation, and usually the more critical work of, for example, the French mycological schools (especially Vuillemin) has been overlooked or ignored. The same conservative attitude is adopted by Gilman who (though he mentions Mason’s earlier work and Vuillemin’s terminology of spore forms) states that “for the purposes of this volume the general term conidium covering all these types seems adequate ~. I must admit that a consistent application of Mason’s newer morphological results throughout the Deuteromycetes would at present be impossible, but much more use should be made of the principles already established. This would, of course, give ample opportunity for wrong interpretations (did not Mason himself frankly admit this ?) but these would surely be a challenge to critical investigations. And this is what is still badly needed in systematic mycology! I am afraid that compilations like Gilman’s Manual leave mycology more or less where it is and, besides, to tyros they offer no more advantage than the popular German publica- tions of twenty or more years ago, though most of the descriptions are more elaborate and certainly better.
W. J. LursenaRmMs.
XO) The Journal of South African Botany.
CiuaupE E. ZoBeLtt: Marine Microbiology. Waltham, Mass., Chronica Botanica Co. Price $5.00. Johannesburg, Central News Agency.
By marine microbiology, Dr. ZoBell means the study of marine bacteria, yeasts and moulds. Other small marine creatures (such as plankton) are dealt with only in relation to these. The author has made a number of important contributions to this field of work. The book is chiefly concerned with marine bacteria: their characteristics ; rdéle in the transformation of organic matter and sulphur compounds and in the nitrogen and phosphorus cycles of the sea: also as food for various animals. Consideration is given to marine microbiology in relation to the technology of sea foods and to such economic problems as deteriora- tion of nets and cork. The last chapter (Chap. X VIII) is entirely devoted to the microbiology of fresh water and waters of greater and lesser salinity than oceanic water, which subjects receive some attention throughout the book. j
The bibliography is extensive and consideration of published work is one of the main features of the book which ~ provides a key to the relevant literature “. As this is scattered and frequently inaccessible, Dr. ZoBell has rendered a considerable service. In such a general review gaps in our knowledge, controversial issues and instances of lack of quantitative data inevitably emerge, and this makes the book useful to workers and prospective workers in this field.
The book is well produced and in addition to 12 diagrams contains a sketch of the Scripps Institution and a number of pleasing vignettes (by A. B. Carlin). Reference to previous and subsequent pages where related subjects are considered, makes the book easy to use.
Wm. Epwyn Isaac.
JOURNAL
OF
SOUTH AFRICAN BOTANY VOL, XI,
PLANTAE NOVAE AFRICANAE.
“Ex Africa semper aliquid novi.’—Pliny.
SERIES XXVI.
By Proressor R. H. Compron, M.A., and Caprarn (8S) T. M. Savrer, R.N. (Ret.).
Amphithalea Fourcadei Compton. (Leguminosae-Liparieae.)
Fruticulosus ramosus, dense foliatus, omnino (praeter corollam) dense appresse argenteo-pubescens. Ramuli angusti, cicatricibus pro- minentibus. Folia erecta vel expansa, simplicia, integra, elliptica, sessilia, acuta, mesonevro infra prominente, marginibus parum incrassatis- Flores singuli, axillares, sessiles, foliis occulti. Calyx subaequaliter semi- lobatus. Petala purpurea, longe unguiculata, glabra. Ovariwm 1—2-ovu- latum. Fructus dehiscens, semine uno.
Hab. Cape Province. Humansdorp Division: Witte Els Berg, along eastern spur, 3,000 ft., Dec., 1925, Fourcade 3208; Kareedouw Pass, 1,700 ft., June, 1927, Fourcade 3242; Kareedouw Peak 2,000 ft.. 21 Dec., 1933, Compton 4556 (Type, in Herb. Nat. Bot. Gardens) ; Witte Els Bosch Peak, S.W. slopes, 16 Nov. 1941, Hsterhuysen 6766. Union- dale Division: Formosa Peak, Jan. 1940, Stokoe 7392: Helpmekaar Peak, 4,000 ft., 28 Jan., 1941, Compton 10449, Esterhuysen 4597. Knysna Division : Concordia, 700 ft., Keet 914. George Division: Montagu Pass, 1,500 ft., 2 Nov., 1894, Schlechter 5793.
A small much-branched densely leafy shrub. Branches, leaves, calyx and ovary covered with a dense “appressed silvery pubescence. Branches slender, the leaf-scars prominent. Leaves erect to spreading.
$2 The Journal of South African Botany.
entire, elliptical, sessile, acute, the midrib prominent below, 6—9 mm. long, 4—5 mm. wide, flat or somewhat inflexed, the margins slightly thickened. Flowers axillary, sessile, solitary, concealed among the leaves. Calyx 4 mm. long, sub-equally lobed about half-way down, the ventral lobe slightly the longest, the two dorsal lobes joined a little higher than the others: lobes narrow, attenuate, pubescent inside as well as outside. Vexzllum light purple, claw 2 mm. long, limb quadrate- orbicular, 4 mm. long and wide, emarginate, with short incurved auricles, the sides somewhat reflexed. Alae dark purple, claw slender, 2 mm.
Fic. 1. Amphithalea Fourcadei. Portion of plant, nat. size. 1. Flower, front view. 2. Flower, side view. 3. Calyx, laid out, seen from within. 4. Vexil- lum. 5. Alae, from within. 6. Carina, from outside. 7. Carina, from within. 8. Old stamen. 9. Stamens. 10. Gynaecium. 11. Legume and calyx. 12. Legume, sagittal section. All x 5. (Compton 4556.) Del. W. F. Barker.
long, limb 3 mm. long. Carina dark purple, claws 2 mm. long, limbs 3 mm. long. Ovary 1—2-ovulate. Fruit 1-seeded, dehiscent, 6 mm. long.
This new Amphithalea has been collected on several occasions and has been variously placed in herbaria as A. imbricata (L) Druce (=A. densa E. & Z.), A. violacea Benth. and A. phylicoides E. & Z. It is, however, well distinguished by its growth and leaf characters from all these species, and it occupies a distinct area along the Tsitsikamma Mountains from the Montagu Pass eastwards to the Kareedouw Pass. IT am indebted to Dr. H. G. Fourcade for assistance in connexion with this plant, which I name in his honour.
Plantae Novae Africanae. 83
Erica extrusa Compton. (Ericaceae-Ericoideae.) § Hermes.
Fruticulus humilis ramosus. Ramuli divaricati, ascendentes, minute puberuli. Folia 3-nata, erecta, non appressa, parum incurvata, linearia, suleata, apiculata, glabra. Flores terminales et axillares ad apicem ramulorum. Pedicel angusti, elongati, minute puberuli. Bracteae
| 6 5
Fie. 2. Hrica extrusa. 1. Portion of plant, nat. size. 2. Whorl of leaves x 5.
3. Leaffrom below x 5. 4. Leaf,side view x 5. 5. Flower x 10. 6.Corolla x 10. 7. Bracts x 10. 8. Sepal x 10. 9. Anther, back view x 20. 10. Anther, front view « 20. 11. Stamen, side view x 10. 12. Gynaecium x 10.
(Compton 16817.) Del. W. F. Barker.
remotae, anguste lineares. Sepala ad corollam firme appressa, rigida, sulcata, apiculata, marginibus minute glandulosis. Corolla glabra, rosea, urceolata, quadrangula, angulis inter sepala extrusis, segmentis erectis, ovatis, obtusis, denticulatis, tubo subaequalibus. Filamenta tenuia. Antherae inclusae, brunneae, late cristatae. Ovarium glabrum, Stylus angustus, stigmate exserto, capitellato. i atlas
84 The Journal of South African Botany.
Hab. Cape Province. Caledon Division: Palmiet River, Elgin’ Dec., 1941, Stokoe 7957: Palmiet River Valley, Grabouw, Jan., 1943, Stokoe 8577 ; East foot of Kogelberg, 16 Jan., 1944, Esterhuysen 9995 ; Aries Kraal, 18 Nov., 1944, Barker 3354, Compton 16508; 30 Dec., 1944, Compton 16817 (Type, in Herb. Nat. Bot. Gardens).
A small much-branched bushy shrublet, usually less than 20 cm. high. Branches divaricate, ascending, minutely puberulous. Leaves 3-nate, erect, not appressed to stem, a little longer than the internodes, linear, slightly incurved, 4—8 mm. long, 1-0—1-5 mm. wide, apiculate, suleate, glabrous. Flowers terminal and axillary towards the end of the branches. Pedicels, bracts and sepals red. Pedicels 4—6 mm. long, slender, minutely puberulous. /racts narrow-linear, minutely denti- culate, scattered $ to ? along pedicel, 1-5—2-0 mm. long. Sepals narrow, closely appressed to corolla, 2-5 mm. long, rigid, apiculate, suleate, with minutely glandular margins. Corolla urceolate, glabrous, pink, 4 mm. long, 3 mm. diam., quadrangular, the angles expanded at the base into 4 gibbosities between the sepals, the segments erect, broadly ovate, obtuse, minutely toothed, about equalling the tube in length. Filaments slender, 2 mm. long. Anthers included, smooth, brown, 1 mm. long, with a pair of broad lobed basal crests, the pore ? the length of the cell. Ovary glabrous, broadly ovoid. Style 2-5 mm. long, slender. Stigma capitellate, exserted.
This small summer-flowering heath grows in local societies on coarse sand or quartzite gravel derived from the Table Mountain Sandstone hills of the Palmiet River valley below Grabouw. The basal almost saccate protrusions of the corolla between the sepals are distinctive and have suggested the specific name, and the glabrous ovary and cristate anthers assist in separating it from other species of the § Hermes (into which, as a matter of fact, it does not fit very well).
Euryops breviloba Compton. (Compositae-Senecioneae.)
Frutex erectus, paullum ramosus, omnino glaber. Rami teretes, medullati. Folia viridia, erecto-patentia, pinnati-partita, lobis utrinque 4—7, oppositis vel sub-oppositis, divergentibus, inaequalibus, linearibus, supra canaliculatis, apiculatis. Capitulum solitarium, pedunculo longo, nudo, erecto, terminale. Jnvolucrum labriforme, lobis c. 21, acuminatis. Receptaculum planum. Flosculi radii ec. 21, ligulati; flosculi disci numerosi: omnes flavi. Achaenia glabra. Pappi setae serrulatae, caducae.
Hab. Cape Province. Ceres Division: Karoo Poort, 25 Aug. 1935, Compton 5425; 27 Sept., 1944, Compton 16069 (Type, in Herb. Nat. Bot. Gardens); 27 July, 1941, Bond 1191, Esterhuysen 5494; Bolus
Plantae Novae Africanae. 85
2613 ; cult. at Kirstenbosch, 357/44 : Ceres Wild Flower Show, 30 Sept., 1928, Hutchinson 619. Clanwilliam Division: Matjesrivier, Cedarberg, Aug., 1943, G.H.H. Wagener 112.
An erect little-branched shrub up to 100—150 cm. high, glabrous throughout, and with little or no surface deposit of resin. Stems terete, smooth, slightly decurrent-angled below edges of petiole, with a large pith. Leaves green, erecto-patent, 5—10 cm. long, once pinnatipartite, lobes 4—7 on each side, opposite or sub-opposite, spreading, the middle lobes the longest, up to 2 cm. long, usually shorter, rachis and lobes linear, c. 2 cm. wide, furrowed above, apiculate. Capitula solitary, terminal on an erect naked peduncle 10—18 em. long. Involucre 12—18 mm. diam., flat-based with erect sides c. 5 mm. deep, bordered by c. 21 erect narrowly triangular lobes which are c. 4—5 mm. long, acuminate with puberulous margins and tips. Receptacle flat or slightly concave, honeycombed. fay-florets c. 21, female, ligulate yellow, tube 4 mm. long, limb 30—40 mm. long, 5 mm. wide, oblanceolate, obtuse. Disc- florets numerous, hermaphrodite, tubular, yellow, 6 mm. long gradually tapering to base, lobes 1 mm. long. Achenes obovate with 5 longitudinal ridges, glabrous. Pappus of several white serrulate bristles, 2—3 mm. long, caducous.
A bandsome plant, more erect than the well-known Huryops Athanasiae Less., and not so tall or straggling, and well distinguished therefrom by the much shorter lateral segments of its leaves. The leaves also distinguish it sharply from H. Serra DC. (with which it has been much confused in herbaria) which has narrow serratilobed leaves with many teeth on each side forwardly directed.
De Candolle’s description of H. Serra reads “ glaber, foliis elongatis sessilibus coriaceo-rigidis utrinque lobulis dentibusve grossis acutis dis- tantibus utrinque 10—12 pectinato-serratis . . .” (Prod. VI, 444).
E. breviloba is a conspicuous plant just outside the debouchment of Karoo Poort on the Ceres-Calvinia Karoo, on quartzite rocks at the foot of the Zwartruggens Mountains. It maintains its distinctive characters in cultivation.
E. Serra is a more widely distributed species, though apparently never very abundant : it occurs in localities with a much higher rainfall than EF. breviloba (e.g. Tulbagh, Drege ; Roodesand Pass, Tulbagh side, Stokoe, 1 Aug., 1938; slopes of Great Winterhoek Mts., near Saron, Nov., 1941, Stokoe 8116 ; S. slopes of Mt. Lebanon, Houw Hoek, Pillans 2651).
¢
Euryops Wageneri Compton. (Compositae-Senecioneae.) Frutex erectus, paullum ramosus, omnino glaber. Folia erecto-
86 The Journal of South African Botany.
patentia, pinnatipartita, lobis anguste linearibus, erecto-patentibus, subelongatis. Capitulum solitarium, pedunculo erecto, robusto, nudo, terminali. Jnvolucrum labriforme, supra parum latiore, lobis c. 21,
x A B Cc D
Fic. 3. Leaves of Euryops spp., nat. size. A. EH. breviloba (Compton 16069). B. E. Wageneri (Wagener 101). C. #. Serra (Pillans 2651). D. E. Athanasiae, rather smaller than the average. (Compton 9497.) Del. W. F. Barker.
attenuatis.. Floseult radi. ¢.)21, ligulati, aurantiaci,, speciosi : flosenls discy numerosi.. Pappi setae parum_plumosae, caducae..
Hab. Cape Province. Clanwilliam Division: Cedarberg, ne. a Sept. 1943, G. 2. Wagener 101 (Type, in Herb. Nat. Bot. Gardens) : cult. at Kirstenbosch 354/43 :_Ezelbank, 4,000 ft., Schlechter. 8836..
A tall, erect, little-branched shrub, stove throughout (oe ike
Plantae Novae Africanae. 87
tips and margins of the involucral segments). Leaves relatively numerous and close-set, erecto-patent, once pinnatipartite, 5—8 cm. long, lobes about 4—6 on each side, opposite or sub-opposite, erecto-patent, 1—3 em. long, apiculate, rachis and lobes furrowed above, c. 1-5 mm. wide.
Peduncle terminal, erect, stout, terete or finely striate, nude, up to 25 em. long and 4 mm. diam. Capitulum solitary. Involucre flat-based or slightly intruse, up to about 15 mm. diam., broadening above to about 20 mm. diam., with about 21 erect teeth which are 7—8 mm. long with attentuate points which are puberulous tipped and margined. Ray-florets c. 21, female, orange-coloured, tube c. 4 mm. long, limb up to c. 42 mm. long, 10 mm. wide, oblong-lanceolate, obtuse, with a minute apiculus. Disc-florets numerous, hermaphrodite, tube 6-5 mm. long, lobes 1-5 mm. long. Pappus 2:5 mm. long, of straight or slightly flexuous bristles, slightly plumose, caducous.
This conspicuous shrub bears a general resemblance to H. Athanasiae Less. from which it differs in the smaller leaves with shorter segments, the relatively longer and stouter peduncles, the deeper involucral cup and longer segments, and the orange colour of the ray-florets which are distinctly wider and give the capitulum a more handsome appearance. I name it in honour of the late Mr. G. E. H. Wagener, J.P., of Cedarberg, an enthusiastic observer of the plants of his district, whose death deprived the Kirstenbosch Herbarium of a valued contributor.
Euryops breviloba and E. Wageneri together with HL. Serra DC. and HE. Athanasiae Less. form a natural group of species: they are readily distinguished from one another by their leaves, typical examples of which are illustrated in Fig. 3.
Marasmodes undulata Compton. (Compositae-Anthemideae.)
Fruticulus diffusus, ramis divaricatis, rigidis, glabris, sparse glandu- losis. Folia alterna, erecto-patentia, sessilia, linearia, glabra, glanduloso- punctata, mucrone parum reflexo. Capitulum solitarium, terminale, sessile. Jnvolucrum obconicum; bracteae numerosae, multiseriatac, exteriores foliaceae marginibus apicibusque scariosis, imteriores anguste obcuneatae, fere totius scariosae, apicibus latioribus, obtusis, undulatis, sub-expansis. Flosculic.21. Corolla infra tubulosa, supra campanulata, parum glandulifera, Jobis expansis. Antherae basi obtusae, parum ex- sertae. Stylus robustus. Achaenia glabra, in humiditate mucosa. Pappi squamae paucae, inaequales.
Hab. Cape Province. Paarl Division : gravelly flats north of Huguenot, 26 April, 1946, Compton 17987. (Type, in Herb. Nat. Bot. Gardens.)
A small diffuse shrub. Branches terete, slender, divaricate, rigid, glabrous, with a few scattered superficial glands. Leaves alternate,
88 The Journal of South African Botany.
sessile, erecto-patent when young, spreading when old, linear, slightly auricled at base, glabrous, glandular-punctate, with a short slightly reflexed apiculus, 4—6 mm. long, 1 mm. wide. Capitula solitary, ter- minal, sessile, 7—8 mm. long, 6—7 mm. diam. Jnvolucre obconic, the bracts numerous, multiseriate, the outer ones foliose with a narrow
Fic. 4. Marasmodes undulata. 1. Portion of plant, nat. size. 2. Leaf from below x 5. 3. Leaf, side view x 5. 4. Capitulum x 5. 5. Part of re-
ceptacle and bracts x 5. 6—9. Bracts x 5. 10. Floret x 10. 11. Pappus scales x 10. 12. 3 stamens x 10. 13. Style and stigma x 10. 14. Achene after boiling x 10. (Compton 17987.) Del. W. F. Barker.
scarious margin and tip, grading to the innermost ones which are about 4 mm. long and 2 mm. wide, almost wholly scarious, narrowly obcuneate with a broader half-spreading obtuse undulate tip. Receptacle flat. Florets c. 21, glabrous, the tube cylindrical 1-5 mm. long, 0-8 mm. diam., the campanulate portion slightly glandiferous, 1-0 mm. long, the lobes spreading, ovate, yellow, 1-0 mm. long. Ovary 1-5 mm.
Plantae Novae Africanae. 89
long, glabrous. Pappus of a few (6—I10) unequal colourless scales, up to 0-8 mm. long. Anthers blunt at the base, shortly exserted. Style stout, lobes truncate, bristly at the tip.
An inconspicuous shrublet, distinguished from other species of Marasmodes by its rather larger solitary capitula with the innermost involucral bracts broadly scarious and undulate at their obtuse apices. The achenes have a surface coating of mucilage which swells up when wetted and becomes adhesive.
Relhania latifolia Compton. (Compositae-Inuloideae.)
Frutex erectus, rigidus, aromaticus, omnino glaber. Rami teretes, foliis occulti. Folia numerosa, coriacea, erecto-patentia, orbiculari- obovata, integra, perobtusa sed apiculata, multipunctata, venis incon- spicuis. Capitula ce. 5—10, in umbella simplice terminali, partim foliis occuita, breve pedunculata, cylindrica, basi obconica. Lnvolucri bracteae multiseriatae, erectae, imbricatae, scariosae. Paleae angustes, acutae, transparentes. Flosculi radii c. 16, breviter ligulati, feminei; flosculi disci tubulares, hermaphroditi. Achaenia glabra. Pappi squamae paucae, parvae, acutae. ;
Hab. Cape Province. Clanwilliam Division: Het Kruis, 29 Sept., 1943; Compton 15016 (Type, in Herb. Nat. Bot. Gardens), Leighton 85.
An erect, rigid, aromatic shrub, glabrous in all parts. Branches arcuate-erect, terete, concealed by overlapping leaves when young, with distinct leaf-scars later. Leaves numerous, coriaceous, entire, erecto-patent, green, orbicular-obovate, obtuse but apiculate, the veins inconspicuous, glandular-multipunctate on both surfaces, the margin slightly thickened, very uniform in size on the same branch, c. 6—9 mm. long, 5—7 mm. wide. Capitula c. 5—10 in a simple terminal umbel, their peduncles and bases hidden by adjoining leaves. Peduncles 3—4 mm. Jong, erect, slender. Jnvolucre cylindrical, obconic at base, 9—10 mm. long: bracts multiseriate, erect, imbricate, scarious, the innermost slightly diverging and more transparent, linear, 7 mm. long, 1 mm. wide. Paleae 6 mm. long, narrow, acute, transparent. lorets c. 16 ligulate and 36 tubular, yellow : ray-florets tubular except at the expanded tip, female, the tube 2-5 mm. long, the limb 2-5 mm. long, 1-0 mm. wide, the ovary 3-5 mm. long; disc-florets hermaphrodite, the ovary 3-5 mm. long, the corolla with a basal slender portion 3:0 mm. long, an upper wider portion 1-5 mm. long, and lobes 0-5 mm, long. Ovaries glabrous, slender. Pappus a cup of a few transparent pointed scales, 0-3 mm. long.
Relhania latifolia is related to R. squarrosa L’Hér., a widely distri- buted and “ variable ” species, and it might possibly be regarded as a
90 The Journal of South African Botany.
variety thereof. It is distinguished by its broad flat obtuse erecto- patent (not squarrose) leaves, its slightly larger capitula, and the narrower upper portions of the tubes of the disc-florets and ligulate portions of the ray-florets. Certain Piquetberg gatherings may form a connecting link between it and R. squarrosa.
Metalasia langebergensis Salter. (Compositae-Senecioideae) § Eu- metalasia.
Sectioni Condensatae accedit, atqui nulli specei singulari prope affinis. Squamae involucrales bi-seriatae admodum MM. quinqueflorae DC., sed spathulatae et apice cuspidatae et roseo-tinctae, non valde acutae albaeque. Fruticulus robustior, rigidior. Folia congesta, fasciculata. Capitula satis numerosa, 5-flora. Squamae extimae sub-membranaceae, non foliaceae. Corolla longior. Pappus apice vix ampliatus, acutus.
A rigid shrublet up to 60 cm. high, the younger branches sparsely lanate, rather densely leafy right up to the inflorescence. Leaves sessile, fascicled, narrow-linear, mucronate, spirally twisted in the lower half, glabrous beneath, strongly involute™, the tomentose upper surface almost concealed : subtending leaf in each fascicle up to 7 mm. long, the re- mainder shorter. Corymbs dense, hemispherical, slightly compound, 1-5—2 cm. in diam., with 20—80 congested capitula. Capitula cylin- drical, very shortly pedicellate, 6—7 mm. long, 1-6—2 mm. in diam., 5-flowered. Outer involucral scales spirally arranged in several ranks, woolly-matted in the lower part, acuminate, subspinous, reddish brown, and at length spreading at the apex, the outermost very small, subulate, gradually lengthening and broadening inwards, their shape almost merging into the shape of the inner scales : inner scales free, sub-biseriate, spathulate, cuspidate, white, more or less rose-tipped, about 1-2 mm. wide towards the apex. Pappus bristles white, acute, serrulate, scarcely thickened at the tip. Corolla narrow-tubular, red in the upper half, 3 mm. long, the lobes narrowly ovate-deltoid.
Hab. Cape Province: Riversdale Div.; summit of Garcia’s Pass, Salter 6771 (type in Bolus Herb.). Flowering July. Co-type has been sent to Kew.
This species belongs to the sub-section Condensatae, but has no very near affinity. The inner involucral scales are bi-seriate as in VW. quinqueflora DC., but spathulate, cuspidate and pink-tinged towards the apex, not sharply acute. It is a much more robust and rigid shrublet with closely fascicled leaves and with many more capitula in each head.
* In this and the following species of Metalasia the leaves, owing to the twisting at the base, have the appearance of being revolute with the tomentose surface facing
the ground, whereas morphologically they are involute, the tomentum being really on their upper surface.
Plantae Novae Africanae. 91
The outermost involucral scales are sub-membranous, not foliaceous, the corolla longer and the pappus bristles acute and scarcely thickened at the apex.
Metalasia rhoderiodes Salter (Compositae-Senecioideae) § Eu-metalasia. M. divergenti Don proxime accedit, sed ita differt :—Ramuli apicem versus plerumque roseo-tomentosi. Capitula numerosiora, lana rosea lenta inter se dense constricta. Squamae exteriores uniseriatae, in lana,
6, sub-uni- seriatae, albae vel raro pallide roseo-tinctae, interdum squamis 1—2
nisi tamen ad apicem, dense circumstrictae : interiores 5
parvis in intervallo additis.
A diffuse shrublet up to 40 em. high, with densely woolly-tomentose branches, usually pink towards the tips of the branchlets. Leaves fascicled, lanceolate, linear-lanceolate or oblong, pungent-mucronate or sub-apiculate, usually more or less involute and spirally twisted, white- tomentose above, cobwebby beneath, the subtending leaf in each fascicle commonly 0-5
1 cm. long, the axillary much smaller and rarely glab- rescent above. Inflorescences terminal, solitary or in shortly pedunculate pairs. Floral bracts linear-lanceolate or spathulate, acuminate, mucro- nate, lanate, the tips only free. Capita very compact, hemispherical, 1—1-5 cm. in diam. Capitula usually numerous, sessile, cuneate in outline, 4—4-5 mm. long, 5- or more rarely 3—4-flowered, bound together in the lower half with tough pink cottony wool. Outer involucral scales linear, acute, l-seriate, deeply embedded in the wool, the subulate tips only free: intermediate scales 1—2, free, small, white, scarious, or wanting: inner scales usually 5—6, sub-uniseriate, linear, white or rarely tinged with pink, acute, about 0-7 mm. broad. Pappus bristles serrulate, white, very slightly thickened towards the acute apex. Corolla 3—3-5 mm. long, the tube tapering slightly to the base, 0-6—0O-7 mm. broad at the apex, deep orange-red in the upper half: ; lobes narrowly ovate-deltoid, very short.
Hab. Cape Province: Paarl Div.; stream-side, Witte River Valley, Esterhuysen 8681 (type in Bolus Herb.) ; shale band between Bailey’s Peak and Pie Blane, 3,500 ft., Hsterhuysen 8524, Stokoe 8161; moist cliffs on S.E. slopes of Limietberg, 3,000 ft., Hsterhuysen 1601 ; Wellington Sneeuwkop, 4,000—5,000 ft., Hsterhuysen 13000: Worcester Div. ; Slangkop Mts., Cossacks, Hsterhuysen 8620: Caledon Div.; East side of Kogelberg, 2,000 ft., Leighton 877, Compton 16859, 16864; Palmiet River Valley, Stokoe 8730. Flowers Dec.—Apl.
This very distinct mountain species may be at once recognised by the pink cottony wool which binds the capitula together, the tomentum on the upper part of the branchlets being also frequently, though not in-
92 The Journal of South African Botany.
variably, pink. It is perhaps most nearly related to M. divergens Don, but differs in its far more numerous capitula and the uniseriate outer involucral scales, which are closely embedded, except at the extreme tip, in the pink wool of the flowering head, and so break away from the capitulum on dissection.
Metalasia riparia Salter (Compositae-Senecioideae) § Hu-metalasia.
M. muricatae Less. affinis sed ita differt :—Capitula valde breviora, in dimidio inferiore lana pallide aureo-viridi dense constricta. Swub- capita densissima, multiflora, glomerulis plerumque 4—10 itidem lanato- constrictis. Corolla valde brevior, inferne ampliata. Squamae involu- crales exteriores scariosae, glabrae, liberae, 2-seriatae: interiores 1- seriatae, minus acutae.
A large robust erect shrub 1-5—2-5 m. high. Branches ascending, the younger woolly-tomentose, rather densely leafy almost to the apex, straightish, usually with few branchlets. Leaves fascicled, spirally twisted. involute, lanate beneath, white-tomentose above, mucronate : subtending leaf in each fascicle upto 1 cm. long, the remainder much smaller. Flowers white. Inflorescence corymbose, often 3—5 cm. in diam., usually with 10—15 pseudo-umbellate tomentose branches about 1-5—2 em. long, each bearing one leaf-like bract, usually near the apex. Swub- capita 10—15, 1—1-5 em. in diam., compound, densely woolly-matted : glomerules 4—10, rarely less, shortly pedicellate, with an outer involucre of sub-fcliaceous mucronate bracts matted together below. Capitula about 10 in each glomerule, subsessile, 3-flowered, about 4 mm. long, the outer involucral scales 2-ranked, narrow-linear or subulate, scarious, sharply acute, white or pale straw-coloured, the inner 1-ranked, longer, linear, obtuse or subacute, white, 0-5—0O-6 mm. broad, shortly over- topping the corolla. Pappus bristles white, serrulate, broadening and oblanceolate at. the tip. Corolla tubular, yellow, about 2-7 mm. long, slightly swollen in the lower half, the lobes erect, deltoid, 0-3 mm. long. Achene (mature) oval, black, 1-5 mm. long, minutely papillose.
Hab. Cape Province: Caledon Div. ; streamside near Palmiet River mouth, Salter 5179 (type in Bolus Herb.) Levyns 5333; Platteberg and Paardeberg, near Palmiet River mouth, Stokoe 9075: streamside, Onrust River Mt., Hsterhuysen 4922. Flowers Nov.—Feb. Co-types were distributed to the British Museum and Kew.
Rather local and growing in dense thickets on stream sides. Its nearest affinity seems to be M. muricata Less., but it differs from all the many forms of that species as follows :—Capitula much shorter, closely bound together with pale greenish yellow wool in the lower half. Suwb- capita very dense, many flowered, the individual glomerules also bound
Plantae Novae Africanae. 93
together with wool. Corolla much shorter, somewhat swollen in the lower half. Outer involucral scales scarious, glabrous, free, in 2 ranks : inner scales in I rank, less acute.
Metalasia xanthocephala Salter (Compositae-Senecioideae) § Eu- metalasia.
M. aureae Don affinis sed ita differt :—Corymbus minor simpliciter subumbellatus, eramosus, apice planior. Flores saturatius flavi. Capi- tula latiora, basin versus minus attenuata, densius lanato-stricta. Jn- volucri squamae exteriores apicem versus non, carinatae, interiores non mucronulata. Pappus praecipue apicem versus minute plumosus. Corolla longior, in tubi faucibus non ampliata.
A rigid, widely branching shrub, 60—90 cm. high, the younger branches and branchlets thinly lanate. Leaves fascicled, linear, spirally twisted, sharply acute, subspinous, sparsely woolly or almost glabrous beneath, strongly involute, the white-woolly upper surface almost con- cealed : subtending leaf in each fascicle usually 0-6—I em. long, some- times reflexed, much longer than the remainder. Capitula cylindrical, 7—8 mm. long, 1-5—2 mm. in diam., 5-flowered, shortly pedunculate, usually 8—24 in close terminal simple corymbs 0-8—1-3 cm. in diam. : peduncles shortly white-woolly, 1—3 mm. long. Outer involucral scales in several series, spirally imbricate, closely matted together with white tough cottony wool below, the outermost minute, progressively lengthen- ing inwards, linear, smooth, acute, mucronate, pale yellowish brown, the 5 innermost scales equal, free, scarious nairowly liear-cuneate, obtuse or subtruncate at the apex, yellow-tipped, as long as the corolla, about 1:3 mm. broad at the apex. Pappus bristles slightly longer than the corolla, minutely plumose, particularly towards the apex, with a slightly widened yellow limb. Corolla about 4 mm. long, yellow, the tube tapering a little to the base: lobes ovate-deltoid, about 0-5 mm. long. Achene oval, 1-2 mm. long, minutely scabrid.
Hab. Cape Province : Malmesbury Div. ; Oude Post, between Malmes- bury and Hopefield, Salter 4407 (type in Herb. Bolus); Kalabas Kraal, Salter 7179; Cape Div.; flats 2 miles N.E. of Durbanville, Pillans 8529; Paarl Div.: between Mulder’s Vlei and Phesante Kraal, Salter 4393 ; low clay ridges 1} miles 8.W. of Joostenberg (Hercules Pillar), Pillans 7618; flats north of Paarl, Compton 18121; Tulbagh Div. ; near Wolseley, Salter 7301. Flowers April—June. Co-types were dis- tributed to the South African Museum, Cape Town, the British Museum and Kew.
An affinity of WM. aurea Don, but differing in its smaller simple flat- topped unbranched corymb, deeper yellow flowers, rather broader
94 The Journal of South African Botany.
capitula, less attenuate below and more densely woolly-matted, the outer involucral scales not keeled on the upper part and the inner scales not mucronulate, the minutely plumose, rather than serrulate pappus bristles and the corolla is longer and not appreciably widened at the throat of the tube.
The description of M. aurea in the Flora Capensis, Vol. III, p. 270, was compiled from dried specimens. The flowers are actually lemon- yellow and the corolla is about 3 mm. long with the tube distinctly widened at the throat. Harvey states that “the heads (capitula) are rather few or very numerous,” but they are, in fact, almost invariably very numerous. It seems possible that Drége’s small specimen in Herb.
SNES
i
N ‘
‘ .
] (e, 5 4.
Fie. 5. % Metalasia xanthocephala. 1. Pappus bristle x 10. 2. Capitulum x 6. 3. One of the five free involucral scales x 9. 4. Corolla and achene x 9. (Salter 4407.) Del. T. M. Salter.
Sond. from Berg River, marked M. octoflora DC., which Harvey says “in, no respect differs from MW. aurea” was, as its locality suggests, actually M. xanthocephala in which the heads are comparatively few-flowered. Harvey, having never seen the two species in the living state, may well have thought them the same and it is also significant that JZ. aurea is a species with a more easterly distribution.
It is, at any rate, inconceivable that De Candolle’s description of M. octoflora—branches adpressedly pubescent, capitula ovate-cylindrical, 8-flowered—could possibly have been applied to the species described here. If he had ever dissected the capitulum of this plant he must surely have noticed that the pappus bristles were minutely plumose
and not serrulate.
SOME NOTES ON THE GENUS HERMANNIA, WITH DESCRIPTIONS OF FIVE NEW SPECIES.
By Carrarn (S) T. M. Satter, R.N. (Ret.).
The large genus Hermannia, in which it is now generally considered that Mahernia should be embodied, is, at least as far as the South African species are concerned, greatly in need of revision, but the task is one which is not to be undertaken lightly.
In the preparation of the present paper, in which I have proposed five new species, all of which (except H. confusa) I have observed in the field, it has involved the scrutiny of a large number of scattered descrip- tions of species published since the rather doubtful account in Flora Capensis, Vol. I, now over eighty years old. I have been struck parti- cularly by the difficulties involved owing to the brevity of most of the descriptions and by the great importance in this genus of field notes such as the size, habit, habitat, colour of the flowers, etc., details which have been but rarely recorded with the specimens collected. The colour of many of the flowers soon changes or is lost in drying. As an instance of this, reference may be made to my notes, under H. scabricaulis, as regards the colour of the flowers in H. lacera (E. Mey. ex Harv.) Fourcade (Maherma scabra BE. & Z.). Dr. Fourcade, who knows H. lacera in the living state, has kindly given me the true colour “ yellow with a purple as opposed to the guesses of Ecklon and Zeyher
»
base, drying reddish “ violaceous ” and Harvey “reddish ”’. I have also alluded to Harvey’s citation of Drege 7307a here. This is Mahernia pilosula Harv., a species with reddish orange flowers, but it is possible that there has been some accidental interchange in the numbers Drege 7307 and 7307a. The difference in the colour of the flowers was of great importance to me in suggesting that more than one species had been cited under one name, apart from other less apparent structural differences.
The revelations which have come to light in the case of Mahernia scabra are for the most part the result of some knowledge of the plants in the living state. It may be only a coincidence, but it seems open to question whether some of the other species would bear critical examina- tion. It is but rarely, however, that so many of the old cited specimens are represented in South Africa. Nearly all those which I have been able to examine are in a terrible state of preservation,
96 The Journal of South African Botany.
Again, to illustrate the importance of habitat notes, H. ternifolia Presl., in which the leaves are normally cuneate, truncate or emarginate, has been observed sprouting from the roots on recently burnt ground with such abnormal leaves that its identity could only be determined with difficulty, a somewhat disconcerting fact in a genus in which the classification is at present greatly dependent on the leaves.
Many of the species, especially in the § Mahernia, are prostrate or decumbent in habit, but it is not always easy to detect this from dried specimens.
It is therefore to be hoped that collectors will supply fuller particulars
of this‘nature with their specimens while the facts are fresh in their ~ memories.
More detailed descriptions are also necessary. Whereas a species may be distinguished from what is supposed to be its nearest affinity, it must be remembered that there are others still to come which may differ in some character not even mentioned in the existing description.
Even in the Bolus Herbarium alone, there are very large numbers of unidentified species and, with the improved facilities of modern transport, more and more are being added from year to year.
Hermannia scabricaulis Salter sp. nov. (Sterculiaceae-Mahernia) § Lacerifoliae.
H. lacerae (EK. Mey. ex Harv.) Fourcade (typus E. & Z. 398) affinis sed ita differt :—Partes herbaceae, nisi tamen rarissime folia, omnino sine pilis stellatis. Stipulae econnatae. Folia incondite crenato-dentata, vix lacerata et non pinnatifida. Calyx brevior, 5—6 mm. longus lobis attenuatis. Stamina breviora 5—6:5 mm. longa: filamentorum lobi e basi remotiores.
A procumbent woody shrublet, densely clothed with glandular- capitate hairs. Stipules free 4—6 mm. long, 3—4 mm. broad, obovate, acute variously toothed. Leaves petiolate, the blades elliptical, oval, ovate or obovate, mostly 2—4 cm. long, 1-5—2-5 em. broad, irregularly crenate-dentate, scarcely lacerate, usually glabrous above, minutely glandular-pilose beneath and on the margins, very rarely with a few scattered stellate hairs: petioles 0-5—1 cm. long. lowers dull pink. Peduncles twice or more longer than the leaves, like the pedicels and bracts glandular-pilose. Bract not or scarcely sheathing, 4—5 mm. long, about 4mm. broad, irregularly toothed. Calyx widely campanulate, 5—6 mm. long, glandular-pilose, the teeth ovate-lanceolate, attenuate, acute, rather longer than the tube. Petals about 1 em. long, the obovate limb 4 mm. long, rather longer than the pubescent claw. Stamens 5—6-5 mm. long: filaments cruciform, about 1:2 mm. broad across
Some Notes on the Genus Hermannia, with Descriptions of Five 97 New Species.
the lobes : lobes 3 mm. from the base, stellate-pilose : anthers 3—4 mm. long. Ovary oval, densely pilose with glandular and stellate hairs : style sparsely pubescent on the lower half. Capsule oval, sharply ridged, about 1-3 cm. long.
Hab. Cape Province: Cape Div.; Blaauwberg, Zeyher 130 (type in Bolus Herbarium); Kuils River, Zeyher; Tygerberg Pillans 4762 : Cape Peninsula ; near Muizenberg, Bolus 4942, Wolley- Dod 3686, Marloth 573; near Retreat, Pillans 4008; Maitland, Guthrie 603; Clovelly, Michell ; Chapmans Bay, Adamson 2259; west of Vasco da Gama Peak, Salter 6218: Caledon Div. ; Houw Hoek, Guthrie 4688: Rivers- dale Div.; Eland’s Kop, Mwir 1776.
An, affinity of H. lacera (EK. Mey. ex Harv.) Fourcade (Mahernia scabra Ki. & Z. of the Flora Capensis Vol. I, p. 214, excluding Zeyher, Blaauwberg, a different species which is here taken as the type of H. scabricaulis, and also Drege 7307a). H. scabricaulis differs from H. lacera as follows :—Herbaceous parts, except very rarely the leaves, entirely without stellate hairs. Stipules free. Leaves much less deeply divided, the petioles shorter, up to 1 cm. long. Calyx shorter, 5—6 mm. long, the teeth attenuate. Stamens shorter, 5—6-5 mm. long: lobes of the filaments more distant from the base.
Harvey bas confused at least two, or perhaps three distinct species under Mahernia scabra EK. & Z. in Flora Capensis. His description appears to apply in part to Zeyher’s Blaauwherg specimen which is not the same as H. & Z. 398, his type, but probably Drege, from Zondag’s River (MM. lacera E. Mey.) should be considered as the type under the new combination H. lacera Fourcade. Drége 7307a, from Paarl! district, as represented in the National Herb. Pretoria (ex Herb. Mus. Paris) with a ticket number in Drége’s hand-writing, is identical with the type of Mahernia pilosula Harv., but Harvey cites Drege 7307, not 7307a under that species. Possibly the numbers have been wrongly affixed on some Herbarium sheets. WV. pilosula is a species with reddish-orange flowers. There seems to be a discrepancy in Harvey's description, for neither in the type (H. & Z. 407) nor in Drége or Pappe, nor in the many specimens of later collecting, are the bracts ‘“‘ connate into an incised hood ”’, being, rather, 3-partite almost to the extreme base with the medial lobe forked and the two lateral narrow-linear.
E. & Z. 398 as originally described in Enum. Plant I (1835) is said to have “ violaceous” flowers. I am indebted to Dr. H. G. Fourcade for the information that the flowers of his specimens (F. 4675) were “yellow with a purple base’, but they become reddish in drying, the state in which they were seen by Harvey.
98 The Journal of South African Botany.
Hermannia dichroma Salter sp. nov. (Sterculiaceae-Mahernia). § Verticillatae.
H. humifusae E. & Z. affinis sed ita differt :—Vfolia elongatiora, in- condite crenato-lobulata, minus lacerata. Pedunculi breviores. Bracteae valde minores, fere 4-5 mm. longae, basi vaginantes. Petala lacti- coloria vel rarius obscure rosea: lamina latior, fere orbicularis, quam unguiculus 2-5-plo longior. Filamenta ad lobos 1-2 mm. lata, lobis basi propinquioribus. Ovarium latiore, subglobosum.
A widely diffuse or decumbent shrublet, the stems up to 40 cm. ‘long, densely and shortly. glandular-scabrid : flowering branchlets with a few admixed stellate hairs. Leaves oblong or oblong-elliptical, rarely obovate, obtuse, irregularly crenate-lobulate, scarcely lacerate, mostly 2—5 cm. long, 0-5—1-8 cm. broad, with rather sparse stellate hairs on both faces: petioles glandular-scabrid, 2—8 mm. long. Stipules free, lanceolate or ovate-lanceolate, entire, 3—7 mm. long. Flowers rather numerous, pale cream-coloured or more rarely dull pink, very sweet-scented. Peduncles 2-flowered, as long as or longer than the leaves, like the pedicels glandular-pubescent with admixed stellate hairs : bracts connate in a laxly sheathing involucre about 4:5 mm. long, with 2 short ovate-deltoid teeth at the apex. Calyx widely campanulate, 5—5:5 mm. long, stellate-pilose, the teeth ovate-deltoid, attenuate, ciliate, about 2-5 mm. long. Corolla campanulate: petals 8—9 mm. long, the limb almost orbicular, stellate-pilose near the base inside, 23 times as long as the concave claw. Stamens 6 mm. long: filaments cruciform 2 mm. above the base, 1-2 mm. broad across the lobes: lobes greenish, densely stellate-pilose. Ovary subglobose, about 2-5 mm. long, densely and minutely stellate-pilose: style 4:5—5 mm. long, pubescent near the base.
Hab. Cape Province: Clanwilliam Diy. ; 5 miles west of Clanwilliam, Salter 2745 (type in Bolus Herbarium) ; Graafwater, Compton 4300 (dull pink); Nardouw Pass, Salter 3620, 7534. Flowers September. Co- types were sent to the British Museum and to Kew where it could not be identified.
An, affinity of H. humifusa E. & Z., but differing as follows :—Leaves more elongate and less lacerate. Peduncles shorter. Bracts very much smaller, sheathing. Petals cream-coloured or more rarely dull pink, the limb much longer than the claw. Filaments narrower across the lobes, the lobes nearer the base. Ovary more globose. It differs from H. chloroleuca Diels (ex desc.) in the colour of the flowers, not white suffused with green, in the much smaller floral parts and in the calyx, which bears stellate hairs. It is superficially like H. collina Schltr., but the upper part of the filaments (above the lobes) is not sigmoid as in that species.
Some Notes on the Genus Hermannia, with Descriptions of Five 99 New Species.
Hermannia oligantha Salter sp. nov. (Sterculiaceae-Eu-Hermannia) § Cuneifoliae.
H. alnifoliae LL. affinis sed ita differt : pilis stellatis densis minutis et pilis stellatis laxis majoribus et pilis capitatis adjectis. Folia late ovata vel obovata vel fere orbicularia, nequaquam basi cuneata. Flores majores: racemi eramosi pauciflori. Calyx 5—6 mm. longus: lobi apicibus longioribus subulatis instructi. Petala fere 8 mm. longa, laminis 3 mm. latis. Stamina 4-5 mm. longa :
Partes herbaceae juniores
filamenta apicem versus 1-5 mm. lata ; antherae 2mm. longae. Ovariwm sicut stilus longius.
A small much-branched diffuse woody shrublet about 30 cm. high. Branches glandular-scabrid, the younger herbaceous parts densely pilose with very minute stellate hairs, glandular-capitate hairs and with some large weak white tubercular-based stellate hairs admixed. Stipules usually attenuate from an oblique gibbous base, acute, sometimes falcate, 2-5—6 mm. long. Leaves subsessile or shortly petiolate, broadly ovate, obovate or nearly orbicular, 4—9 mm. long, 3-5—7 mm. broad, grey, deeply wrinkled, pilose with the three kinds of hairs, but the glandular- capitate few and inconspicuous: petioles up to 3 mm. long. Flowers few, yellow-orange, 1—2, rarely more, in simple unbranched terminal racemes. Bracts tritoliolate, the medial lobe linear, the two lateral like the stipules but smaller. Peduncles 2—3 mm. long. Calyx sub- membranous, 5—6 1m. long, the teeth deltoid at the base, narrowly attenuate to a subulate apex, rather shorter than the tube. Corolla urn-shaped : petals about 8 mm. long, the limb obovate, as long as the claw. Stamens 4-5 mm. long: filaments cuneate, 1-5 mm. broad at the apex, broadly membranous at the margins: anthers 2 mm. long. Ovary 2 mm. long, densely hairy: style 3 mm. long, puberulous on the lower half. Capsule subglobular, about 5 mm. in diam., with three obtuse apical horns.
Hab. Cape Province: Clanwilliam Division ; 5 miles east of Graaf- water, Salter 2750 (type in Bolus Herbarium); top of Nardouw Pass, Pillans 7093 ; Bull Hoek, Schlechter 8384 (a fragment) : Namaqualand ; Kamieskroon, Compton 11106: also identified at Kew as being the same as Pearson 6275. Flowers July—Sept. Co-types were sent to the British Museum and Kew.
An. affinity of the profuse flowering H. alnifolia L., but differing in. having gland-tipped hairs in addition to the two kinds of stellate hairs, simple few-flowered racemes, the flowers considerably larger, and longer subulate tips on the calyx-teeth.
Hermannia confusa Salter sp. nov. (Sterculaceae-HKu-Hermannia) § Pinnatifidae.
100 The Journal of South African Botany.
Fruticulus erectus, 20—40 cm. altus. Caulis rigidus, inferne lignosus. ~ Rami graciles, plus minusve stellato-pilosi pilis glandulosis minutis admixtis. Stipulae liberae, oblique ovatae vel lanceolatae, valde attenu- atae, auricula gibbosa obtusa induta. Folia petiolata, pinnatisecta vel bi-pinnatisecta, supra glabra, infra stellato-pilosa: segmenta ultima anguste linearia, apice saepe seta sola, plerumque setis paucis apicem versus induta: petioli ad 1 cm. longi. IJnflorescentiae longae, graciles. Pedunculi satis remoti, plerumque 2-flori, 1—3 cm. longi. Bracteae minutae, tripartitae. Calyx 5—6 mm. longus: lobi 2—2-5 mm. longi et lati, e basi deltoidea valde attenuati, ad apicem cuspidati, sinibus rotundatis. Corolla flava, fere 9 mm. longa: petalorum laminae late ovales vel ovatae, auricula gibbosa obtusa marginis exterioris basi in- structae : ungues apicem versus ciliati. Filamenta cuneata, 1-6—2 mm. lata, apice sparse minuteque stellato-pilosa: antherae in ambitu lanceolatae. Ovariwm obovatum. ad angulos et apicem versus dense stellato-pilosum.
An, erect shrublet, 20—40 cm. high, the stem woody below. Branches slender, more or less stellato-pilose with minute gland-tipped hairs admixed. Stipules free, obliquely ovate to lanceolate, 3—5 mm. long, strongly attenuate, the gibbose auricle obtuse. Leaves petiolate, pinnati- sect or bi-pinnatisect, mostly 1—3 cm. long, glabrous above, minutely glandular-pilose, stellato-pilose beneath, the ultimate segments narrow- linear, 0:5—1 mm. broad, somewhat decurrent, acute or obtuse, often with one terminal bristle, sometimes with a few more simple bristles near the apex: petioles up to 1 em. long. Inflorescences long, slender, usually minutely glandular, with or without stellate hairs. Pedwncles rather distant, mostly 2-flowered, 1—3 cm. long, shorter than, the inter- nodes. SBracts very small, tripartite. Calyx widely campanulate, 5—6 mm. long, the tube glabrous or rarely sparsely stellato-pilose : lobes 2—2-5 mm. long and broad, strongly attenuate from a deltoid base to an acute cuspidate, rarely stellato-ciliate apex, the sinus rounded. Corolla yellow, urceolate : petals imbricate, about 9 mm. long, the limb broadly oval or ovate, slightly shorter than the claw, with a distinct gibbous auricle at the base of the outer margin: claw arcuate, convex, ciliate on the upper part. Filaments cuneate, 4 mm. long, 1-6—2 mm. broad, with membranous wings and with a few small stellate hairs at the truncate apex : anthers lanceolate in outline, 3mm. long. Ovary obovate, densely stellato-pilose on the ridges and towards the apex: style pubescent on the lower half. Capsule sub-globose, indented at the apex, about 5 mm. in diam., the stellate hairs on the ridges towards the apex largest.
Hab. Cape Province : Caledon Div. ; between Caledon and Babylon’s Tower, LH. & Z. 374 (in Bolus Herb.) ; Hartebeeste River, Zeyher 2001 ;
Some Notes on the Genus Hermannia, with Descriptions of Five 101 New Species.
hills near Caledon and Somerset, H. & Z.; Zwartberg, Schlechter 5557, Guthrie (Bolus Herb. 17034) ; Houw Hoek, Guthrie 3377, Bolus ; between Caledon, and Napier, Hsterhuysen 3044: Swellendam Div. ; Bushmans River, Compton 11909: Bredasdorp Div.; Napier, Compton 8978 : Stellenbosch Div.; near Somerset, Hottentots Holland, H. & Z. 408 (Mahernia pinnata Ki. & Z.) ; Sir Lowry Pass, Wasserfall 383 : Worcester Div. ; Hex River, Drége ; near De Doorns, Bolus 13076 ; near Worcester, Leipoldt 3509, (Bolus Herb. 18622): Ceres Div.; Prince Alfred’s Hamlet, Compton 11969; Gydouw Pass, Leipoldt 3508, Bolus: Tulbagh Div. ; near Waterfall, H. & Z. 377 (H. coronopifolia EK. & Z.), Hutchinson 371 ; Tulbagh Road, Schlechter 8995; near Tulbagh Kloof, Bolus ; Winter- hoeksberg, Bolus 5127; Artois, Hsterhwysen 6083: Malmesbury Div. ; near Moorreesburg, Bolus 9951: Piquetberg Div.; Piquetberg, Bolus 13524; Grooenvlei, Guthrie 2551; Piquenier’s Kloof, Schlechter 4940 ; Kapitein’s Kloof, Pillans 8027.
This species comprises the specimens cited by Harvey in Flora Capensis Vol. I, p. 203 under Hermannia tenuifolia Sims in Bot. Mag. t. 1348, viz. H. & GZ. 374, H. & Z. 377 (H. coronopifolia HE. & Z.), H. & Z. 408 (Mahernia pinnata E. & Z.), H. & Z. hills near Caledon and Somerset, Zeyher 2001 and Drége, Hex River, together with a large number of conspecific specimens of later collecting. I have not seen Drége, between Breede River and Bokkeveld, but in all probability it is also conspecific.
My reasons for removing these specimens from H. tenwifolia are as follows :—
Although the figure (Bot. Mag. t. 1348) with its 3-word description— foliis pinnatifido-liniaribus—has been considered as validly published, it seems advisable to treat it as an imperfectly known species, for the figure might just as well be intended to represent H. pinnatisecta Salter sp. nov. (a species unknown to Harvey) which follows here and ha’ similar leaves. It is in reality no more than a “ pretty ” picture and it is impossible to determine the shape of the petals and, indeed, whether the margins are intended to be sinuate or merely undulate : the calyx is quite characterless and, of course, the sexual parts of the flower are hidden. I do not think the exact identity of Sims’ species can ever be known with certainty. I have not seen the defective specimen in Herb. Banks which Sims states to be “ like it”.
But a further difficulty arises. Owing to the inadequacy of the original description, Harvey must have obtained the details given in the Flora Capensis elsewhere, and it would be assumed from one or more of the specimens which he cites. The five referred to above are certainly conspecific inter se, but it has been found that his description differs
102 The Journal of South African Botany.
in many important respects from my own observation of H. & Z. 374 and other undoubtedly conspecific specimens. The stipules are re- markably gibbous at the base and strongly attenuate, not amplexicaul ; the leaves are petiolate, not subsessile ; the bracts are tripartite and there is no sign of a terminal bristle on the calyx lobes, though a few specimens have one or two stellate hairs ; the limb of the petal, described as ovate, is, in fact, asymmetrical having a basal auricle, nor is the claw downy ; the filaments are almost too narrow to be described as obovate, certainly not broadly obovate and the anthers are lanceolate in outline, not hastate. The character of the stiff hair at the apex of the leaf seg- ments appears to be of little value, for these hairs seem to be caducous, some specimens, notably Zeyher 2001, being entirely without them, while in others there are several simple hairs close to the apex.
I am unable to explain these discrepancies and have therefore given a full description of this species as H. confusa, for there appears to be no point in attaching my description to a species of such uncertain identity as H. tenucfolia.
Hermannia pinnatisecta Salter sp. nov. (Sterculaceae-Ku-Hermannia) § Pinnatifidae.
Fruticulus decumbens. Caules numerosi, non, rigidi, saepe eramulosi, inferne glabrescentes, superne minute stellato-pilosi. Stipulae ovatae, attenuate acutae, 2—4 mm. longae. Folia glabrescentia, juniora pilis paucis stellatis induta, superiora remota, petiolo incluso 1-5—3 cm. longa, 0-6—2 cm. lata, profunde pinnatisecta vel bi-pinnatisecta : segmenta ultima linearia, basin versus coarctata, decurrentia. Flores pauci, flavi, racemosi. Bracteae 2—3-partitae, amplectentes, (2am: longae. Pedunculi internodis valde breviores, stellato-pilosi. Calyx 7 mm. longus, glanduloso-punctatus, laxe stellato-pilosus, lobis deltoideis leviter acuminatis, sinibus subangulosis. Corolla urceolata, 1 cm. longa : petalorum laminae obovatae, raro obscure auriculatae, 4—5 mm. latae. Filamenta cuneata, 1-8 mm. lata : antherae in ambitu lanceolat ae, acutae, 3mm. longae. Ovariuwm ovale, saltem ad angulos stellato-pilosum.
A low decumbent shrublet. Stems numerous, usually simple, scarcely woody, 20—50 env. long, usually branching from the root or below ground, glabrescent towards the base, minutely stellato-pilose above. Stipules ovate, attenuately acute, 2—4 mm. long. Leaves glabrescent, the younger with a few small stellate hairs, solitary, the upper widely distant, including the petiole usually 1-5—3 cm. long, 0-6—2 cm. broad, deeply pinnatisect or bi-pinnatisect, the ultimate lobes linear, obtuse or subacute, tapering towards the base, somewhat decurrent, usually 1—2 mm. broad: petioles 0-3—1l cm. long. Flowers few, racemose, distant,
Some Notes on the Genus Hermannia, with Descriptions of Five 103 New Species.
solitary, yellow. bracts 2—3-partite, clasping, 1—2 mm. long. Peduncles much shorter than the internodes, stellato-pilose. Calyx widely cam- panulate, about 7 mm. long, gland-dotted, sometimes laxly stellato- pilose, the teeth 2°5 mm. long, deltoid, slightly acuminate, the sinus subangular. Corolla urceolate: petals 1 cm. long, the limb sub-sym- metrical or rarely with an obscure auricle at the base of the outer margin, 4—5 mm. broad, as long as the arcuate claw. Filaments cuneate or narrowly cuneate-obovate, 3:5 mm. long, 1-8 mm. broad, with mem- branous wings: anthers lanceolate in outline, acute, 3 mm. long. Ovary oval, 2 mm. long, densely stellato-pilose, at least on the ridges: style 4-5 mm. long, very minutely pubescent. Capsule 6—7 mm. in diam., indented at the apex, stellato-pilose on the ridges.
Hab. Cape Province: Malmesbury Div.; Mamre Hills, Compton 14928 (type at Kirstenbosch); near Yserfontein, Salter 1347; Koper- fontein, Compton 9457; Darling, Schlechter 5339; Slangkop, M. Gillett Bol. Herb. 23340; Moorreesberg, Leipoldt Bol. Herb. 18628: Clan- william Div.; Van Putten’s Vlei, WM. Gillett 4058. Flowers Aug.—Sept.
Var. 8 auriculata Salter. Calyx sinibus inter lobos rotundatioribus. Flores flavi vel rarius albi. Petalorum laminae angustiores, 5—6 mm. latae, auricula gibbosa obtusa conspicua marginis exterioris basi instructae.
Sinus between the calyx lobes more rounded. lowers yellow or more rarely white. Limb of the petals wider, 5—6 mm. broad, with a conspicuous gibbous auricle at the base of the outer margin.
Hab. Cape Province: Malmesbury Div.; Yserfontein, Compton 17374 (type at Kirstenbosch); near Langebaan, Leighton Bol. Herb. 23339 ; Steenbergs Cove, Compton 15942 (corolla white).
H. pinnatisecta, judging by the leaves, is an affinity of H. confusa Salter, the preceding species and the imperfectly known H. tenwifolia Sims in Bot. Mag. It differs in its lax decumbent habit, symmetrical stipules, solitary flowers, larger corolla and calyx, the latter sparsely stellato-pilose with much less attenuate teeth and the entirely glabrous filaments.
A NOTE ON URGINEA PYGMAEA DUTHIE. By R. 8. Apbamson.
This inconspicuous late-flowering species has recently been gathered by Captain T. M. Salter in a new locality, Rondebosch Common, where it was growing on recently burned ground. As it was previously known only from the original locality on the Stellenbosch Flats, this discovery marks an, extension of the range and an, addition to the flora of the Cape Peninsula.
The Rondebosch specimens have been compared with the type speci- men, (Duthie 1603 in Herb. Univ. Stellenb.) and are certainly identical. The type sheet consists of a number of specimens at different stages of development. An examination of the specimens from the Rondebosch locality has revealed, however, certain, features which were either not mentioned in or which do not show real agreement with the original description and figures (Duthie, Ann. Univ. Stellenb. VI. A. 2, 10, 1928, Pl. 2). As the species is little known and as there seems a probability of its discovery in further stations, these points seemed worthy of record.
The Rondebosch specimens have 3—6 leaves per bulb not only 1—3 as described. The leaves are dark green but not glaucous and when fully developed are horizontal or prostrate on the ground. The specimen figured (l.c. fig. 2) which has erect leaves would appear to be either immature or to have been grown, under abnormal conditions.
In the notes following the description, it is stated that 1-flowered scapes are commonest, 2-flowered occasional, and 3-flowered rare. These specimens have most commonly 2 or 3 flowers, some have 4, while 1-flowered specimens seem to be without doubt poorly developed in- dividuals. On the type sheet the flowering specimens are mostly 1-flowered but the fruiting specimens are just like those from Rondebosch.
The perianth does not quite agree with the figure (l.c. fig. 10): the segments are longer in, proportion to their width than is. shown, and are united for not more than a third their length. The stamens are attached at or above the level of union of the segments. In colour each perianth segment has a broad stripe, chestnut-brown, inside, dull chocolate-brown, outside, with the marginal part translucent and very thin. These brown flowers are a very characteristic feature and assist in rendering the plant inconspicuous but were not mentioned in the description. Indeed, so
106 The Journal of South African Botany.
inconspicuous are the flowers that their discovery on Rondebosch Common was due to observation of the bright yellow anthers not the perianth.
The seeds figured (1.c. fig. 7) are not entirely characteristic. Young seeds are tetrahedral with a very loose testa. As maturity is reached the seeds become flattened, those at the outside of the capsule becoming quite flat, those in the centre of the capsule remaining somewhat angular like those figured.
The leaves form immediately after flowering time and appear to be full grown in May or June.
U. pygmaea is most nearly allied to U. minor and might be confused with that species especially in the dried state (cf. Duthie 1.c. 12: Adam- son, this Journal VIII, 242, 1942). It is distinguished by the horizontal leaves which are somewhat widened towards the tip, the short pedicels and by the brown flowers with united perianth segments. U. minor has erect filiform leaves, longer pedicels with the flowers in a small corymb, and perianth segments almost free and white inside.
ROELLA RETICULATA. By R. S. Apamson.
Highty years ago Sonder (FI. Cap. 3, 592, 1865) stated that the rather common, plant generally known, as R. reticulata A.DC. was not the same as the plant originally described by Linnaeus under that name. As the matter has not been taken further and as the statement if true means that the generally used name is not valid, it seemed advisable to in- vestigate the problem and attempt to reach a settlement. In what follows, very considerable help from Dr. T. A. Sprague is gratefully acknowledged.
Roella reticulata was first described by Linnaeus (Sp. Pl. ed. 1, 170, 1753) as follows :—
Roella foliis ciliatis ; mucrone reflexo. Roella foliis imbricatis. Roy. lugdb. 248. Campanula capitis bonae spei, foliis reti- culatis spinosis. Pet. mus. 21, f. 157. The description was repeated without change in the second edition (Sp. Pl. ed. 2, 241, 1764).
As there is, according to Jackson’s index, no contemporary specimen of the plant in the Linnaean Herbarium, it is not possible to ascertain directly what the plant intended really is and recourse must be had to the attached references. The figure given by Petiver for his plant is of a shoot without flowers which though quite unlike any species of Roella is a fair representations of Gorteria cilhkaris L. (Cullumia ciliaris (L.) R.Br.). This identification is supported by the fact that Burmann (Pl. Afr. Rar. 152) quoted Petiver’s plant as being the same as his Carlina foliis imbricatis oblongis reticulatis et in aculeam aduncum desinentibus, which is itself quoted as a reference by Linnaeus under Gorteria ciliaris (Sp. Pl. ed. 2, 1284, 1764). ;
It has not been possible to check the reference to Van Royen. How- ever, Richter (C. Linn. Syst., etce., 181, 1840) says that the other refer- ences refer to Gorteria ciliaris and that R. reticulata is probably that plant. Linnaeus’ own diagnosis is certainly more applicable to Gorteria ciliaris than to any species of Roella. The fact that in 1764 Linnaeus described Gorteria ciliaris but made no alteration in the species of Roella suggests that he was not familiar with the plant be had described in 1753. It would seem that he described it wholly from the literature and that the
108 The Journal of South African Botany.
figure of Petiver is the type. From Petiver’s description he obtained the epithet “ reticulata’, and the position he assigned to the plant may well have been influenced by Petiver’s unexplained use of the name Campanula.
If then, Roella reticulata L. is really Gorteria ciliaris, the plant described by A. De Candolle in 1830 (Monog. Campan. 174) cannot bear that name though it is generally applied to it. When describing his species, De Candolle says that Petiver’s figure, which is referred to by Linnaeus and which appears to be the type, represents a different plant. He justifies his use of the Linnaean name on account of a specimen (in Herb. DC.) which is stated to have been given to Vahl by Linnaeus himself. Though this specimen has not been seen, it cannot possibly be the type of Roella reticulata LL. and cannot affect the argument that that name applies to Gorteria ciliaris. The further references that are given by De Candolle, to Lamarck and to Poiret, do not afford any help. The former (III, 2, 8, 1793) merely expands the original diagnosis by a few words and adds “ Gorteriae ciliaris habitus’. The latter (Enc. Meth. 6, 232, 1804) states that the species is not a Roella.
It is thus apparent that FR. reticulata L. and R. reticulata A.DC. are different plants and that the latter must have another name. Three names seem available which were used for species closely allied to or even, synonymous with R. reticulata A.DC. These are :—R. Dunantu A.DC. Monog. Campan. 175, 1830; R. gracilis Eck. & Zeyh. Enum. 387, 1837; and R. prostrata KH. Mey. ex A.DC. in DC. Prodr. 7, 447, 1838.
Of these Rk. Dunanti is sufficiently distinct to retain independant specific rank. . gracilis is a composite species, in part R. ciliata L. and in part KR. reticulata A.DC., and so may be disregarded. The third species, R. prostrata, though the type specimen has not been seen, is evidently from the description closely allied to R. reticulata A.DC.
An, examination, of a number of specimens, both in the field and in collections, has shown, that there is a considerable range of variation in habit, in the size and closeness of the leaves, in the presence or absence of teeth on the sepals, in the length and colour of the corolla, and in other lesser features. Indeed, there is a complex group of individuals, not sharply delimited, from which De Candolle selected two forms for descrip- tion out of a much larger possible number. Whatever selection was made there appears to be overlapping in features and, until a full study of all the forms, both genetical and other, is undertaken, the only practic- able course seems to be to group all of them under one species. Sonder (1.c.) seemed to realise this when he reduced R. prostrata to a variety. Unfortunately the characters he gives for his yariety are not satisfactory.
Roella reticulata. 109
Shortly before the issue of the volume of the Prodromus in which De Candolle described his species, the name #. prostrata was published by Ecklon and Zeyher (Enum. 387, 1837) but without any description. The specimens to which they attached this name are not of De Candolle’s plant but are Rk. spicata L.f. De Candolle described his species quite independently and without previous knowledge of the use of the name. This is shown by the fact that the new species described in Ecklon and Zeyher’s Enumeration are in the Prodromus only included in an appendix : the main body of the work was written before the other book was available.
Since it appears that R. prostrata A.DC. is synonymous with R. reticulata A.DC., and since Ecklon and Zeyher’s use of the name is without description and hence without nomenclatorial standing, it is suggested that De Candolle’s name be applied to the species. The synonymy is :—
R. prostrata E. Mey. ex A.DC. in DC. Prodr. 7, 447, 1838, non E. & Z. Enum. 387, 1837 (nomen nudum).
R. reticulata A.DC. Monog. Campan. 174, 1830, non L. Sp. Pl. 170, 1758.
R. gracilis HK. & Z. Enum. 387, 1837 (p. pte.).
THE SOUTH AFRICAN QUININE TREE. By HE. P. PHuitxtes.
In the year 1926, the National Herbarium forwarded a supply of dry bark of the “ Quinine tree’ to the Johns Hopkins University, U.S.A.* The material was collected from trees growing in Pondoland and was identified as Rowvolfia caffra Sond. (National Herbarium No. 3358.) At the time the question arose as to whether two species of Rowvolfia occurred in South Africa and the specimen, quoted above was submitted some years later to Kew for an opinion. The Director of Kew in his reply suggested a series of specimens should be examined critically and those from the Magaliesberg—the type locality of R. caffra—compared with specimens from Natal (the type locality of R. natalensis) and Pondoland.
The writer recently had an opportunity of undertaking such an examination. The calyx may vary in length from 1-5 mm. to 2-5 mm. and the tube from saucer-shaped to shortly campanulate ; the corolla varies in length from 4-5 mm. to 6-5 mm. but otherwise does not differ in the specimens examined ; the stamens are similar in all the specimens ; the dise surrounding the ovary may be saucer-shaped and not partly enveloping the ovary or may be cup-shaped and envelope the lower portion of the ovary ; the ovary wall may be smooth or wrinkled ; the style differs slightly in length but the stigma is similar in all the specimens.
The difference in leaf character given in the Key in the “ Flora Capensis ”’ to distinguish the two species does not hold good. Specimens from Natal have lanceolate leaves, while in a Burke specimen (No. 113) one leaf on the specimen is definitely oblanceolate, and a specimen trom the Magaliesberg (Pole-Hvans No. 1522) has both lanceolate and oblanceo- late leaves on the same specimen. The result of the examination of a larger range of material than was available when the descriptions in the “ Flora Capensis ” were drawn up, indicates that there is no vegetative or floral characters which would justify us recognising two species of Rouwvolfia in South Africa. This view is copfirmed by Dr. I. J. Craib of the Division of Forestry, who informed the writer that foresters only recognise one species. The distribution of the species is quite normal. As the specific epithet ‘caffra’? appeared first in Sonder’s work (Linnaea, vol. 23) the name for the South African Quinine tree should be Rowvolfia caffra Sond. (= R. natalensis Sond.).
* Soa American Chemical Society, vol. 54, p. 2412 (1982) for results of examinatio 1
THE GENERA CHILIANTHUS BURCH. AND BUDDLEJA LINN.
By E. P. PHIttiies.
In the “Flora Capensis ” the two genera Chilianthus and Buddleja are upheld for a group of South African plants. The distinguishing characters as given in the Key are :—
Chilianthus—Calyx deeply lobed ; anthers exserted. Buddleja —Calyx shortly lobed ; anthers included.
The species placed under the two genera may be distinguished, however, by a more easily recognised character, viz., a campanulate corolla-tube in Chilianthus and a cylindric corolla-tube in Buddleja.
In De Candolle’s Prodromus the genera are distinguished as follows :—
Chilianthus—Calyx 4-fid. Corolla-tube short. Stamens more or less exserted.
Buddleja —Calyx 4-toothed or 4-fid. Corolla-tube short or long. Stamens included.
In the “ Genera Plantarum ”’ it is stated under the genus Chilianthus —strongly allied to the genus Buddleja and differs principally on the exserted anthers. Solereder (Pflanzenfamilien)