f $i if f "4 ov - i | ? ® ‘ , i ’ ' 7 : Ls I i - 7 * P : { i on : iJ . a i 2 a j . 7 1 i = ; i : « a } f a ar is ' | = : *. “ ty : i j i» hn, : ‘ t uv ; ' Fi , A ' t ‘ i ~ : ' e A = - 7 ra ’ = ' : ; = ‘ . 5 . 7 2 - ) 2 oe bn a . ~~ ~ “ - : r r/) = : = : a =) J 7 z 7 ; : . - +5 t bs ' : 7 a = oF. aa 7 = 7 : ; , a i A | = - - fd . _ , : ’ 7 ws - = S | 1 t _ - o \ = ul w 7 , 2 - ‘ i ; os 5 ae Sy re - - - e = = - : - = ‘ > A 4 : 7 4 ; a : nh > ole. : neers Ps 5 7 = = 7 be 7 am = . = “ 2 « = = = es - ‘ : 7 ~ : hs ‘ . / = - 7 — - . ‘ . 7 a f = us 7 rs > ian : é ' _ - i _ 7 - = 1 7 =) 7 ? i ‘ yi @ = 7 9 ; BI if hat 7 ie ‘ a : “7 i ; i. a. Ie Vt + . i a: iu hy ; qi F a ry) it i , 5 v ee ° ; >> ~ : - —_ - : i = 7 * - : es + 1a sh 7 ’ 1 7 ae q 5 2, 4 1 “J see i ca 5 yf i 5 . 4 i ies < ‘ ‘ ’ rl . ' i y . + =. = = ’ a > “ vie - i is = i i : 7 7 : 1 i } i - & i 7 ‘ ; a - : - i ’ 7 3 7 a * 7 a i a { me | 7 i 7 : . + e j ak : Ps , 1 - os A [ - - ~- i : , ic : - ' f ocr ‘7 u i uy ‘ 4 i ‘ F ig a ¢ 7 i a - 4 A + « : ' 7 . — ' = - 7 + ce ’ k ' - ; / ' iz ' : ; * = 7 : : * 7 o— ~ “5 1 5 - f i © : 7 7 = ' . ae i ’ a , 1 ; _ - > : , : a y . , . i _ b 7 * ' S . J i : - + 1 : a 0 : : Y a ‘ i . a 7 f se a ) 2 : i. of $ - : ’ 1 7 ‘ pa ' ne a * ft : 7 ; 7 Card ~~ . i i = t Y } 2 ] ’ _ . bd wag : ‘ 7 ¥ a % : a ; _ 7 r Ma t i . a = ; ‘ t om na wo we: 1-.~ af ; oe tea i ’ : ee ee a ANNALS OF THE SOUTH AFRICAN MUSEUM VOLUME XVI ANNALS OF THE SOUTH AFRICAN MUSEUM VOLE VE eX PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY NEILL AND CO., LTD., 212 CAUSEWAYSIDE, EDINBURGH. 1917 — 1933. ts 1 T' : ; e ° = - 7. > —- » a 7 : » ~ Fi - > an 7 , 2 = - 7 _ -_ —_ 7 uy ' - ae a - Dé an ’ - - ; - J - = : 7 : - 7] a 7 ¢ | a P= we . . ) a yayl a e om ~ e ; 7 es =. - ax =) ; » a - - = a ; - 7 : § 7 ‘a - - oe 7 . ; ; Ss : ; 7 _— s 7 ¥ a , © - Z : oo a! a 7 , i a p a < a r be 7 s. ‘ e an _ ’ P A Thi - f z roi ee oe y 7 J poi SDE 7 I | ‘ - . : _ Poe a : - 7 - / J LS - > 7 - a Fe : 7 Ss o of i 1 7 7 a : 7 : 7 7 A 7 = Th .) : . - o i 7 : 7 _ ¥ o _ 7 : rg - - * 7 7 f te 7 : 7 « a : - 7 -5 ae a 7 7 = , pil ’ y - eer y 7 <. = y - i @. hh _ >» : : - 7 4 ; y ie Se : J 7 eal : y : a in =o ’ 7 ie a 7 7 ; Mi - ae : 7 7 s i 7 : Pe —— . 7 i 7 i - 7 7 - = a 7 _ > & us a - a - bee : - ary - a , - - i = a - a - Oy i 7 ~ = - - 7 » e@ _ 7 7 7 7 a | i — : a as y . wet - Po & j = 2 ny = i 7 ~s oa : ye B © - t er = er); = _ 7 - “a 7 i : er aS 7 — - - ‘ ‘ - 7 cf. . 7 ¥ - - = a@ 7 iy oF . = . a oe y 7 4 a on = 4 : : 7 : ; A _ 7 re oO 7 fi a) : 7 a _ _~* ? - 2 _ a re a 7 ; » a - - o et ‘ . = - - _ uF . 7 < , ee ; A nl ® 7 7 =] i : > , = - ae : — A , . ¥ = rn 7 7 r - a TV ng = J , _ —_ - : ~~ ’ : . : _ = . 7 ' —_ ; e vo ' ? § 7 7 7. & ; _ = 7 : ' wy : 7 : : : : 7 = md “"— > 7 : - i ae a ; a rn = wan) ao: ~ a a y : Oy fate : a “ ’ =) = 5 : a i — te 7 ~ id AL i. oe Ee ia =e ie ; 7 Ed a ro j a ones —— 7 -@ : 7 = a 7 _ “ 0 as i. arenas, 0° Asm 7 te ‘ A ee’ a - - — - u - a - a Det o . = _ - 7 es - ne 7 Ps - ‘ 7 = te” : P a 7 : i : « > . : a ‘ 7 _ 7 7 _ . - i : : - - _ 7 = q y - ie 7 p_' : ‘1; >'> o= 5 F < +> =—S 1) i 7 7 : a 7s Fae . ; ee 7 , oie 7 - 7 = ee ' Pree ———— TRUSTEES OF THE SOUTH AFRICAN MUSEUM. Sir Toomas Murr, C.M.G., M.A., LL.D., D.Sc., F.R.S. J. G. VAN DER Horst, Esq. Prof. Wiit1am ApaAm JOLLY, M.B., Ch.B., D.Sc., F.R.S.S.Afr. W. J. THORNE, Esq. W. Brinton, Esy. SCIENTIFIC STAFF OF THE SOUTH AFRICAN MUSEUM. Epwin Leonarp Gitt, D.Sc., Director. KepreL Harcourt BarnarpD, M.A., D.Sc., F.L.S., Assistant Director; in charge of Fish and Marine Invertebrates. Miss STAR GARABEDIAN, M.A., Assistant in charge of the Botanical Department. REGINALD FREDERICK LAWRENCE, B.A., Ph.D., Assistant in charge of Reptiles, Batrachians, and Arachnids. ALBERT JOHN HeEssb, B.Sc., Ph.D., Assistant in charge of the Entomological Department. LrguWE Dirk Boonstra, D.Sc., Assistant in Palaeontology. Stpney Henry Haueuron, B.A., D.Sc., F.G.S., Honorary Keeper of the Geological and Palaeontological Collections. Astitey J. H. Goopwin, M.A., Honorary Keeper of the Ethnological and Archaeo- logical Collections. LIST OF CONTRIBUTORS. S. GARABEDIAN. Contributions to a Knowledge of the Flora of South-West Africa. No. 1. List of Grasses. Plate VIII E: P. PxHiures. A Contribution to the Flora of the Leribe Plateau and Environs: with a Discussion on the Relationships of the Floras of Basutoland, the Kalahari, and the South-eastern Regions. Plates I-VII M. A. Pocock. Volvox and Associated Algae from Kimberley. With Field Notes by J. H. Power. Plates XXV-XXXVII, and Seven Text-figures M. A. Pocock. Volvox in South Africa. Plates XX XVIII-XLIX, and Ten Text-figures F. Ricu and M. A. Pocock. Observations on the Genus Volvox in Africa. Plates IX-X XIV, and Six Text-figures vii PAGE 381 473 523 - ” DATE OF ISSUE OF THE PARTS. Part 1, June 21st, 1917. Part 2, April, 1925. Part 3, July, 1933. Wile VII. VIII. TX.-XXIV. XXV.-XXIX. XXX. XXXI-XXXIII. XXXIV. XXXV. XXXVI-XXXVII. XXX VITI-XLIX. LIST OF PLATES. pe Lobelia Tysonii Phillips. Anthrixia fontana MacOwan. Blepharis espinosa Phillips. Utricularia humilis Phillips. { Lotononistrisegmentata Phillips. WahlenbergiadepressaW.& E. \ Lobelia aquatica Phillips. Euryops Annae Phillips. ( Heliophila basutica Phillips. Psammotropha androsacea Fenzl. | Wahlenbergia basutica Phillips. W. depressa W. & E. ( Helichrysum calocephalum Schltr. H. radnii Sp. Moore. \ Crepis polyodon Phillips. H. chionosphaerum D.C. Map of Basutoland. Map of South-West Africa. South African Volvoz. Volvox gigas. Volvox gigas and africanus. Volvox africanus. Volvox rousseletii forma griquaensis. Volvox rousseletii forma griquaensis and Kirchneriella africana. Sphaerocystis poweri. Volvox in South Africa. xi (An index of genera in Part 1 will be found on pp. 372-377.) Agrostis Amphilophis Andropogon Anthephora Aristida Arundo Brachiaria Bromus Chloris Cymbopogon Cynodon . Dactyloctenium Danthonia Dichantium Digitaria . Diplachne Echinochloa Ehrhata Eleusine Elyonurus Elytrophorus Enneapogon Enteropogon Entoplocamia Eragrostis Eremopogon Eriochloa Eudorina . Eulalia Kustachys INDEX OF GENERA. Contained in Parts 2 and 3. A PAGE 400 384 386 395 400 398 389 426 421 386 420 422 398 385 388 418 391 425 421 383 426 423 | 420 426 410 385 388 479 383 42] Xili Fingerhuthia Hemarthria Heteropogon Hyparrhenia Imperata . Kirchneriella Leptochloa Leucophrys Microchloa Monelytrum Odyssea Oropetium Panicum . Paspalum Pennisetum Pentaschistis Perotis Phragmites Pogonarthria Polypogon F O PAGE 425 383 387 387 383 389 419 407 419 419 392 390 396 398 407 399 409 399 XIV Schismus . Schizachyrium . Schmidtia Setaria Sorghum . Sphaerocystis Sporobolus Stipa Themeda . Trachypogon Tragus. Trichopteryx Tricholaena Tripogon . Triraphis . Index of Genera. PAGE 426 385 424 393 384 501 407 406 388 387 406 398 394 423 422 U PAGE Urochloa . p ; ; . 390 V Vetiveria . ; 3 : . 084 Volvox . : 427, 480, 529, 647 Vossia : A 3 5 Bis} W Wilkommia : : : ~ 419 xX Xyochlaena : é : . 395 ANNALS SOUTH AFRICAN MUSEUM. (Vou. XVI.) 1. -A Contribution to the Flora of the Leribe Plateau and Environs : with a Discussion on the Relationships of the Floras of Basutoland, the Kalahari, and the South-Hastern Regions.—By KE. P. Puiuuipes, M.A., D.Sc., F.L.S., Assistant. CONTENTS. PAGE Il. InrTRopuUcTION . : ; : ; ? é 2 Il. Tue Lerise Fiora ; ; : 4 : ; ; 3 Orography and Geology. ; : é ha ce Meteorology : ; : : 2 4 Oecological and Biological iment : ! : ‘ 5 Native Names and uses of Plants . : : : 10 Alien Flora . : : : . 12 Systematic Constituents of the Flora : : : 5 ils} Ill. Tue Eastern Mountain ReG@ion. — . ; : 5 leh Orography ; é : . : oe lle Geolovy : : : : Le PUL? Meteorology : : : : . : S LY Systematic Constitucnts of the Flora ¢ ; : ot hee IV. THe Kauanarr FLora ; : . E : SB Introduction : : ; : e523 Meteorology : F : 25 Systematic Constituents of the F toes : : : A AD V. A ComparkisON OF THE FLORAS OF ''H# HASTERN MOUNTAIN bP AND KaLAHARI REGIONS ; 2 : , : Beau VI. Tur SourH-Easrern REGION . : : ; P 5 6) 2 Annals of the South African Museum. VII. A Comparison or THE FLoRA oF THE EastERN Mountain AND SoutH-EasteErRN Rea@ions . : ; : d > Bil VIII. Summary ; . 3 ; ; ; : 5 OY IX. List or Species occuRRING IN THE Eastern Mounrain Ruaion, witH Descriprions or NEW SPECIES : : 5 33 X. InDEX TO GENERA ‘ : d : : : . 372 I. INTRODUCTION. THE present paper is the outcome of an investigation undertaken by the writer, of the flora of Leribe in northern Basutoland. Hitherto Basutoland has not received the attention from botanical collectors which it deserves, and the only collections as far as I am aware, are those of Cooper, who journeyed through Basutoland (unfortunately his plants are not properly localised); the Rev. J. Buchanan, who botanised in the Leribe District ; Mr. J. Thode, who collected on the Basutoland side of the Drakensbergen; and Mr. E. E. Galpin, who explored the high mountains separating Basutoland from Barkly East and as far inland as the Buffalo River Waterfall. The results of Mr. Galpin’s investigations were read before the South African Association for the Advancement of Science in 1908. The Rev. H. Dieterlen and his wife, Madame A. Dieterlen, of the French Protestant Missionary Society, who were stationed for many years at Leribe, have been deeply interested in the local flora. Madame Dieterlen has made an exhaustive collection of the native plants, the bulk of which she forwarded to the South African Museum herbarium for identification. The writer has thus had unique opportunities during the past seven years of becoming fairly well acquainted with a part of the Basutoland flora, and more especially with that of Leribe. In February 1913, he paid a visit to Leribe for the purpose of studying the flora on the spot. When naming Madame Dieterlen’s plants and comparing them with the specimens in our herbarium, I was particularly struck by the fact that so many of the species also occurred in the Eastern parts of South Africa such as the Transkei, Komgha, Pondoland, East Griqualand, Natal, etc., and then came to the conclusion that Basutoland ought not to be included in the Kalahari Region. This led me to make a more detailed study of the subject, with the result that I have attempted in the following pages to prove what Bolus * first suggested, viz. that Basutoland and parts of the surrounding country form a distinct floral area. * “Sketch of the Floral Regions of South Africa,’ Science in South Africa. Cape Town. 1905. Flora of the Leribe Plateau and Environs. 3 The principal collections (besides those of Madame Dieterlen and my own) which have been examined and embodied in the list at the end of the paper are the following: Dr. Bolus’ and Mr. H. G. Flana- gan’s, made in the neighbourhood of Witzie’s Hoek and the Mont- awux-Sources ; collections made by Messrs. J. Thode, J. M. Wood, and M. Evans on the Drakensbergen aud the higher parts of Natal ; Mr. Galpin’s and Drége’s collections from the high mountains separating Basutoland from Barkly East; Burke and Zeyher’s, from the Orange Free State. My thanks are due to Mrs. F. Bolus, B.A., the Curator of the Bolus Herbarium, South African College, Cape Town, who has allowed me every facility for consulting that collection, and who has verified the naming of some species of Orchidaceae and Mesembrianthemum ; to Dr. 8. Schonland, M.A., of the Albany Museum, Grahamstown, who examined some critical species of Crassulaceae; to Dr. O. Stapf, F.R.S., the Curator of the Kew Herbarium, who named some of the Gramineae anl Cyperaceae ; to the late Dr. J. M. Wood and Mr. E. E. Galpin, F.L.8S., who kindly sent me for examination some type specimens which were in their respective collections. ‘lo Madame Dieterlen I am under a deep debt of gratitude for her hospitality, which was extended to me and my wife when at Leribe; for her efforts in obtaining the Sesuto names and uses of the local flora; and for allowing me the use of her large herbarium. — It is due solely to her untiring zeal and energy that our knowledge of the Leribe flora has been gained. Il. PHE LERIBE FLORA. OROGRAPHY AND GEOLOGY. The portion of the country investigated botanically was the Leribe Plateau, its slopes, ravines, and the surrounding piains (see fig. 1). Leribe Plateau is situated in the Leribe District, Northern Basuto- land, not far from the borders of the Orange Free State. The plateau stands as an isolated table-land, 5-6000 ft. high, sharply marked off from the surrounding plain by walls of rock which rise perpendicularly for 2-800 ft. The area of the summit is roughly 30 sq. miles. At the southern end is a peak, Qoqolosi Peak, which reaches nearly 8000 ft. in altitude. The lower slopes of the plateau merge gradually into the plain, and numerous ravines, with rocky stream-beds, are found all round its boundaries. The plateau is built up of Cave-sandstone belonging to the Storm- berg Series (see also geology of the Eastern Mountain Region) with a 4 Annals of the South African Museum. capping of dolerite on Qoqolosi Peak. The plain and valley of the Caledon River are composed of silt washed down by the streams and mountain torrents which are suddenly formed during the heavy summer rains. Ihigets als LERIBE PLATEAU oh wt LERIBE = DISTRICT METEOROLOGY. I have been unable to obtain records of the temperature at Leribe, but from data available we know that extremes of heat and cold are experienced during the course of the year. During the months of June and July the temperature falls below freezing point during the night. Occasionally frost occurs as early as May and lasts until September, and snow falls once or twice a year but melts after a few hours, except on the highest peaks, where it sometimes remains for days or weeks. In summer, from November to March, severe hail- storms accompanied by thunder and lightning are frequent, and, as will be seen from the accompanying table, these are also the rainy months. ‘The rains are usually preceded by northerly winds, but strong westerly winds blow from August to October. The country is Flora of the Leribe Plateau and Environs. 5 often subjected to drought, when the vegetation presents a scorched-up appearance. The mean average rainfall at Leribe during the five years 1909-1918 was 27°01 in., most of which falls from December to March. The dry months are from June to September. Rainfall at Leribe Mission Station.* : 3 Average for 1909. 1910. OTE 1912. 1913. 5 years. January . 10°12 6°84 225 1°49 Dito 4-68 February . 811 5°96 110 = 10°44 44.2 6:00 March . 3°32 3°14 6°25 1:59 3°75 3°61 April me GO? 0°87 2°75 4:27 1-68 2-11 May ess) 0-64, 97 2°02 0-19 1°44. June 70700 0-26 0-21 0-79 0-09 0:27 July - Odd O15 0°85 0:24. 0-02 0:27 August . 0°67 0-06 0-76 0:00 = 0°65 0°42 September _1°23 177. 0:05 0-01 0°65 0°74 October . 0°38 5°38 2-29 1:34. 316 2251 November 2°60 1:47 2°24 1:16 2°61 2°01 December 6°06 1°59 0°65 6°28 0°00 2°9] OECOLOGICAL AND BroLoGicaAL FEATURES. The general aspect of the plains, slopes, and plateau is that of a grass-land, though in a few localities thickets composed principally of Cussonia paniculata and Huclea lanceolata, intertwined with Rhoicissus cuneifolia, are seen under the precipitous cliffs. The valley and plain are largely cultivated with maize and Kaffir-corn, while the plateau affords good grazing for numerous head of cattle, sheep, and goats. On the plain the predominant grass is Hragrostis plana, among which many herbaceous and sub-herbaceous plants are found. Owing to the extensive cultivation it is not surprising to find also a large number of imported weeds, and such species as Datura Stramonium, Verbena officinalis, Chenopodium murale, C. ambrosioides, Alternanthera Achyrantha, Amaranthus paniculatus, ete., are commonly met with. On the west slopes of the plateau the predominant grasses are Andropogon hirtus, Eragrostis curvula vay. conferta, Anthisliria imberbis var. mollicoma, and Hragrostis plana: Tricholaena setifera, Elionurus argenteus, Setaria flabellata, Andropogon contortus, Penni- setum Thunbergi are common, while less frequent are Hragrostis * Records kept by the Rev. H. Dieterlen, 6 Annals of the South African Musewm. brizoides and E. gummiflua. Growing among the grasses are numerous annuals and perennials, mostly of a herbaceous character and low growth, few, except round marshy patches, exceeding 2 ft. in height. Some of the typical plants found in these localities were Gnaphalium undulatum, Cynoglossum micranthum, C. enerve, Acrotome inflata, Cyanotis nodiflora, Conyza podocephala, Helichrysum rugu- losum, Nemesia foetens var., Rhus Sonderv var., Mahernia coccocarpa, Zinma multiflora, Salvia repens, Solanum indicum, Sutera filicaulis, Hieracium capense, Rhus discolor, Polygala rarifolia, Dianthus scaber, Striga elegans, Berkheya setifera, Vernonia Kraussii. Under, or in the shade of, rocks or large boulders scattered on the slopes are found Oldenlandia Heynei, Tephrosia capensis, Stachys aethiopica var., Eucomis undulata, Richardia albomaculata, Lobelia Erinus, Dicoma anomala, and numerous ferns, all shade-loving plants. In damp _ patches Cyperus usitatus is found growing subsocially. The few bushes present are not sufficiently numerous in individuals to affect the land- scape. Those which were recorded are Asclepias fruticosa, Leonotis mollis, Rhus erosa, and, Solanum indicum. The southern slopes of the plateau do not bear such a luxuriant grass formation as the western and northern slopes, The predominant grass present is Hragrostis plana. Eragrostis curvula var. conferta, so common on the western slopes, is here of secondary importance. It was on this side of the plateau that the heaths, orchids, and most of the bulbous plants were collected. The rest of the vegetation is of a very similar character to that described above, but only more evident owing to the less dense covering of grass. Many of the species taken here are common to the other slopes of the plateau, though some, such as Pentanisia variabilis, Crassula parvula, Monsonia biflora, Pharnacewm detonsum, Wahlenbergia depressa, Phytolacca heptandra, etc., were confined to this locality. In some places Salvia stenophylla and Phytolacca heptandra form large patches to the almost complete exclusion of other plants. The western and northern ravines, as stated above, are thickly wooded. When the sides of the ravine are steep the bush extends to under the cliffs, but when the sides slope gently to the bed of the ravine they are covered with grass. Leucosidea sericea is the pre- dominant tree. ~ Of secondary importance are Kiggeleria africana, Cassinopsis capensis, Ilex capensis, Rhamnus prinoides, while specimens of Plectronia ciliata and Celastrus buxifolius are rare. The under- growth is composed mainly of Myrsine africana, but Huclea coriacea, Cliytia pulchella and Buddleia salvifolia are common; Rhus pyroides being less frequent. At Lefi’s Kloof, the grassy slopes on either side Flora of the Leribe Plateau and Environs. 7 of the broad shallow opening to the ravine proper are the habitats of the only Protea found at Leribe, viz. P. caffra. In habit it closely resembles P. grandiflora of the Western Province, and at a distance might easily be mistaken for that species. The area which the plants occupy is very limited, as no specimens are found in the deeper parts of the ravine. Numerous shade-loving plants, almost all herbaceous, are found growing in the ravines. Some are found on the damp grassy slopes in the shadow of the cliffs or under trees, others grow deeper down in the ravine under rocks. Among many such plants gathered were Alepidea amatymbica, Hypoxis Gerrardi, Mysotis afro- palustris, Geranium canescens, Galiumn rotundifolium, Scabiosa Colum- baria, and numerous ferns. The streams from the ravines, as they enter the level plain, form small marshes. Such situations are chiefly occupied by Marisews congestus and Polygonum serrulatum, while bushes of Cluytia natalensis are common round the wet margins. The summit of the plateau presents the appearance of a large flat covered with short grass. Nowhere is there the same luxuriance of grass-growth as in the valley and on the slopes, and bush of any description is quite absent. The low grass-formation is due chiefly, I consider, to the amount of grazing by cattle, sheep, and goats, and also to the fact that the tall grasses found at the lower levels are absent on the summit, their place being taken by Aristida adscensionis, Elionurus argenteus, Digitaria monodactyla, and Andropogon contortus. It is very possible that the more succulent grasses have been destroyed on the plateau by overstocking, leaving the coarse grasses in possession. I might mention that very few, if any, cattle are allowed to graze on the lower slopes in summer, owing to the difficulty of keeping them from damaging the crops. Among the plants gathered on the plateau were Lobelia (Metzleria dregeana), Rhynchosia Totta, Venidium arctotoides, Alchemilla Woodii, Dicoma anomala, Crassula basutica, etc. The damp banks of streams and pools were the habitats of hydrophilous plants such as Polygonum spp., Rumex spp., Salvia repens, Helichrysum similimum, Xyris capensis, Utricularia spp., Limosella major, Ranunculus Meyeri, and numerous Cyperaceae. Qoqolosi Peak carries a different,formation from that of the surround- ing plain. As the peak is ascended the grasses disappear and their place is taken by succulent plants, plants of a dwarf habit, and small shrubs. The succulent plants were mostly species of Crassula ; Lotononis Woodii, Melolobium adenodes, Heliophila basutica (Pl. V, fig. 1) were typical of the dwarf plants; Anthospermum pumilum, Sutera aspalanthoides, S. pristisepala, Melolobium microphyllum were representative of the stunted bush, 8 Annals of the South African Museum. Among the plants which go to make up a plant community we may expect to find plants of very diverse habit and form, and the flora of Leribe is no exception to this when the individual species are studied. The conditions under which plants live is the dominating factor in determining their form, so that it is not surprising to find that nearly all the species, with the exception of those inhabiting the damp shady ravines, exhibit marked xerophytic characters, and many, especially from the higher altitudes, assume a sub-alpine habit. (See Ase. Vel.) Annual plants as Cotula anthemoides, Arctotis stoechadifolia, Ursinia annua, Sebaea exigua and Exochaenium grande, which spring up after the summer rains have commenced, are not numerous. The majority of the species are perennials, and exhibit various devices whereby they are enabled to withstand the extreme drought during the winter months from June to September, when less than 2 in. of rain falls for that period. Plants with long tap-roots, thick tuberous roots, and underground woody stems, are common. Sonchus nanus, Ipomoea simplex, Arthrosolen gymnostachys, Lotononis basutica, and Argyro- lobium spp. may be taken as typical of the tap-rooted plants. Senecio erubescens, S. serra, Helichrysum undatum var. pallidum, H. latifolium, Gerbera piloselloides, G. viridifolia have many thick roots. Helichrysum platypterum, H. psilolepis, Othonna natalensis, and Crepis polyodon (PL. VI, fig. 3) are representative of plants with woody underground stems. A few plants, such as Brachystelma foetidum and several Orchidaceae, possess underground tubers. Under this category must also be mentioned a number of Monocotyledons with underground bulbs and corms. Plants with a tufted habit forming cushions are not numerous ; the two most typical representatives of this form are Gazania armerioides and Psamimotropha androsacea (Pl. V, fig. 2). Prostrate plants are represented by Helichrysum ericaefolium var. albidulum, H. caespititium, H. chinosphaerum (Pl. VI, fig. 4), H. Randii (Pl. VI, fig. 2), Lightfootia denticulata, Lobelia dregeana, Ursinia montana, Herniaria hirsuta, Argyrolobium nanum, etc. 'The prostrate creepers met with were Convolvulus capensis var. plicata, C. ulosepalus, Ipomoea oblongata var. hirsuta, and Dicoma anomala. The principal climbers noted were Clematis brachiata, Rhoicissus (Cissus cuneifolia), Cynanchum virens, and Riocreuxia picta. Sarco- stemma viminale and Helichrysum Sutherlandi were the only two pendent species met with. They were both growing on cliff faces and hanging down in festoons. Succulent plants constitute a very small percentage of the flora both in species and in individuals. Cotyledon orbiculata may perhaps form an exception to the latter statement, as Flora of the Leribe Plateau and Environs. 9 near the cliffs at the entrance of the large ravine between the Mission Station and Jonathan’s village the writer came across a large area on which this species was growing. The other succulent plants noted were Mesembrianthemum spp., Crassula spp., Aloe spp., Bulbine spp., and Stapelia flavirostris. Many plants only possess radical leaves which lie more or less flat on the ground. Some of these recorded were Wahlenbergia androsacea, Berkheya (Stoebea aristosa), Haplocarpha scaposa, Gerbera piloselloides. G. viridifolia, Sonchus nanus, Chironia palustris, Manulea crassifolia and Crabbea hirsuta. For species with the habit of trees or bush the reader is referred to the enumeration of plants found in the ravines, with the addition of Printzia pyrifolia, Osteospermum moniliferum, Rhus spp., Buddleia salvifolia, Halleria lucida, and Lycium spp., which form bushes from 4-8 ft. high. The leaves exhibit xerophytic characters in various ways and degrees. In indumentum all stages from a pubescent to a woolly leaf are met with. Plants with leaves woolly on both surfaces are Senecio macro- spermus, Helichrysum callicomum, H. cephaloideum, H. psilolepis, H. aureo-nitens, H. calocephalum (PI. VI, fig. 1), H. adenocarpum, Leo- tonyx squarrosus, Sopubia cana, ete. Helichrysum Mundii, H.latifolium, HI. undatum var., Pentzia pyrifolia, Berkheya Kuntzii, B. onopordifolia, B. montana, B. alba, Chilianthus corrugatus, Stachys rugosa var. have leaves woolly on the under-surface only. In Gerbera piloselloides, Mentha longifolia sub-sp. capensis, Argyrolobium nanum, Lotononis basutica, etc., the leaves are hairy and not woolly. Solanum indicum and S. tomentosum have tomentose leaves. The only plants with glandular leaves are Sutera pristisepala, Drosera burkeana, and D. ramentacea var. Scabrid-leaf plants were more frequent’ and were represented by Striga lutea, S.elegans, 8. Thunbergii, Melasma basuticum, and Verbena venosa. The amount of leaf-surface developed shows the same great range of variation, viz. from reduced leaves in Wahlen- bergia denudata and Harveya spp. to the broad leaves of typical mesophytes found in the ravines. Ericoid leaves were the most common, and were noted in Helichrysum athriziifolium, H. rugulosum, H, squarrosum, Pentzia virgata, Athriaia elata, Metalasia muricata, Stoebe cineraria, Eriocephalus punctulatus, Wahlenbergia depressa, Inghtfootia denticulata, Lobelia decipiens,and Passerina ericoides. In Gazania armerioides the leaves are involute: in Sutera atropurpurea, Selago longipedicellata, and Walafrida densiflora the leaves are small and faseicled. Among the plants which develop thorns may be mentioned Lycium spp., Solanum spp., Asparagus spp., and Celastrus (Gymnosporia) buxifolius. All the parasitic plants found belong to the Serophulariaceae. 10 Annals of the South African Museum. There were 4 species of Harveya, 4 species of Striga, and 2 species of Melasma. Insectivorous plants were represented by 2 species of Utricularia and 2 species of Drosera. I should not consider this a floral region of great beauty where the size and colouring of the flowers are concerned. This conclusion has been arrived at from a study of the flowers of the various species, and from an attempt to group them (excluding the Cyperaceae and Gramineae) into the following three Classes,* viz. : I. Large, gaily-coloured flowers, or when a number of small flowers are grouped in large and conspicuous heads. II. Medium-sized flowers. III. Inconspicuous flowers. Class I. (e.g. Brunsvigia, Ipomoea, Gladiolus, ete.), approximately 46 per cent. Class II. (e.g. Ranunculus, Silene, Erica, Sutera, etc.), approxi- mately 46 per cent. Class IIT. (e.g. Rhus, Celastrus, Phytolacca, etc.), approximately 8 per cent. The great bulk of the species produce dry fruits, either dehiscent or indehiscent; fleshy fruits are only found in less than 5 per cent. of the total number of species. The dry dehiscent and dry indehiscent fruits are represented in about equal proportions. ‘The Compositae, Gramineae, Cyperaceae, etc., are the largest representatives of the latter group (included among these are those fruits such as are found in the Umbelliferae and Labiatwe which split up into one-seeded portions), while the Liliaceae, Leguminosae, Orchidaceae, Scrophu- lariaceae, etc., contain the largest number of species which produce the former type of fruit. It is not within the province of this paper to go into any further details regarding the fruit and seeds; it will suffice to mention that such forms of fruit as (1) hairy fruits (e. 4. Clematis, Compositae), (i1) Spiny fruits (e.g. some Boraginaceae), (ii) winged fruits (e.g. Rumex) are met with. Hairy seeds occur in the Asclepiadaceae, Salix, and Hriospermum ; winged seeds in many of the Monocotyledons (e. g. Gladiolus, Urginea, etc.). Native Names anv Uses or Puants. In un interesting paper t on ‘ Zulu Medicine and Medicine-Men,” = Pay the author im the Introduction states: “It is by no means an * O. Stapf, ‘The Flora of Kinabalu,” ‘'I'rans. Linn. Soc.,’ ser. 2, vol. iv, 1894. + Rey. Alfred Bryant, “Zulu Medicine and Medicine-Men,” ‘Annals of the Natal Government Museum,’ ii, 1 (1909). Flora of the Leribe Plateau and Environs. 1 exaggeration to affirm that comparatively the average Zulu can boast of a larger share of pure scientific knowledge than the average European.” This statement applies with equal force to the natives of Basutoland. As the reader will observe in the following list of plants, all the species collected in Basutoland by Madame Dieterlen and the author have Sesuto names. In the majority of cases each species has a name of its own by which it is distinguished from closely allied species or from plants with a similar appearance, but, as for example in the Juncaceae, Cyperaceae and Orchidaceae, there is usually but one “generic”? name which includes all the species of the particular orders. The natives like the pre-Linnean botanists, in many cases express the name in a short sentence, using as a basis either the particular habit of the plant when it presents any striking feature, such as the development of thorns, excessive hairiness, etc., or its use when it possesses any medicinal or economic value. It is difficult to translate into English many of the Sesuto names, and in all cases a literal translation is given. Where “meaning unknown” appears after a Sesuto name it indicates that the natives can give no reason for the plant being so called, not that they do not know it. All the particulars as to the uses the plants are put to have been given me by Madame Dieterlen, who has had a unique opportunity, during her nineteen years’ residence at Leribe, of obtainmg such information. The reader will notice that no mention is made of the uses of many of the plants, and this means that the natives from whom Madame Dieterlen obtained her information either did not know of any use the particular plant was put to, or would not tell, fearing that they might be imparting valuable secrets. In a great many instances, when a native doctor is called to attend a patient, he first consults his divining-bones to find out the cause of the ailment, ana after satisfying himself on this point, reconsults them to find out the particular medicine to be administered. The Basutos, like other South African natives, are great believers in witch-craft, and they are firmly convinced that most of their ailments are due to the evil influence of some person who may be even living at a distance. To avert such evil influences or to break their spell when they are at work, the natives employ divers “ medi- cines”’ and charms. ‘These are administered in various ways, as decoctions, lotions, powders, etc., or sprinkled in the huts or on their persons, burnt in their courtyard, etc., as charms. The ‘“ medicines ” prepared from various plants are supposed to turn away lightning, bring rain, or increase their crops, and there are few things in their a 12 Annals of the South African Museum. daily life that cannot be affected, either for good or ill, by a particular preparation from a plant or plants. ALIEN FLORA. There is a considerable alien flora at Leribe, consisting of 48 species spread over 38 genera and 18 Orders. Most of the species are now widely spread over South Africa. The plants mentioned im the list below .have not been included in the Leribe flora on the following pages : Papaveraceae. Argemone mexicana, Linn. Cruciferae. Brassica pachypoda, Thellung, var. Lepidium Schinzu, Thellung. Nasturtium officinale, R. Br. Caryophyllaceae. Stellaria media, Vill. Spergula arvensis, Linn. Malvaceae. Malva parviflora, Linn. M. verticellata, Linn. Zygophyllaceae. Tribulus terrestris, Linn. Geranvaceae. EKrodium ecicutarium, L’ Her. Leguminosae. Medicago laciniata, All. Phaseolus Mungo, Linn. Onagraceae. Epilobium hirsutum, Linn. E. tetragonum, Linn. Oenothera biennis, Linn. Oe. tetrapetala, Cav. Oe. villosa, Thunb. Compositae. Erigeron canadense, Linn. Gnaphalium luteo-album, Linn. Xanthium spinosum, Linn. Bidens leucantha, Willd. Senecio vulgaris, Linn. Cnicus lanceolatus, Willd. Sonchus oleraceus, Linn. Solanaceae. Solanum nigrum, Linn. Physalis peruviana, Linn. Nicandra physaloides, Gaertn. Datura stramonium, Linn. Nicotiana glauca, R. Graham. N. rustica, Linn. Scrophulariaceae. Vernonia anagallis, Linn. Labiatae. Mentha aquatica, Linn. Amarantaceae. Amaranthus paniculatus, Linn. Achyranthes aspera, Linn. Chenopodiaceae. Chenopodium ambrosioides, Linn. C. murale, Linn. C. album, Linn. C. foetidum, Schrad. Roubieva multifida, Moq. Polygonaceae. Polygonum aviculare, Linn. P. lapathifolium sub.-sp. macu- latum, Dyer & Trim. Rumex Acetosella, Linn. Flora of the Leribe Plateau and Environs. 15 Huphorbiaceae. Gramineae. Kuphorbia Peplus, Linn. Poa annua, Linn. Bromus unioloides, H.B.K. Urticaceae. Hordeum secalinum, Schreb. Cannabis sativa, Linn. Zea Mais, Linn. (cultivated). Tue SysteEMATIC CONSTITUENTS OF THE FLORA. The following list deals only with the area discussed in the previous pages, and unlike most lists of this nature, which after all are only approximate to the truth, it may be accepted as representing the true composition of the flora of the Leribe plateau and its environs. I do not think there is a single species of Phanerogams or Ferns from this locality that has not been collected by Madame Dieterlen. ° > Orders. Genera. Species. Cryptogamia vasculares fe ee lO. 38 Phanerogamia. Dicotyledons . = G45 220, 920 04:78 Monocotyledons. Seloe ee LOO Re 271 Total (Phanerogams) 77 329 749 Proportion of Monocotyledons to Dicotyledons Teele me Proportion of genera to species”. : 5 Sis Peay Orders. ae see 1. Compositae . : . 144 : . 19°22 2 Gramineae . ; 87 SWS! 3. Cyperaceae . , 2 08 a AD 4. Liliaceae. « | Oe 5 P07. 5. Scrophulariaceae . rte AD oa) 6. Leguminosae : Be AS . 8°20 7. Asclepiadaceae . cand 3°60 8. Orchidaceae . -) 25 BERR 9. Amaryllidaceae . Me . 2°80 10. Crassulaceae ok 0) ll. Iridaceae ae miley, : . 240 12. Campanulaceae lA es MLNS )7/ 13. Rubiaceae. ae 1-87 14. Geraniaceae . 13 lee 15. Cruciferae. 12 : BP PRD 16. Anacardiaceae ; 10 : 1:33 17. Labiatae : : : 10 : : 3s 14 Annals of the South African Museum. 18. Umbelliferae A J 9 ; 20) 19. Juncaceae . ; : 8 : » 6-06 20. Solanaceae . , . 8 ; e106 The following are the remaining Orders arranged according to the number of species in each: Boraginaceae, Convolvulaceae, Ficoideae, Rosaceae, Santalaceae (7) ; Huphorbiaceae, Malvaceae, Polygalaceae, Polygonaceae, Ranunculaceae, Sterculiaceae (6) ; Cucurbitaceae, Eri- caceae, Gentianaceae, Thymelaeaceae (5) ; Caryophyllaceae, Ebenaceae, Selaginaceae (4); Amarantaceae, Commelinaceae, Urticaceae (3) ; Acanthaceae, Aroideae, Celastraceae, Dipsiceae, Droseraceae, Hyperi- cineae, Illecebraceae, Lentibularieae, Loganiaceae, Portulaccaceae, Salicineae (2); Ampelideae, Araliaceae, Bixineae, Capparideae, Hriocaulaceae, Halogoraceae, Linaceae, Menispermaceae, Myricaceae, Myrsineae, Naiadaceae, Olacinaceae, Oleaceae, Onagraceae, Papa- veraceae, Pittosporaceae, Proteaceae, Resedaceae, Restiaceae, Rham- naceae, Valeriaceae, Verbenaceae, Xyridaceae (1). The following are the largest genera: Helichrysum (30 species), Senecio (21), Crassula (13), Andropogon, Cyperus, Hypoxis (11), Lotononis, Rhus (10), Eragrostis (9), Bulbostylis, Juncus, Schizo- glossum, Scilla, Scirpus, Wahlenbergia (7), Asclepias, Asparagus, Digitaria, Gladiolus, Solanum, Sutera, Thesium (6). There are 189 genera represented by one species only. Ill. THE EASTERN MOUNTAIN REGION. All writers on the phyto-geography of South Africa, despite the divergence of their views as to the subdivisions of South Africa into botanical regions, have agreed in including Basutoland in a region generally termed the “ Kalahari Region,” the limits of which vary according to their individual ideas.* Thus Thodet includes the 8.W. Protectorate, Bechuanaland, the Orange Free State, Basutoland, the Transvaal Hoogeveldt (south of the 26th parallel), and parts of the N.E. Districts of the Cape Province in his “ Kalahari Territory”; Marloth defines a region which he calls the “ High Veldt,” and while * The reader should refer to a paper by R. Marloth (“ Wissenschaftliche ergebnisse der Deutschen Tiefsee-Expedition auf dem Dampfer, ‘ Valdivia,’ ” 1898-1899, ii, 3, Pls. 4 and 5), where a set of instructive maps is given showing the historical development of South African Phyto-Geography. + Justus Thode, ‘The Botanical Regions of Natal, determined by Altitude,’ Durban, 1901. t R. Marloth, “The Phyto-Geographical Subdivisions of South Africa,” ‘Report of the Brit. Assoc. Adv. Sc.,’ Cape Town, 1905. Flora of the Leribe Plateau and Environs. 1153 excluding much which comprises the ‘“ Kalahari Region” of Thode, agrees so far as to include Basutoland. Marloth however notes that “In the eastern parts the country is mountainous, and the vegetation consequently more varied.”’ Bolus,* whose paper is the classic on South African Phyto-Geography, states in the introduction to the “ Kalahari Region’: “It must at once be stated that this vast region is as yet so imperfectly explored as to its physical divisions, its aspect, its climate, and the systematic constituents of its vegetation, that it is impossible at present to do more than offer a very general view of a country which will need many years of study, and which will hereafter almost certainly require to be divided into several Regions or, at least, to be subdivided into Provinces.” As now treated it is bounded on the west by the still less explored Western Region, on the south by the Upper Region, on the south-east and east by the mountains of the South-Eastern Coast Region, on the north by the great Tropical Region, which is beyond the scope of our inquiry. It thus includes : 1. The higher eastern mountain country, forming parts of Cape Colony, Natal, and Basutoland, with an altitude higher than 3500 cr 4000 ft. (This, in our view, will probably hereafter require separa- tion as a Region or Province.) 2. Almost the whole of the Orange River Colony, of the Transvaal, and Bechuanaland. I quote the above in full, as Bolus was the first to realise the mistake of including Basutoland and parts of the surrounding country with the major portion of the Orange Free State, the Transvaal, and Bechuanaland. He was however unable to follow up his view owing to the lack of sufficient material. In the ‘ Flora Capensis,’ the standard work on South African Systematic Botany, Basutoland is also included in a Kalahari Region almost identical with that of Bolus. It is a portion of the first part of Bolus’ “ Kalahari Region” that is dealt with in this paper. OROGRAPHY. The Eastern Mountain Region, as the writer has limited it for the present, is a large tract of country lying between the 28th and 30th degree of latitude, and at its widest lmits extending from the 27th degree of longitude on the west to about midway between the * H. Bolus, “Sketch of the Floral Regions of South Africa,’ ‘ Scie South Africa,’ 1905. 16 Annals of the South African Museum. 29th and 30th degree in the east; both at its northern and southern extremities Sit becomes narrower. A glance:at the map will show that it has}a well-defined natural boundary on the west in the shape of a continuous range of mountains running almost parallel with the Caledon River (Pl. VIT). The change in the flora westward Fia. 2. 4Q000 F*...... aR @000 : 6.000 4,000 2000 SEA LEVEL. points to the above mountain-range being almost a true floral boundary. The southern limits of this region I have’provisionally regarded to be the Wittebergen, which separates Basutoland from Barkly*Hast, though Iam led to believe from a study of] what is known of the flora of Aliwal North, Wodehouse, and Barkly East, that when these districts are better explored botanically, the Stormberg Range will prove to be the southern boundary. On the east the boundary line follows the contour of the Drakensbergen at an altitude of about 5000 ft., below Flora of the Leribe Plateau and Environs. 17 which the flora assumes a different character. The Natal portion of the Eastern Mountain Region would correspond to Thode’s “ Mountain Region.”* The northern portion of our Region converges on the Drakensbergen and does not exceed 50 miles in breadth. I have not extended my enquiries in detail beyond the 28th parallel, though Bolus extended this region northward as a narrow strip, as far as Barberton. The Hastern Mountain Region as defined above then includes the whole of Basutoland, a strip of the Orange Free State from Nelson’s Kop on the north to the Orange River on the south, which forms a rough horse-shoe, a small portion of the Cape Province on the south, and a portion of Kast Griqualand and Natal bordering the Drakens- bergen. This area forms a well-defined elevated table-land above 5000 ft. in altitude on its outermost limits, with a general elevation of 8000 ft. in the greater part of Basutoland, and rising to peaks 10,000 ft. and over on the Drakensbergen (see fig. 3). Basutoland, which forms the major portion of the Eastern Mountain Region, is a great tract of high country, 10,300 square miles in area, of which the Drakensbergen is the south-eastern edge. The top of the Drakensbergen slopes gently westwards, and the sloping ground is covered with grass, heather, bog, and rocks; all the cliffs are on the Natal side. A long range of mountains, the Maluti Mountains, traverses Basutoland in a south-westerly direction, but throughout the country is mountainous and in some parts difficult of access. GEOLOGY. For a more detailed account of the geology the reader is referred to papers by Dr. A. L. du Toitt and Mr. F. F. Churchill,t from which the following has been extracted. The whole of Basutoland is built up of Stormberg beds divided up as follows in descending orders (see fig. 3).§ 4. Volcanic beds. 3. Cave Sandstone. 2. Red beds. 1. Molteno beds. 1. Moureno Brps.—This formation crops out over flattish ground * Thode, loc. cit. 7 A. L. du Toit, “The Forming of the Drakensbergen,” ‘Trans. S. Afr. Philos. Soc.,’ vol. xvi, 1, pp. 53-57. t F. F. Churchill, “ Notes on the Geology of the Drakensbergen,” loc. cit., vol. x, 3. § This sketch has kindly been drawn for me by Mr. 8. H. Haughton, B.A., F.G.S. ap — 18 Annals of the South African Museum. at the foot of the Drakensbergen, and on the western borders of Basutoland. Fia. 3. ORANGE ee Beth leh FR TATE- pete Leomith: alas 3 Drakensberg Lavas Storm berg Sedimentary. 3 Ker Aarree Beds. aicies (0) 4 ower Karrco Beds, Zo: re me \ Jes a . “4 : 2? & re = me LS be at v i) Sy § : : 7 4 oS “ & 4 = ¥ 4 ~ a = dd g S < ) « ¢ x Q | 2. Rep Brps.—These extend into Basutoland, and are exposed in the valley of the Orange River. They can also be followed along the flanks of the mountains of the Basutoland border from Kornet Spruit, Flora of the Leribe Plateau and Environs. 19 past Mafeteng, Maseru, Tholtse, to the head of the Caledon River. The high ground to the north of Ficksburg and round Harrismith is built up of Red beds and Cave Sandstone. 3. Cave SanpsronE.—So called on account of the unequal weather- ing which causes the formation of large caves. It varies in thickness from 2-400 ft., and rests on the Red beds. 4. Voutcantc Beps.—Consist almost entirely of basic lavas which penetrate the sedimentary rocks. On the Natal side of the Drakens- bergen the dolerite capping does not extend lower than 5900 ft. to 6100 ft. above sea-level. The whole of these Stormberg beds rest on the Karroo beds which crop out in the Orange Free State and in the lower parts of Natal (see fig. 3). The soils derived from the Stormberg Series are very uniform in texture and fine in grain, and contain relatively large proportions of lime.* The reader will have noticed that both the altitude and the geological formation mark off the Eastern Mountain Region into a well-defined area. METEOROLOGY. The temperature and rainfall records from this Region are very meagre. What records I have been able to obtain are given in the following tables : TasLe I.—Annual Rainfall, 1901-1908. ore :7 1901. 1902. 1903. 1904. 1905. 1906. 1907. 1908. Mean. Mafeteng . 5600 — — — 21-66 26:55 25:17 37°37 22°67 26-68t Mohalie’s Hoek — 27:22 29°74 18°94 20°85 28°72 28-19 41°30 28:02 27:87 2 Maseru. . 5065 29°46 82°24 23°20 28°47 30°08 30°98 39:09 21:64 24-89 Teyateyaneng. 5690 37:00 33°78 20°52 2449 30°28 28-00 39°30 27-21 30-07 Moyeni Quth- ing : . 6000 42°81 — 22:01 24°47 37-53 — 4612 — 34587 Quacha’s Nek . 6236 28°74 27:00 27°63 25°31 32°35 27°68 46:19 43°37 32-28 + Average for five years. The only other records are for Harrismith, 25-43 inches (1905) 24°91 inches (1918), and Thaba Unchu, 19-09 inches (1904). For the rainfall at Leribe the reader is referred to a table on p. 5. * C. F. Juritz, “The Fertility of some Colonial Soils, as Influenced by Geological Conditions,” ‘ Trans. S. Afr. Philos. Soc.,’ vol. XViil, p. 27. 20 Annals of the South African Musewn. TasLE Il.—Rainfall during the six Summer Months (October—May),. 1901-1908. an 1901. 1902, 1903. 1904. 1905. 1906. 1907. 1908. Mean. Mafeteng . 5600 — — — 17:01 19°79 20°03 26°95 15°84 19°20* Mohalie’s Hoek — 20°67 20°76 12°18 16°62 22:06 23°66 29°64 19°42 20°62 Maseru. . 5065 24°02 22°99 17:10 19:05 25°50 26°37 30°36 15°62 22°62 Teyateyaneng. 5690 30°46 24°96 1430 18°98 25°69 23°58 29°75 20°42 23°51 Moyeni Quth- ing : . 6000 32°54 — 1460 18°38 2949 — 32°63 — 25°52* Quacha’s Nek 6236 23°98 21:95 20°30 24°10 28°68 24-82 36°74 — 25 59f * Average for five years. + Average for seven years. TasE ITl.—RKainfall during the six Winter Months (April—September), 1901-1908. Ait JN y901. 1902, 1903. 1904, 1905. 1906. 1907. 1908. Mean, Mafeteng . . 5600 677 801 — 465 676 S14 1042 683 694* Mohalie’s Hoek -— 655 898 676 423 666 458 11:66 860 7°24 Maseru , . 5065 5:44 925 610 442 458 461 873 602 6:14 Teyateyaneng . 5690 654 882 622 551 459 442 955 679 663 Moyeni Quthing 6000 1027 90! 741 609 804 — 1849 — 905+ Quacha’s Nek . 6236 4°76 505 7°35 121 367 2°86 12°45 1035 5:96 * Average for seven years. y Average for six years. fo) ” TaBLE 1V.—Rainfall during the Months of May—August, the period of Least Rainfall. eee 1901. 1902. 1903. 1904. 1905. 1906. 1907. 1908. Mean. Mafeteng . . 5600 140 189 — 2:14 266 101 480 3:46 2:48* Mohalie’s Hoek . — 238 3°78 3:26 1:93 2:66 1:25 491 4:44 3:07 Maseru : - 5065 0:82 3:99 2:02 2:50 1°60 1°20 2°63 271 2175 Teyateyaneng . 5690 1:88 2:11 1°32 3:26 0938 O84 2:91 380 2:13 Moyeni Quthing . 6000 4°19 545 319 295 365 — 587 — 421+ Quacha’s Nek . 6236 1:03 2:90 223 077 0:93 063 2:99 256 1°75 * Average for seven years. + Average for six years. A reference to the above tables will show that it is a region of summer rains; over 75 per cent. of the total rainfall takes place from October to May. The average annual rainfall is about 29°39 in., which compares favourably with other parts of South Africa, except the Karroo, Upper, and certain parts of the Kalahari Regions. The rainy months are usually January-February, when severe thunder- storms often accompanied by hail, are frequent. During March and April the rainfall diminishes, and a “dry period” (May—August), during which less than 5 in. of rain are registered, intervenes between Flora of the Leribe Plateau and Environs. 21 the first rains which begin to fall about September or October, The “dry period” is not so severe as that which occurs during the corre- sponding months in the western part of the Kalahari, as a comparison with the Table VII on p. 25 will show. The only temperature records available are given in the following table : Taste V.—Temperature. Altitude Number Absolute Absolute in feet. of years, maximum. minimum. Teyateyaneng : : 56900 MG) eee S260 Hy . ) 3a05 B Mohalie’s Hoek : — 5 82:02 30;5eb Leribe . : ; 5210 2 83:0° F. : 32 2e0ne Moyeni Quthing . : 6000 2 SPHOP IBS 5 9 Gi0ytehy 10h Butha Buthe : : 5500 1 Sele BL og | BGR 10 It will be seen from the few data above that the extremes of temperature are by no ineans great. This table should be compared with Table VIII on p. 25. The few figures give no indication of the temperature conditions on the high mountain peaks, some of which rise to 10,000 ft. or more. These peaks are among the highest found in South Africa, but none rise above the line of permanent snow, though in winter many of them are capped with snow for days or weeks at a stretch. The proper exploration of the peaks in this Region will undoubtedly yield much that is of botanical interest. Throughout this Region frosts are of frequent occurrence on the low ground in winter. SysTEMATIC CONSTITUENTS OF THE FLORA. Orders. Genera. Species. Cryptogamia Vasculares , = : 25 ; 53 Phanerogamia— Dicotyledons . : 76 ols . 1041 Monocotyledons . 15 , 147 Yao Total (phanerogams ) oil 466 1558 Proportion of Monocotyledons to Dicotyledons =. 1: 208 Proportion of genera to species. : 7 SU eka) Predominating Orders. Number of | Percentage of species. the whole. 1. Compositae. ; : 286 ; 18:05 2. Gramineae. ; 146 peal 3. Liliaceae : : 101 : 6°37 4. Leguminosae . d : 95 : 5°99 22 Annals of the South African Musewm. Number of Percentage of species. the whole. 5. Orchidaceae. : : 89 : 561 6. Scrophulariaceae_ . 86 : 542 7. Cyperaceae . : : 5 : 4°73 8. Asclepiadaceae . : 5 : O21 9. Amaryllidaceae . : 39 : 2°46 10. Iridaceae ; ; : 36 : 2°27 11. Crassulaceae . , ; 35 ‘ 2°20 12. Labiatae , ; : 34 : 2°14 13. Campanulaceae : 31 MOSS) 14, Geraniaceae 29 1°83 15. Selaginaceae 29 1:83 16. Gentianaceae . 21 1-32 17. Umbelliferae . 21 132 18. Cruciferae 20 1:26 19. Ericaceae 20 1:26 20. Rubiaceae 20 1:26 21. Ficoideae 18 L138 The following are the remaining Orders arranged according to the number of species in each: Polygalaceae (15); Anacardiaceae, Boraginaceae, Convolvulaceae (13) ; Rosaceae, Solanaceae (12); Mal- vaceae, Ranunculaceae (11); Polygonaceae, Thymelaeaceae (10) ; Caryophyllaceae, Cucurbitaceae, EHbenaceae, Euphorbiaceae, Sterculi- aceae (9); Acanthaceae, Juncaceae, Santalaceae (8); Onagraceae (7) ; Amarantaceae (6); Chenopodiaceae, Proteaceae, Verbenaceae (5) ; Celastraceae, Dipsaceae, Naiadaceae, Urticaceae (4); Aroideae, Com- melinaceae, Eriocaulaceae, Lentibulariaceae, Loganiaceae, Oleaceae, Papaveraceae, Portulaceae, Tiliaceae (3) ; Dioscoreaceae, Droseraceae, Gesneriaceae, Hypericinaceae, Halogoraceae, Illecebraceae, Lythraceae, Sapindaceae (2); Ampelideae, Apocynaceae, Araliaceae, Bixineae, Capparideae, Hydrocharideae, Lllicineae, Lineae, Menispermaceae, Myricaceae, Myrsinaceae, Nyctaginaceae, Olacinaceae, Passifloraceae, Piperaceae, Pittosporaceae, Phytolaccaceae, Primulaceae, Resedaceae, Restiaceae, Rhamnaceae, Salicaceae, Saxifragaceae, Valerianeae, Xyrideae, Zygophyllaceae (1). The following are the largest genera: Helichrysum (61 species) ; Senecio (56); Crassula (30); Erica, Sutera (18); Disa, Lotononis (17) ; Sebaea (16); Andropogon, Hypoxis (15); Kniphofia, Selago (14); Argyrolobium, Cyperus, Eragrostis, Eulophia, Gladiolus, Schizoglossum, Wahlenbergia (13); Berkheya, Euryops, Habenaria, Indigofera, Pelargonium, Rhus, Zaluzianskya (12). Flora of the Leribe Plateau and Environs. 23 Among the Phanerogams enumerated in the list there are 273 species (17-23 per cent. of the total Phanerogamic flora) endemic. The more important orders, of which more than 20 per cent. of the species are endemic, are as follows : Number of Number of Stet species. endemic species. BEES Selaginaceae : 29 : 12 ‘ 41:3 Ericaceae. : 20 : 6 : 30:0 Campanulaceae . 31 ; 8 : 25°8 Compositae . . 286 . 74 : 25°8 Crassulaceae , 35 : 7) : 25:7 Leguminosae 95 : 24. : 25°2 Gentianaceae E 21 : 5 : 23°8 Liliaceae : lO : 23 : 22:7 Orchidaceae . : 89 : 20 ; 22°4, Scrophulariaceae . 86 ; 19 : 22:0 The more important genera, of which more than 20 per cent. of the species are endemic, are as follows : Gone: Number of Number of Porcentace species. endemic species. Be. Berkheya. : 12 7 58°3 Lotononis . 3 V7 9 52°9 Disperis ; F 10 5 50°0 Euryops : ; 12 6 50°0 Hebenstreitia 3 10 5 50:0 Kniphofia. : 14 7 50°0 Zaluzianskya : 12 6 50:0 Wahlenbergia 13 6 46°1 Stachys : : 3 5 45°5 Alepidea.. A 9 4 44-4, Selago . : ; 14 6 42°8 Erica . : : 18 6 33°3 Sebaea . ; : 16 5 31-2 Crassula : é 30 9 30-0 Helichrysum : 61 18 29°5 Sutera ; : 18 : 5 : 27°7 Senecio ; ‘ 56 : 12 F 21-4 TV. THE KALAHARI FLORA. To arrive at the composition and the affinities of the flora of the Eastern Mountain Region it has been necessary to draw up fresh lists of the species occurring in the Kalahari and South-Eastern Regions, 24 Annals of the South African Museum. as for our purpose the lists given by Bolus are useless for comparison, for the reason, as I have pointed out above, that he included East Griqualand, Basutoland, etc., in his Kalahari Region, and the former also in the South-Eastern Region. These fresh lists have been compiled from the collections in the South African Museum and the Bolus herbaria, and from records given in the ‘Flora Capensis.’ Bolus was conscious of the heterogeneous nature of his Kalahari Region, as he states it ‘‘ will hereafter almost certainly require to be divided into several Regions, or at least to be subdivided into Provinces.” From an examination of collections made in the Transvaal round Houtbosch (and Pietersburg) and Barberton (above 4000 ft.), I also support this view, though I think Bolus was wrong in including Barberton in his proposed Eastern Mountain Region. Unfortunately I have not been able to examine sufficiently large collections from these localities to make any definite statement as to the true affinities of the floras, but from the data brought together I have no hesitation in saying that they cannot be included in the Kalahari Region. This being my view, I have excluded from the Kalahari Region as under- stood in this paper, all species which occur in the neighbourhood of the high mountain range forming the northern contination of the Drakensbergen. The following reasons have led me to adopt this view: 1. The presence of at least 7 species of Hricaceae in these regions, viz., Erica alopecurus, Harv., EH. alticola, G. & B., E. Atherstonei, Diels., Hi. cerinthoides, Linn., EF. drakensbergensis, G. & B., FE. subverticillaris, Diels., and H. Woodii, Bolus. 2. The presence of a large orchid flora. From Barberton 30 species have been recorded, and from Houtbosch and Pietersburg 44 species. The Kalahari Region contains 27 species of Orchidaceae, of which only 7 species are common to either Barberton or Houtbosch. The presence of epiphytic orchids is also significant. Megaclinium Sandersoni, Oliv., Angraecum sacciferum, Lindl., and A. pusillum, Lindl, occur at Barberton: Angraecum sacciferum, Lindl., and A, tricuspe, Bolus, at Houtbosch. 3. The paucity of species of Asclepiadaceae. ‘This important order of the Kalahari Region, which ranks fourth (with 119 species) among the predominant Orders, is as far as my information goes, ouly represented at Barberton by 16 species and at Houtbosch by 12 species. This paucity of species is also very noticeable in the Acanthaceae and Convolvulaceae, both of which are characteristic Orders of the Kalahari Region. It is probable that when the floras of Barberton and Houtbosch are ® Flora of the Leribe Plateau and Environs. better known they will prove to be rich in species which do not occur in the Kalahari Region, as I find from the small collections examined viz., 3382 species from Barberton, and 424 species from Houtbosch and Pietersburg, 49°09 per cent. of the former and 50-47 per cent. of the latter have not been recorded from the Kalahari Region. METEOROLOGY. Taste VI.—Rainfall, Kalahari Region.* Altitude Mean annual fall aise WHTIGTS in feet. for ten years. TOPE Upington : oe 2500) ae 8°67 c 2°24. Prieska . : . 38300 . 11°41 : 3°07 Kimberley : . 4042. 20°44: : 462 Vryburg . : . 4300. 25°50 - 3°85 Mafeking : - 4194. 26°98 5 3°57 Pretoria . ‘ . 4471 ~—~—«. 26°66 : — Johannesbure . 5 66m << 30°26 a, 4°56 Bloemfontein . . 4510. 25°39 : 6°57 * Ex Bolus. Taste VII.—Rainfall during the Months of May—August, of Least Rainfall. Alt. 1901. 1902. 19032 1904, 1905. 1906. 1907. Upington : - 2800 0:00 O18 005 0°20 045 1°46 0°52 Prieska : . 8800 O11 O00 O15 049 1°27 1°86 1°59 Griquatown 2 3560) 0:23) (0:70) (0:50 0218 40:92 ko) ASS Douglas : s 3000) (0:22) 0:79 07349 0:33 E22 4 OAs Kimberley . . 4042 012 098 O95 O54 O'42 O82 1:04 Vryburg . - 43800 045 O67 OY5 O19 O26 O04 1°56 Bloemfontein . 4510 — — 206 159 067 O67 — * Average for five years. Taste VIII.—Temperature, Kalahari Region.* Alt. Number of Absolute metres. years. maximum, Kenhardt . : : 830 6 112°0° F. Prieska . : LOS ; 1 105;8° 9H. Kimberley ; . 1248 : 12 107°2° F. Johannesburg . . 1764 5 96°0° F. Kroonstad . ; . 1384 : 1 ; 105'8° F. Bloemfontein . 5 dletes7/ F 12 : 109:0° F. * Ex Bolus. Mean fall during Mean fall during gix summer months. 6°43 8°34 15°82 21°65 23°41 25°70 18°82 the Period 1908. Mean, O87 0°46 3°35 1°10 1°32 0°87 113 0-96 erin (ORsii 1°35 0:68 261 0-:94* Absolute minimum, 20:0° F. 250° F. 215° 21:0° F. 25:0° F. 16°0° F. These tables should be compared with those under the Eastern Mountain Region. The extreme drought, lasting from May to August, and the greater extremes of temperature are the outstanding differences. 26 Annals of the South African Museum. SYSTEMATIC CONSTITUENTS OF THE FLORA. Orders. Genera. Species. Dicotyledons. . 95 540 : 1752 Monocotyledons : 16 : 129 : 448 Total . dul : 669 : 2195 Proportion of Monocotyledons to Dicotyledons 1: 3°95 Proportion of genera to species ; : lass Predominating Orders. Number of Percentage of species. the whole. 1. Compositae . ; ; 257 : 11-70 2. Leguminosae . : ; 255 : 11-70 3. Gramineae. : ; 140 ' 6°37 4. Asclepiadaceae F . 119 : 5°42 5. Liliaceae : : : 104. . 4-73 6. Scrophulariaceae — . : 95 ; 4°32 7. Acanthaceae . E . 77 ; 3°50 8. Cyperaceae . ; 63 2°87 9. Labiatae ‘ : ; 57 5 2°59 10. Convolvulaceae : : 55 s 2°50 11. Rubiaceze é f : 49 : 2°23 12. Sterculiaceae . 4 ; 49 : 2°23 13. Euphorbiaceae , : 45 : 2°05 14. Inidaceae : : : 42 : 1Oo 15. Ficoideae : : 4 41 ; 1:86 16. Malvaceae : : . 30 : Lesi7 17. Amaryllidaceae : 29 : 133 18. Anacardiaceae , , 28 ‘ 1:27 19. Amarantaceae ; ‘ 28 : 27 20. Orchidaceae. : : 27 : 1:23 21. Verbenaceae . , : 27 : 1:23 22. Tiliaceae : : ; 25 : 1:13 23. Cucurbitaceae : . 24, : 10g 24. Polygonaceae . 22 : 1-00 25. Selagineae : 22 : 1:00 The following are the largest genera: Indigofera (48 species) ; Helichrysum (45); Senecio (36); Ipomoea (31).; Hermannia (29) ; Sutera (27); Acacia (26); Rhynchosia (24); Asclepias, Eragrostis (23); Rhus, Tephrosia (21); Blepharis, Schizoglossum (20); Cro- tallaria, Mesembrianthemum (19) ; Cyperus (18) ; Andropogon (17) ; Flora of the Leribe Plateau and Environs. 27 Aristida (15); Gladiolus, Ocimum (14); Grewia, Lotononis, Ver- nonia (13); Anthericum, Barleria, Convolvulus, Panicum, Sebaea, Selago (12). V. COMPARISON OF THE KALAHARI AND EASTERN MOUNTAIN FLORAS.* 1. Orders occurring in the Kalahari which are not recorded from the Eastern Mountain Region: Nymphaceae, Elatineae, Ochnaceae, Connaraceae, Hamamelideae, Rhizophoraceae, Lemnaceae, Myrtaceae, Loasaceae, Plumbagineae, Pontederiaceae (each represented by one species); Malpighiaceae, Burseraceae (each represented by 2 genera and 3 species); Tune- raceae (2 genera and 2 species) ; Combretaceae (2 genera and 13 species) ; Sapotaceae (3 genera and 5 species); Rutaceae (5 genera and 6 species) ; Pedalineae (5 genera and 9 species); Bignoniaceae (1 genus and 3 species); Plantagineae (2 genera and 2 species) ; Loranthaceae (2 genera and 5 species) ; Haemadoracea (2 genera and 3 species). The most striking feature to be noted is the absence of the Com- bretaceae and the Pedalineae, which are characteristic Kalahari Orders. 2. Orders occurring in the Eastern Mountain Region which are not recorded from the Kalahari : Piperaceae, Restiaceae (1 genus and | species) ; Aroideae (1 genus and 38 species). 3. Orders well represented in the Kalahari, but poorly represented in the Eastern Mountain Region : Kalahari. Eastern Mountain Region. Order. No. of No. of No. of No. of genera, species. genera. species. Capparideae 7 9 : if ; 1 Malvaceae i 30 : 4 a al Sterculiaceae. 6 . 49 2 9 Tiliaceae 3 25 2 3 Zygophyllaceae . 4 8 if it Celastraceae 6 v MES iL 4 Rhamnaceae : 4, : ip il i Ampelideae 5 15 : 1 1 Leguminosae . 51 . 255 eG 95 Lythraceae : A, sea eilul : a 2 Cucurbitaceae. 6 . 24 ; 6 9 * See Fig. 4. 28 Annals of the South African Museum. Kalahari. Eastern Mountain Region. Order. No. of No. of No. of No. of genera, species. genera, species. Ficoideae . a lel 2 41 : A eels Rubiaceae . We at i(5) 5. 49 : 9 ey 20) Asclepiadaceae . 34. 5 LT . 15 5 pl Loganiaceae : 5 8) 2 : 3 A 3 Convolvulaceae . 7. 5 : 13 Acanthaceae a6 Cs 4: 7 8 Verbenaceae : 7 27 4, 5 Amarantaceae . 14 % 28 4 6 Polygonaceae . 4 oars 2 10 Euphorbiaceae . 15 ~ 45 3 9 Urticaceae ; 5 See! 3 4, The Capparideae, 'Tiliaceae, Cucurbitaceae, Convolvulaceae, and Acanthaceae are all characterestic Kalahari Orders, the rest occur commonly in other parts of South Africa, but are better represented in the Kalahari than in the Eastern Mountain Region. 4. Orders well represented in the Eastern Mountain Region, but poorly represented in the Kalahari : Eastern Mountain Region. Kalahari. Orders. No. of No. of No. of No. of genera. species. genera, species. Polygalaceae—. 2 ils 2 2 : 7 Geraniaceae : 5 + 929 : 5 . 18 Crassulaceae . 4 5) : 3 , 2 iG Umbelliferae . Il ral : 8 6 lil Ericaceae . ; 2 . 20 ; 1 : i Orchidaceae es a te : 9 Pee Ah All these Orders are characteristic both of the South-Western and South-Eastern Regions. The large predominance of Ericaceae and Orchidaceae stands out as the striking difference between the Eastern Mountain Region and Kalahari floras. 5. Predominant genera in the Kalahari compared with the same genera in the Hastern Mountain Region : No. of No. of No. of No. of Genera. spp. in Spp. in | Genera. spp. in spp. in ki E.M.R.* | ie E.M.R.* Indigofera . 48 . 12 | Hermannia , ~ 299). 5 Helichrysum. 45 .~ 61 | Sutera. a Re ee 18 Senecio . et ESO. eee OO | Acacia . x it AG Pac ] Tpomoea. GAP ee 8 | Rhynchosia . 24 7 * E.M.R. (Eastern Mountain Region). K. (Kalahari Region). Flora of the Leribe Plateau and Environs. 29 No. of No. of No. of No. of Genera. spp. in spp. in Genera. spp. in spp. in Ke E.M.R.* | Ken E.M.R.* Asclepias . 23 Le eGledwolus. . ida . Is Eragrostis 23 13° Ocimum 5 LAS ws 0 Rhus 21 12 Grewia . Peasy te Os 2 Tephrosia 21 So ultmetenonise . 18... 17 Blepharis eA 2 Vernonia Als} 6 Schizoglossum 20 15 Anthericum . 12. 6 Crotalaria . 19 ik Barleria. qh as | ae Ht Mesembrian- Convolvuius . 12 4 fthemum . 419 = . 9 | Panicum 47 Pe 7 Cyperus . See ells | Sebaea . Pes. LG Andropogon . 17 . 15 | Selago . meee le te eA Aristida «| llsje 8 | * E.M.R. (Eastern Mountain Region). K. (Kalahari Region). The predominance of species of Ipomoea, Hermannia, Acacia, Blepharis, Crotalaria, Ocimum, Grewia, and Barleria in the Kalahari should be specially noted. 6. Predominant genera in the Eastern Mountain Region com- pared with the same genera in the Kalahari: No. of No. of | No. of No. of Genera. spp. in| spp. ae | Genera. spp. in spp. in E.M.R.* | E.M.R.* os Helichrysum. 61. ye Cyperus. eleanor 10s) Senecio . 2) 456 36 Eragrostis ~~ 9 is 97 23 Crassula . 30 7 Eulophia Ror Gs) ae 9 Erica. a 18 1 Gladiolus yt AS. ee eek. Sutera. sis 27 Schizoglossum 13 . 14 Disa. ely. ] Wahlenbergia 138 . 8 Lotononis . 17 13 Berkheya 2» 12 ee 0. Sebaea . : 16 12 Euryops ; 12 : 5 Andropogon . 15 WE Habenaria . 12. . 6 Hypoxis A AS 9 Indigofera . 12 . 48 Kniphofia . 14 3 Pelargonium. 12 =. 7 Selago . . 14 12 Rhus. SP leo Argvrolobium 18 5 Zaluzianskya. 12 0 * E.M.R. (Eastern Mountain Region). K. (Kalahari ReeBay, 7. There are 423 species (or 27°19 per cent.) of the Eastern Mountain Region flora which also occur in the Kalahari, but of this number 317 species (or 20°42 per cent.) are common to the Kalahari, the EKastern Mountain Region, and the South-Hastern Region. 30 Annals of the South African Museum. 8. The proportion of Monocotyledons to Dicotyledons is greater in the Kalahari Region than in the Eastern Mountain Region (viz. 1: 3°28 and 1: 2°03 respectively) though the actual number of species is smaller. VI. THE SOUTH-EASTERN REGION. The South-Eastern Region as defined by Bolus will require some alteration, as he included in this region the whole of Natal and East Griqualand, parts of which belong to the Eastern Mountain Region. Bolus himself recognises this, as he notes “‘ the (boundary) line should keep to a contour of about 3500 ft. of altitude, above which the vege- Fia. 4.—Diagrams showing the differences between the predominance of the twelve largest Orders in the following floras: I. Eastern Mountain Region. II. Kalahari Region. III. “ Kaffrarian Region.” IV. Natal under 4000 ft. 1. Compositae. 2. Gramineae. 3. Liliaceae. 4. Leguminosae. 5. Orchidaceae. 6. Scrophulariaceae. 7. Cyperaceae. 8. Asclepiadaceae. 9. Amaryllidaceae. 10. Iridaceae. 11. Crassulaceae. 12. Labiatae. tation begins to assume the character of the steppes of the Kalahari Region, but no such contour-line is marked on any map.” I am at present attempting to redefine the South-Eastern Region in the light of what we now know of the Eastern Mountain Region. As yet my results are incomplete, though I am led to favour the view of Thode* that this region can be split up into three provinces. Marlotht recognised the diversity of the flora when he mapped off a ‘ South- Eastern Coast Belt.” A flora of East Griqualand, and of King Williams Town, Komgha, Pondoland, Transkei, and Tembuland (the latter group I shall now call the ‘‘ Kaffrarian” flora) has been com- piled for the sake of comparison with the Eastern Mountain Region. The flora of East Griqualand presents problems which make it * Thode, loc. cit. + Marloth, loc. cit. Flora of the Leribe Plateau and Environs. 31 advisable to treat the subject in a separate paper, and for comparison with our Region I will only use the “ Kaffrarian”’ flora and the flora of Natal wnder 4000 ft., above which it assumes a totally different aspect. VII. COMPARISON OF THE FLORAS OF THE EASTERN MOUNTAIN AND SOUTH-EASTERN REGIONS.* Taste I.—The Relative Positions of the Five Predominant Orders. Eastern Mountain “ Kaffrarian ” Natal under Region. Region. 4000 ft. 1. Compositae. | 1. Compositae. 1. Compositae. 2. Gramineae. 2. Leguminosae. 2. Leguminosae. 3. Liliaceae. 3. Liliaceae. 3. Graminae. 4. Leguminosae. 4. Gramineae. 4. Liliaceae. 5. Orchidaceae. 5. Orchidaceae. 5. Orchidaceae. The five predominant Orders are the same in each case, the Eastern Mountain Region differs from the other two floral areas in the important position taken by the grasses and the comparatively few Leguminosae. The Compositae also form a far larger percentage of the total flora than in either the Kaffrarian or Natal areas. Itis only when the floras are examined more critically that the marked differ- ence between the Eastern Mountain and the South-Eastern Regions becomes apparent, as the following table shows: TaBLeE II, riers Ne SN ee Capparideae —. 0 6 : 12 Rutaceae . : : 0 : 10 : 6 Ampelideae 1 13 : 16 A pocynaceae 1 9 1 Gesneriaceae 2 “ 8 Tiliaceae 2 8 13 Oleaceae 3 13 10 Celastraceae 4 34 : 23 Acanthaceae 7 50 f 71 * «K” (Kaffrarian Region). “N” (Natal, under 4000 ft.). E.M.R. (Eastern Mountain Region). All the principal genera occurring in the Eastern Mountain Region are also well represented in the South-Eastern Region with the excep- * See Fig. 4. 32 Annals of the South African Museum. tion of Selago, Eragrostis, Schizoglossum, Euryops, and Zaluzianskya, which are more numerous in the former region. The large number of Cycadaceae in the South-Hastern Region should be specially noted, as no member of the Order has yet been recorded from the Eastern Mountain Region, and it is more than probable that they are totally absent from this region. There are in the Eastern Mountain Region 812 species (or 52°31 percent.) which also occur in the “ Kaffrarian” flora, though of this number 317 species (or 20°4 per cent.) are common to the Kalahari Region, the Eastern Mountain Region, and the “ Kaffrarian ” Region. | VIIl. SUMMARY. The Eastern Mountain Region is a large plateau lying between the 28th and 30th degree of latitude and the 27th and 30th degree of longitude. It includes the whole of Basutoland with a portion of the Orange Free State, the Cape Province, Hast Griqualand, and Natal. On the confines of its boundaries it has an elevation of about 5000 ft., rising to 10,000 ft. and over on the Drakensbergen. The area is built up of Stormberg beds which he on the Karroo beds. The rainfall averages about 29°39 inches per annum, most of which falls during the six summer months. The average maximum temperature is about 82° F., and the average minimum temperature about 33°F. In the greater amount of rainfall and lesser extremes of temperature the Eastern Mountain Region shows considerable differences from the Kalahari Region. The Eastern Mountain Region must be regarded as a grass steppe, and in this respect differs somewhat from the Kalahari and South-Eastern Regions, where the Grasses do not form such a predominant portion of the flora. The paucity of species of Capparideae, Ampelidaceae, Asclepiadaceae, Convolvulaceae, etc., separates it from the Kalahari Region, and the few species of Acanthaceae, Celastraceae, Oleaceae, Rutaceae, etc., which oceur within its limits distinguish it from the South-Eastern Region. Among the genera the number of species of Hrica, Zaluzanskya, Wahlen- bergia, Hulophia, Disa, ete., which occur in the Hastern Mountain Region sharply separate it from the Kalahari Region, while the number of species of Selago, Eragrostis, Schizoglossum, Huryops, and Zalumanskya distinguish it from the South-Hastern Region. Speaking generally the Hastern Mountain Region has an affinity with the South-Hastern Region, but is totally distinct from the Kalahari Region. Flora of the Leribe Plateau and Environs. 33 IX. LIST OF SPECIES OCCURRING IN THE EASTERN MOUNTAIN REGION, WITH DESCRIPTIONS OF NEW SPECIES. DICOTYLEDONS. RANUNCULACEAH. CLEMATIS, Linn. C. brachiata, Thunb. Mountain Slopes. A creeping or climbing plant, 3-9 ft. high, flowers white and sweet scented. Summer. A. Dieterlen, 90! climber on trees in ravine, Phillips, 727! slopes of Lefi’s Kloof, a climber, flowers white, Phillips, 813 ! DistriputTion: Cape. Swellendam. Uitenhage. Albany. Mur- raysburg. Somerset Hast. Komgha. East Griqualand. Natal. Transvaal (Rustenberg). Sesuto: Morarana oa mafehlo. The creeper for the churning sticks. The woody part of the plant, being strong and supple, is used for making churning sticks. THALICTRUM, ILnnn. 1. T. minus, Linn. Dongas and mountain slopes. Flowers green. Summer. A. Dieterlen, 380! Caledon River, Zehyer, 41! Doodman’s Krans Mountain, 8700 ft., March, Galpin, 6568 ! DistRIBUTION : Queenstown. Aliwal North. Griqualand West (near Kimberley). Sesuto: Lefokotsane. The small feeble one. Tord ea ngvale. The hat of the “ngvale.” A ‘“ngvale” is a girl of the initiation school. The plant is used as a vegetable in the young state. 2. T. rhynchocarpum, D. & R. De Beer’s Pass, Drakensbergen, 5—6000 ft., Wood, 5971! Disrripution: Komgha. East Griqualand. Transvaal (Pilgrim’s Rest). ANEMONE, Linn. A. Fanninii, Harv. Eland’s River Valley, near the Mont-aux-Sources, 6000 ft., Flanagan, 2126! Distrisution: Natal. ey) Br Annals of the South African Musewm. RANUNCULUS, Lnnn. R. pinnatus, Poir. Damp spots on veld and mountain slopes. Herb, 5-18 in. high, flowers bright vellow. Summer—Autumn. DistRiBuTion : Endemic ? Sesuto: Sethokoyane. Meaning unknown. One of the first spring plants, and very much liked by the natives as a vegetable. S. oleraceus, Linn. Mountain slopes and river banks. 6-86 in. high, flowers yellow. Summer—Autumn. Mametsana. Mother of the small water (cited in the * Flora Capensis ” as a locality). S. aphyllum, Schltr. Bester’s Vlei, near Harrismith, January, Bolus. Distrigution: Albany. Queenstown. Cradock. Tembuland. East Griqualand. Natal. Zululand. Transvaal (Donkerhvek). S. neglectum, Schltr. Harrismith, Sankey, 271; Natal, Oliver’s Hoek, sources of the Tngela River, 5000 ft., Allison; slopes of the Drakensbergen, Wood, 3418. DistrisutTion: Transkei. East Griqualand. Transvaal (Hout- bosch, Barberton). S. longicauda, Lindl. Veld, mountain slopes and plateau. 12-20 in. high, flowers white or pink.. Summer. A. Dieterlen, 244! 485!; Harrismith, Sankey, 268, 269; Natal, summit Amawahqua Mtn., 6—7000 ft., April, flowers a delicate pink, Wood, 4617!; Oliver’s Hoek, 5000 ft., Allison, Q.U.; south slopes of plateau, flowers pinkish, Phillips, 684!, flowers pale pink, dark above, Phillips, 865!; Bester’s Vlei, near Witzie’s Hoek, 5900 ft., December—January, Bolus, 13482 ! Distripution: Albany. Stockenstroom. Queenstown. Komgha. Tembuland. Pondoland. East Griqualand. Transvaal (Barberton, Lydenburg, Belfast, near Donkerhoek). Sesuto: Lekoesha. See Hulophia flaccida. ’Mametsana. Mother of the small water. (Cited in the “ Flora Capensis ” as a locality). 8. cristatum, Sond. Mountain slopes and plateau. 12-24 in. high, flowers pink, or whitish and crimson. Summer. A. Dieterlen, 636a! 636b!; Natal, Oliver's Hoek, 4000 ft., Allison ; near Van Reenen, 5—6000 ft., Wood, 5533. Distrisution: Port Elizabeth. Bedford. Stockenstroom. Queens- town. Somerset East. Tembuland. Pondoland. East Griqualand. Orange Free State (no locality). Transvaal (Houtbosch, Lydenburg, Barberton, Belfast, Ermelo, Middleburg). 270 Annals of the South African Museum. SesuTo: “Mametsana. Mother of the small water. Lekhdlela la Baseng. The “ Lekholela’’ of the Baseng (the name of a Basuto clan). S. ocellatum, Bolus. Mountain slopes and plateau. 12-20 in. high, flowers yellow. Summer. A. Dieterlen, 486!; Harrismith, Sankey, 270; Natal, Oliver’s Hoek, 4—5000 ft., Allison, C., 29. DistrrpuTion: Fort Beaufort. Stockenstroom. Komgha. East Griqualand. Transvaal (Lydenburg, Houtbosch, Pretoria, Ermelo, Barberton). Sesuto: ’Mametsana. Wrongly cited in the ‘ Flora Capensis’ as a locality. MONADENIA, Lindl. M. Basutorum, Rolfe. Basutoland. Heathy summits of the Drakensberg Range, 10000 ft., Thode. DistRIBUTION: Endemic. HERSCHELIA, Lindl. H. Baurii, Kranz. Natal, Bushman’s River Valley, 6-8000 ft., flowers bluish-purple, Wood, 10599. Distripution: Tembuland. Transvaal (Houtbosch, Barberton, Middleburg. ) DISA, Berg. D. stachyoides, Reichb., f. In the valley of the Eland’s River, near the Mont-aux-Sources, 6—7000 ft., January, Flanagan, 1983!; Harrismith, Sankey, 262; Natal, Drakensberg Range, 5—-6000 ft., Wood, 5146. Distrisution: Tembuland. East Griqualand. Natal. Transvaal (Lydenburg, Barberton, Elandspruit Mtns., Houtbosch). D. Sankeyi, Rolfe. Fikilemutu Pass, Drakensberg, Matatiele Dist., 7500 ft., February, McLoughlin, 204! ; Harrismith, Sankey, 264. Distripution: Endemic. D. fragrans, Schltr. Grassy slopes of Quaqua Mtns. and Mopedi’s Peak, Witzie’s Hoek, 6800-8100 ft., March, flowers pale purple or whitish, sweet-scented. Flora of the Leribe Plateau and Environs. 271 Thode, 55; Mont-aux-Sources, 8-9000 ft., Thode; Harrismith, Krook, Penther, 124; Tsitsa footpath, Dist. Maclear, c. 7550 ft., March, flowers very pale pink with purple spots, leaves blotched with purple, Galpin, 6837!; Natal, Van Reenen’s Pass, March, Penther, 86!; summit of Amawahqua Mtn., 6800 ft., Wood, 4565! DistrisuTion : East Griqualand. Transvaal (Houtbosch). D. polygonoides, Lindl. Mountain slopes and plateau. 10-20 in. high, flowers red and pink. Summer. ? i.e. longer than in the type. 0. graminifolium, Thunb. Damp spots on mountain slopes. 10-25 in. high, flowers white. Spring. A. Dieterlen, 729! DistrrpuTron: Cape. Paarl. Tulbagh. Murraysburg. Komegha. Natal. Sesuto: Metsane a manyenyane. Meaning unknown. Nko ea ntja. The nose of a dog. 0. Eckloni, Schltr. Mountain slopes. 10-20 in. high, flowers greenish white. Summer. A. Dieterlen, 420! Distrisution: British Kaffraria. Somerset East. Tembuland. East Griqualand. Natal. Griqualand West. Transvaal (Aapie’s River, Barberton). 312 Annals of the South African Museum. Srsuro: Moretele 0 monyenyane. The “small one” who causes to glide. Used as Urginea capitata but not such a powerful charm. By means of this charm an enemy can cause one’s cows to miscarry, or dry up the milk of a suckling cow. 0. capillaris, Wood and Evans. Between rocks, Bester’s Vlei, near Witzie’s Hoek, 5400 ft., January, Bolus, 8261! Disrrisution: Transkei. Natal (Newcastle). ANDROCYMBIUM, Willd. A. melanthioides, Willd. Plateau. 3-10 in. long, flowers white. Summer. A. Dieterlen, 659!: Natal, under the Drakensberg, Keit, 6. DistripuTion: Worcester (Matjesfontein). Albany. Somerset East. Murraysburg. Victoria East. Graaff Reinet. Swaziland. Transvaal (Barberton, Rustenburg, Houtbosch, Johannesburg, Lyden- burg). Griqualand West. Sesuto: Khukhoana e nyenyane. The tiny “Khukhu” (bulb). Mixed with Polygala spp. it is used in the cure of many diseases. A. leucanthum, Willd. Basutoland, without precise locality, Cooper. Disrripution : Namaqualand Minor. Cape. Knysna. Uitenhage. A. natalense, Bkr. Summit of the Mont-aux-Sources, 9800 ft., January, Flanagan, 2026! Disrripution: Natal. A. longipes, Bkr. Veld and mountain slopes. 6-14 in. high, flowers mauve-pink. Autumn. A. Drelerlen, 516! Disrripution: Albany. Murraysburg. Somerset Kast. Komgha. Srsuro: Khukhoana. The small “ Khukhu.” The dried bulbs are erushed and mixed with fat as an ointment for sore ears. WURMBEA, Thunb. W. capensis, Thunb. Wittebergen, 7-8000 ft., Drege, 3512; Ben McDhui, District Barkly East, c. 9800 ft., March, Galpin, 6870 ! Distrisution : Malmesbury. Cape. Riversdale. Queenstown. Stutterheim. Somerset East. Graaff Reinet. East Griqualand. Flora of the Leribe Plateau and Environs. 315 W. Kraussii, Bkr. Summit of the Mont-aux-Sources, 9500 ft., January, Flanagan, 2027!; Natal, Van Reenen, 5—6000 ft., February, flowers greenish- white. Wood, 9617! Distripution: Tembuland. Pondoland. East Griqualand. W. pusilla, Phillips, sp. nov. Planta pusilla, ad 8 em. alta. Folia 2, 2-4-1 em. longa, 3-8 mm. lata, ovata vel linearia, acuminata, glabra. Inflorescentia spicata, 2-4-florifera, 15-2 em. longa. Tubus perianthii 75-8 mm. longus, 2-2°5 mm. latus, glaber ; lobi 7°5-8 mm. longi, 1:75 mm. lati, lineari-lanceolati, apice obtusi. Filamenta 3 mm. longa, glabra; antherae 1 mm. longae, oblongae. Ovariwm 4°5 mm. longum, 2°5 mm. latum, oblongum, glabrum ; stylus 5-5 mm. longus, basi ‘75-1 mm. latus, apice subulatus; stigma minutum. A small plant up to 8 em. long. Corm not seen. Produced leaves 2, 2-41 em. long, 3-8 mm. broad, ovate or linear, acuminate, clasping the stem, glabrous; leaf-sheath 7-9 min. long in the lower leaf, 1-8-2 cm. long in the upper. Inflorescence spicate, 2—4-flowered, 1-5-2 em. long. Tube of the perianth 7-5-8 mm. long, 2-2'5 mm. broad, glabrous ; lobes 75-8 mm. long, 1:75 mm. broad, linear-lanceolate, obtuse, with two dark glandular swellings 2 mm. from the base, glabrous. Stamens inserted at the base of the perianth-segments; filaments 3 mm. long, lear glabrous ; anthers 1 mm. long, oblong. Ovary 4°5 mm. long, 2°5 mm. broad, oblong in outline, glabrous; styles 3, 5°5 mm. long, ‘75-1 mm. broad at the base, tapering gradually to a subulate apex ; stigmas minute. Mont-aux-Sources, 10000 ft., October, flowers pale, A. Bolus in Herb. Guthrie, 4937. Differs from W. capensis, Thunb., in the dwarf habit and the short 1-4-flowered inflorescence, sheathed in the upper leaf. LITTONIA, Hook. L. modesta, Hook. Damp places at Bester’s Vlei, near Witzie’s Hoek, 5300 ft., Decem- ber, Bolus, 8265!; Nelson’s Kop, Cooper, 880. Disrrrpurion: Pondoland. Natal. ‘Transvaal (Barberton, Lyden- burg, Belfast). XY RIDEAE. XYRIS, Linn. X. capensis, Thunb. Marshy places at Leribe, A. Dieterlen, 602!; damp ground near stream on plateau, flowers yellow, Ph illips, 718 ! 314 Annals of the South African Museum. Distrisution: Cape. Tulbagh. Ceres. Queenstown. Somerset East. Tembuland. Natal. Transvaal (Ma zaliesberg, Rustenberg). COMMELINACEAE. COMMELINA, Linn. C. benghalensis, Linn. Shaded and sheltered spots on veld and mountain slopes. A creeper, 10-40 in long, flowers mauve or blue. Summer. A. Galpinii, Stapf. AN 0 <5 eg VAZOPE NON Ben MeDhui, 9400 ft., Galpin, 6900. i), CO now (No DistriIBuTION : Endemic. f ~) eg wel 348 Annals of the South African Museum. A. bergiana, Trin. forma rigidula. Summit of Doodman’s Krans Mountain, 9650 ft., March, Galpin, 6913. DISTRIBUTION: Endemic. PEROTIS, Azé. P. latifolia, Ait. Van Reenen, 5—6000 ft., March, Wood, 5990. Distripution: Transvaal (Klippan). TRAGUS, Haller. T. racemosus, All. Mountain slopes. 6-15 in. high, spikes greyish. Summer. A. Dieterlen, 240! Distrispution: Uitenhage. Albany. Queenstown. Wodehouse. Graaff Reinet. Albert. Natal. Griqualand West. Orange Free State (near Boshof). Sesuto: Mohloa-tsépe. Iron “mohla.” Meaning unknown. Mabele a litsoere. Sorghum of the Cape canary. Good grazing grass for cattle and goats. SPOROBOLUS, Rk. Br. S. discosporus, Nees. Plateau. 2-6 in. high, spikes cream and brown. Summer. Not common. A. LOTONONIS TRISEGMENTATA, Phillips. LE. P,P. del. e sicco. B. WAHLENBERGIA DEPRESSA, W. & c. Lopenia aquatica, Phillips. Adlard 4 Son g West Newman, Ltd. Wi Plate Vol. XVI. Mus. Ann. S. Afr. Euryoprs ANNAB, Phillips ny West Newman, Adlard & So del, e sicco A. C, Dawson Ann. S. Afr. Mus. Vol. XVI. Plate V. 3. xe 4, x 1. HettopHina Basurica, Phill. 2. PSAMMOTROPHA ANDROSACEHA, Fenzl. 3. WAHLENBERGIA BASUTICA, Phill. 4. W. pepressa, W. & E. forma. Adlard 4 Son 4& West Newman, Ltd. Ann. S Afr. Mus Vol XVI. Plate VI. 1 x al 9 x ql } 3. x 3 a. x : 1. HELICHRYSUM CALOCEPHALUM, Schltr. 2. H. rapDnil, Sp. Moore. 3. CREPIS POLYODON, Phill. 4, H. CHIONOSPHAERUM, D.C. Adlard 4 Son & West Newman, Ltd. pon rs (ted eo i: iD i. TT ae ‘ ar a " ' 7) af - om : my ane t's ‘ ( a i ® ‘ aL a i> es ? © 4 4 q i) ' , } 7 i . {> 7 ; ee LY : AWN i p ~ \ ‘ , 7 ex's i » bah vt j ‘ ‘ 4 é - pt a aa, ’ { Wax : 4 a J yi ‘ig ety” ¥ y i es, ™ é ; 4 - ai f, i. a é s © : . g om ~ : A hs ‘ ' 4. ; i KX # ' a oe: ae bs x + er 7 ‘ — . a. « §— ie ra ’ aD a0 ” | : -_ _ _ : o = 5 _ .¢ | : b= zi a has - = | ae 7 7 7 . _ = a> a a) - 7 - i< > oY io : = we. =! — = « _ - 7 = - Om ‘ —! a9 a Pes oe 7) es 7 a . {= ‘ ¥ > . t yy . - - : , et > a ‘ - - 2 i: aay ' I 1 + a a - a a = a hy b . oe os at u 3. Observations on the Genus Volvox in Africa.—By Fuorence Ricn, Botanical Department, East London College, University of London, and Mary Acarp Pocock, Botanical Department, University of Cape Town. (With Plates IX-XXIV and 6 Text-figures.) CONTENTS, Introduction : A. Volvox aureus Ehrenb. 1. From Potchefstroom. 2. From Weltevreden West. 3. From N’gamo and Old N’gamo, and from the Linyanti River. B. Volvox africanus G. S. West From the Linyanti River C. Volvox capensis sp. nov. 1. From the Cape Flats. Forma rhodesiensis f. nov. (a) From N’gamo and Old N’gamo. (6) From the Linyanti River. D. Volvox Rousseletii G. S. West 1. From the Ussangu Desert. . From Egypt. . From Rietfontein, Transvaal. . From Eliazer Pan and Brakpan, Transvaal. . From Sedudu, Bechuanaland. 6. From Grahamstown. Forma nov. kaokoensis from Kamanyab, Kaokoveld. E. Volvox amboensis sp. nov. . From Ongka, Ovamboland. Summary Conclusion Explanation of Plates Or em Ww bO INTRODUCTION, PAGE 427 439 440 441 Ar the time when, in 1905, Mr. C. F. Rousselet stepped from one of the trains conveying members of the British Association across Rhodesia and collected Volvoz from a pool by the side of the railway- line, very little was known about the occurrence of this genus on the VOL EexXcVib PART 3: 428 Annals of the South African Museum. continent of Africa. Volvor aureus Ehrenb. had been noticed by Schmidle * in material collected in 1899 near the mouth of the Mbasi River, but other records were very scanty. Mr. Rousselet’s sample was submitted to the late Professor West + for investigation, and, though no sexual organs were present, the characters of the vegeta- tive colonies were considered by him amply sufficient to warrant the establishment of a new species to which he gave the name Volvoz Rousseletii. A few years after this, some plankton material collected by Dr. Jakubski of Lemberg (Lwow) from small, temporary pools in the Ussangu Desert {| was found to contain sexual colonies which West stated to belong to this same species, and from an examination of these he was able to complete the diagnosis.§ In 1907 a curious form of Volvox was obtained by Mr. R. T. Leiper from the Albert Nyanza; this was at first described by Professor West || as “ Volvox aureus Ehrenb. (a form),”’ but he subsequently observed characters scarcely in keeping with those of V. aureus, and decided to regard it as a “ new species or new race”’ of Volvox for which he suggested the name V. africanus. It is noteworthy that no trace of this particular Volvox had been seen in the other large lakes of Central Africa, even though they had been more thoroughly investigated than the Albert Nyanza. (It was present, associated with V. Rousseletii, in the material examined by West from the Ussangu Desert.) In 1914, amongst some freshwater algae from Madagascar, Professor F. E. Fritsch ** found a form of V. aureus which he styled forma madagascariensis, and in the same year Brunnthaler ++ recorded V. aureus from Egypt. The latter species, under the name of Volvox minor Stein, was also recorded by Madame Gauthier-Liévre tf from Algeria in 1925. There was no reason why such a commonly occurring genus should not have further representa- tives in Africa, yet, as far as we can ascertain, there are no records of * Schmidle, W., ‘‘ Algen, insbesondere solche des Plankton, aus dem Nyassa- See und seiner Umgebung,” Engler’s Bot. Jahrb., vol. xxxii, 1902, p. 77. + West, G. S., Journ. Quek. Micr. Club, vol. xi, ser. 2, 1910, pp. 99-104. { In what was formerly German East Africa, and is now Tanganyika. § West, G. S., Journ. Quek. Micr. Club, vol. xiii, 1918, pp. 425-428. || West, G. S., Journ. Bot., vol. xlvii, 1909, p. 245. 4] West, G. S., Journ. Quek. Micr. Club, November 1910, pp. 102-103. ** Fritsch, F. E., Ann. de Biologie lacustre, vol. vii, 1914. ++ Brunnthaler, J., Beitr. z. Siisswasseralgenflora von Agypten. Hedwigia, vol. liv, ps 219. tt Gauthier-Liévre, ‘“‘ Quelques observations sur la flore algale de l Algérie dans ses rapports avec le pH,” 1925, C. R. Ac. Sc., vol. clxxxi, p. 927. Observations on the Genus Volvox in Africa. 429 any (De Toni in his Sylloge Algarum, 1889, gives no references) ; there are certainly none in the collections examined by the following :— RaBENHORST, L., Beitr. z. Kryptogamenfl. Siidafrik., Pilze u. Algen, Allg. deutsch. naturh. Zeit., vol. 1, 1855. Reinscu, P., “Contr. ad fl. Alg. aq. dulce. Promont. Bon Spei,” Journ. Linn. Soc., Bot., vol. xvi, 1877. Norpstept, O., Lunds Univ. Arsskrift, vol. xvi, 1880, “‘ De Algis et Characeis I ”’ (Senegal). Wute, N., * Uber einige von J. Menyhardt in Siidafrika gesammelte Siisswasseralgen,”’ Oesterr. Bot. Zeitschr., 1903. West, W. and G. &S., “ Algae from Central Africa,” 1896, Journ. Bots, vol) xxxiv: OsTENFELD, C. H., “ Phytoplankton aus dem Victoria Nyanza,”’ 1908, Engl. Bot. Jahrb., Bd. 41. West, G. S., Ann. 8. Afr. Mus., vol. ix, 1912 (South-West Africa). Fritscu, F. E., Ann. 8. Afr. Mus., vol. ix, pt. 7, 1917 (mostly from the Cape Peninsula). Fritscu, F. E., and Stepuens, E., Trans. Roy. Soc. 8. Afr., vol. ix, pt. 1, 1921 (the Transkei Territories). Fritscu, F. E., and Ricu, F., Trans. Roy. Soc. 8S. Afr., vol. xi, pt. 4, 1924 (Natal). Hopegetis, W. J., Trans. Roy. Soc. 8._Afr., vol. xii, pt. 1, 1925 (Stellenbosch). Fritscu, F. K., and Rica, F., Trans. Roy. Soc. 8. Afr., vol. xviii, pt. 1, 1929 (Griqualand West). HusBer-Pestatozzi, G., Verh. Internat. Ver. Limnol., vol. iv, 1929 (South African plankton). It was therefore of special interest when the present authors had the opportunity of examining Volvox material collected in 1926 from pools on the Cape Flats * by one of themselves and Miss E. L. Stephens (of the University of Cape Town), and also some from Ovamboland collected by Dr. K. H. Barnard and Dr. R. F. Lawrence (of the South African Museum). Two years later, that is, in 1928, Mrs. G. E. Hutchinson, who with her husband was engaged in making a survey of the planktonic fauna of South African vleis and pans, found Volvoz in the Transvaal in five different pieces of water, namely, at Riet- fontein, Brakpan, Leeuwkraal, Weltevreden West, and Potchefstroom * We find that Mr. Steer, Miss Stephens, and others had collected Volvox regularly from the Cape Flats and Lakeside since about 1916, and that Dr. Barnard had already found some in Ovamboland in 1923. 430 Annals of the South African Museum. (Eliazer Pan). In July 1930 one of us and Miss Stephens collected Volvox from a large vlei at Old N’gamo, from pools at N’gamo, from quiet backwaters of the Linyanti (Chobe) River, and from pans and vleis left on the south side of this river from the previous wet season. All these localities, with the exception of the Linyanti backwaters, are temporary pools * (usually known locally as vleis), the majority of them grass-bottomed, 7.e. forming grassy meadowland when they dry up in summer. They occur on soil which is more or less alluvial in character, consisting usually of a mixture of sand and alluvium. The paper that follows deals with the results of an investigation of the foregoing samples, together with one from Egypt collected by Mr. Douglas Simpson in 1924. It is apparent that Volvoz flourishes abundantly in Africa, and its absence from earlier collections is a little difficult to understand. We, like many of our predecessors, realise that considerable varia- tions may occur in one and the same species, not only in the number of constituent cells, in the nature of these cells, in the number of daughter colonies, and in sexuality, but also in the structure of the oospore. There is some doubt as to what should be considered diagnostic features, and there is also some uncertainty as to the number of species of Volvox. The structure of the coenobium has been fully described only in the three European species V. aureus Ehrenberg, V. globator (Linné) Ehrenberg, and JV. tertiws A. Meyer. The two former are the commonest and best known, and differ from one another as follows :— 1. The coenobia of V. aureus are usually smaller than those of V. globator (the latter measure generally 600 to 800 yp). 2. In V. aureus the number of constituent cells is from 200 to 4400 ; in V. globator from 1000 to 15,000. 3. The number of daughter colonies in the asexual coenobia varies from 4 to 14 in V. aureus; in V. globator the number is usually 8 (though it may be more or less). 4. The protoplasts of the cells of the daughter colonies are separated * This is contrary to the experience of Professor Iyengar of Madras, who says that in Southern India Volvow has never been collected from temporary rain-water pools (he adds that the trampling of animals tends to make the water of these pools very muddy), but it agrees with the description given by J. H. Powers (‘“‘ Further Studies in Volvow,”’ Trans. Amer. Micr. Soc., vol. xxviii, 1908, p. 141) of the occurrence of Volvox in parts of America (particularly Nebraska), where Volvox was collected mostly from small rain-pools of scarcely a fortnight’s duration, while ponds 1 to 2 ft. deep yielded no Volvox. The forms of Volvox collected in Nebraska are very like the South African representatives. Observations on the Genus Volvox in Africa. 431 from one another in V. aureus before escape ; in V. globator they are still in contact at this stage. 5. The diameter of the daughter colonies on escape is from 200 to 250 p in V. aureus ; 150 to 200 p (rarely to 320 yw) in V. globator. 6. The protoplasts of the somatic cells in V. aureus are almost round in surface view, more oval in side view ; in V. globator they are smaller, more closely packed, rather angular in surface view, and flattened basally in side view. 7. The protoplasmic processes are very delicate in V. aureus; in V. globator they are thicker, and continuous with processes from the angles of the protoplasts (the chloroplasts extend into these processes). 8. The sexual colonies in V. aureus may be exclusively male or female, i.e. dioecious, or they may be monoecious, and may or may not contain daughter colonies as well; in V. globator they are almost invariably monoecious (7.e. they contain both antheridia and oogonia), and do not, as a rule, contain daughters. 9. The antheridia in V. aureus are numerous (as many as two- thirds of all the cells may form sperm bundles) ; in V. globator there are only from about 5 (or fewer) to 15. 10. The sperm bundles measure from 12 to 18 w in diameter in V. aureus, from 23 to 34 win V. globator. 11. Each sperm bundle is composed of 16 or 32 spermatozoids in V. aureus, from 64 to 256 or more in JV. globator. 12. The spermatozoids in V. aureus have a spherical nucleus and terminal cilia ; in V. globator they are more slender, with an elongated nucleus, and the cilia may be either laterally or terminally attached. 13. The oospores in V. aureus number from 1 to 15 (usually 3 to 8) ; in V. globator they number from 12 to 69 (usually about 30). 14. The oospore in V. aureus measures from 60 to 75 yu in diameter, and its outer wall is smooth ; in V. globator it measures from 44 to 56 p, and its outer membrane, when ripe, is covered with conical spines. 15. A still more fundamental difference between the two was demonstrated by Meyer * during an investigation into the structure of the membrane of the coenobium. In V. aureus the somatic proto- plast is embedded in a thick gelatinous wedge-shaped mass extending far into the interior of the colony, where the radial walls join a central massive star-shaped body (see fig. 1, A:); in J. globator it is enclosed in a short five- or six-sided gelatinous prism tapering very slightly towards the interior of the coenobium from which it is delimited * Meyer, A., ‘‘ Die Plasmaverbindungen und die Membranen von Volvox, etc.,” Bot. Zeit., vol. liv, 1896, p. 187. Fic. 1.—Volvox membranes, stained with methylene blue. A, V. aureus. B, V. africanus. C, V. capensis sp. nov. and V. Rousseletit. In each: 1, Optical section of whole colony. 2, A small portion further enlarged. 3, Cells in surface view. In V. aureus the central star-shaped mass stains deep purple; in V. cupensis the mucilaginous content of the hollow sphere stains slightly lighter purple. In V. africanus the inner membrane is regular and clearly defined when the colony is young, later becomes displaced and distorted by the developing daughter colonies. A,, B,, C; x60; A, x 100; others x 500 (roughly). Observations on the Genus Volvox in Africa. 433 by a flat wall, and the cavity of the coenobium within the layer of cells is occupied, according to Meyer, by a watery fluid. It will thus be seen that the differences between these two are greater than is usually the case between two species of the same genus, and it is not surprising that the suggestion has been made that they should be separated. In 1922 Shaw * placed V. aureus in a new genus which he named Janeto- sphaera, retaining V. globator in the genus Volvox, while he placed V. tertius, in which there are no protoplasmic connections between the protoplasts, in another distinct genus Merrillosphaera. Of these, Volvox is represented in Africa by several species, of which V. Rousseletii is widespread ; the present paper describes further species and forms belonging to this section. Janetosphaera is represented by J. aurea (=V. aureus), differing in no marked degree from that found in Europe, while V. africanus belongs to the third section (= Merrillo- sphaera africana, according to Shaw). None of the recently published comprehensive accounts, however, has adopted Shaw’s nomenclature, hence the present writers are not using it. All who have worked at the group agree that a revision is necessary, cf. Pascher f and Printz.t The latter divides the genus Volvox into subsections, to which he applies Shaw’s generic names. G. 8. West considered that V. africanus was the African species comparable with the European V. aureus, and that V. Rousseletii was similarly comparable with V. globator. The present authors feel that the resemblance between the two latter is greater than that between V. africanus and V. aureus. Moreover, V. aureus itself occurs extensively in Africa. V. africanus is more nearly com- parable with the European species V. tertiws Meyer, its membranes being similarly constructed to those of the latter (see text-fig. 1, B). The features on which West§ based his two African species may be usefully enumerated here as follows :— Volvox africanus : 1. The coenobia, though approximately the same size as those of V. aureus, are of adifferent shape, being constantly ovoid or egg-shaped. 2. The constituent cells number from 3000 to 8000. 3. The daughter colonies vary in number from 1 to 4 (3 isa common number). * Shaw, W. R., The Philippine Journ. of Sci., vol. xx, 1922, p. 478. 7 Pascher in Die Siisswasserflora Deutschlands, Osterreichs und der Schweiz, vol. iv, 1927, p. 462. { Printz, Volvocaceae in Engler’s Nat. Pfl. fam. Ed. II, vol. iii, 1927, p. 59. § West, G.S., op. cit., 1918, p. 426. 434 Annals of the South African Museum. 4. They grow until they almost fill the internal cavity of the parent, becoming flattened by compression; they themselves show well- developed daughter colonies long before their escape from the mother colony, and even a fourth generation is not uncommon. 5. The somatic protoplasts are almost globose, with long, delicate protoplasmic strands (Shaw,* however, disputes the presence of these strands, and in his earlier account West } states that no protoplasmic connections could be observed). 6. The number of oospores averages 74 (70 to 80); wall of oospore thick and smooth. The diagnosis of this species was amplified by Shaw from material collected near Manila in July 1914 and subsequently. He found no protoplasmic connections between the cells; in his material there were sometimes as many as 6 or 8 daughter colonies, the oospores varied from 12 to 43 per coenobium, measuring 43 to 45 w when ripe. The oospores were sometimes associated with from 2 to 6 sperm plates, but more usually the latter were found in colonies producing no other reproductive cells (100 to 400 per coenobium). He was not able to count the number of sperms—evidently most of his antheridia were still young—but thought it was either 64 or 128 per platelet ; he gives the dimensions of one antheridium in the 32-celled stage (26 yz) and of one mature sperm plate (7 uw thick by 20 w wide). The antheridium mother-cell often reached 19 yx before dividing to form a platelet or hollow cup of sperms. According to West this African species differs from V. aureus in the ovoid-ellipsoid shape of its asexual colonies, in the larger number of constituent cells, and in the nature of the daughter colonies. Volvox Rousseletii is described by West as follows :— 1. The asexual coenobia are large, having a diameter of from 1125 to 1240 » when fully mature. 2. The number of constituent cells is large, 25,000 to 50,000, and they are rather closely aggregated. 3. The somatic protoplasts are somewhat angular, with relatively broad connecting strands. 4. Daughter colonies regularly 8. 5. The sexual colonies are dioecious. 6. Sperm bundles are numerous. * Shaw, W. R., loc. cit., vol. xxi, 1922, p. 118; vol. xxu, 11928) pp: 189 and 197. t West, G.S., loc. cit., 1910, p. 103. Observations on the Genus Volvox in Africa. 435 7. The ripe oospores number from 120 to 150; they show a decided concentration towards one pole of the coenobium. 8. The wall of the oospore is clothed with strong conical spines ; average diameter without spines 44 pw, length of spines 11 to 12 p. The foregoing description of V. Rousseletii is somewhat incomplete ; for instance, the sperm bundles, though shown in the microphotographs published by West, are not described by him. West considered that V. Rousseletii differed from V. globator in its larger colonies and the much greater number and closer aggregation of the constituent cells, in the sperm bundles being far more numerous, and the oospores about four times as numerous and showing a con- centration towards one pole; * he also adds that the long conical spines on the oospore “‘ are very different in appearance from the depressed conical warts on the oospores of V. globator.” + The question arises as to how far the peculiarities just enumerated are to be looked upon as of specific importance, and whether they may not, after all, be due to differences in nutrition, insolation, temperature, and so on, but it is not the purpose of the present writers to discuss these points here. We could not help feeling some doubt as to whether the sexual colonies obtained from the Ussangu Desert did really belong to the same species as the asexual colonies collected in Southern Rhodesia, and we considered that if living material could be obtained from the * The present authors, however, find a concentration of the oospores towards one end in all the coenobia of V. globator that they have been able to examine. + Surely this is a rather misleading description of the oospores of V. globator. It is possible that more than one race has been ascribed to this species, but in the figures given by Janet (Le Volvox, Deuxieme Mémoire, 1922, pl. iii, fig. 25, and Troisitme Mémoire, 1923, pl. xx, fig. 4) the spines are shown as bluntly pointed cones, neither depressed nor wart-like. Shaw, in his key to the species of Volvox (loc. cit., May 1922, p. 503), describes the wall of the oospore as “ angularly wavy ” ; yet in another place (p. 486) he speaks of the oospore as having “ a stellate prickly outer membrane ’’—two descriptions which are not in accord with one another. It is possible that observations may have been made on immature specimens, for all spiny oospores show a wavy outline before the spines are fully grown (see fig. 4, C and D). Klein (Vergleichende Untersuchungen iiber Morph. u. Biol. der Fortpfl. bei der Gattung Volvox, Ber. d. deutsch. Naturf. Ges. zu Freiburg i. B., 1890, p. 84) described the exine of the oospore as spiny (stachelartig). In material from Baden-Baden kindly sent us by Mr. Scourfield are colonies containing 12 to 16 ripe and nearly ripe oospores; these have a well-developed spiny exine, the conical spines are from 5 to 8 yw long, the oospores, including spines, having a diameter of 49 to 55 wy. 436 Annals of the South African Museum. identical pool in which Mr. Rousselet originally found it, very satis- factory further information concerning it might be obtained. Conse- quently an attempt was made by Miss EK. L. Stephens in 1927 to locate the spot indicated by Rousselet in his paper to the Royal Microscopical Society in 1906,* and as a result of her inquiries she came to the conclusion that Rousselet had made his collection at Old N’gamo, 5 miles north of the present N’gamo station and 23 miles north of Gwaai. She visited the vlei, which is an extensive one several acres in area in the wet season, with the railway on an embankment built across it, and collected Volvor. An accident unfortunately happened to the material and all the Volvor perished before it could be examined. In July 1930, however, Miss Stephens and one of us were able to revisit the vlei and we found Volvor swarm- ing in parts of it, chiefly in the deep part alongside the embankment. It was particularly abundant in the partially shaded water in a culvert under the line connecting the two parts of the vlei; this part is prob- ably the last to dry up, and it may have been here that Mr. Rousselet collected. On this occasion the dominant form was V. aureus, but another form, very like V. Rousseletiz but still more like the form found on the Cape Flats, was present as well, in all stages both sexual and asexual. But subsequently we found another more detailed account by Rousselet,f made several years later, in which he states explicitly that he collected the Volvor at Gwaai station: “the train stopped for half an hour at this station by the side of a shallow pool formed by the Gwaai River.” Thus his two accounts are contradictory. It has not yet been possible to revisit Gwaai, but we know that there too, for part of the year at any rate, are pools close to the station.t Probably no one can now prove which of Rousselet’s two accounts is the correct one. If the first account is correct, then Old N’gamo must have been the spot where he collected, and it would appear that the doubt expressed above was justified, for while the material found at * Journ. Roy. Micr. Soc., 1906, p. 393. Mr. Rousselet says that he had col- lected it at ‘‘ a station beyond Gwaai, where a stream forms a rather shallow pool close to the station.” Formerly the Victoria Falls train stopped at Gwaai, and also 23 miles farther north (to take in water) at what is now Old N’gamo. Gwaai at present is an unimportant siding and Old N’gamo not even a halt, while a new station has been made at N’gamo. + Rousselet, C. F., Journ. Quek. Micr. Club, vol. xii, ser. 2, 1913-1915. t In April 1925 one of us, unconsciously imitating Rousselet, jumped from the train during an all too brief halt at Gwaai and secured several dips from adjacent pools, but unfortunately without finding Volvox. It was the end of a Observations on the Genus Volvox in Africa. 437 Old N’gamo (and also at N’gamo) possesses asexual colonies like those described by West for Rhodesia, the sexual colonies associated with them are not of the type described from Ussangu and presumed by him to belong to the same species. In any case, the existence of two distinct types has been established, and we have been fortunate enough to obtain from Rhodesia and other localities sexual and associated asexual colonies of both types—the Ussangu type from Egypt, Rietfontein (Transvaal), and Sedudu (N. Bechuanaland Protectorate), and the Old N’gamo type from the Cape Flats as well as from other parts of Rhodesia and the Bechuana- land Protectorate. The main point of difference between these two types is that in one the sexual colonies are dioecious, in the other they are monoecious. Now, whereas in V. globator the sexual colonies are consistently monoecious, it is well known that in V. awrews monoecious as well as the more usual dioecious colonies may occur, and we should hesitate to make this a character of diagnostic importance were it not for (a) its constancy in a particular type, and (0) its association with other slight but well-defined differences. With regard to (a), cultures of the Rietfontein Volvor have been watched constantly for a period of over eighteen months, and the colonies have been found to remain dioecious, with very rare excep- tions in which a few antheridia developed in otherwise preponderat- ingly female coenobia (the reverse—oospores in a male coenobium— was never once seen), while the Cape Flats Volvor during some years of observation (which was particularly close and detailed in.1930 and 1931) has been found to be consistently monoecious.* These two examples, from Rietfontein and from the Cape Flats, are cited in particular, since they are the two whose life-history has been methodi- cally observed, but collections of Volvor from elsewhere appear to confirm these observations. In view of the existence of these two types we propose to create a new species to include the type found at Old N’gamo and on the Cape Flats, and, since it has been fully known longer from the latter locality, to name it V. capensis. It will be seen that this species shows several particularly wet rainy season, and it seems probable that Volvox does not develop in that region until later in the season when the water has grown shallower. It is obviously widespread in Rhodesia. * The results of these observations are embodied in a paper by M. A. Pocock forming the third of the series on Volvox in Africa, and appearing in Part 4 of this volume. 438 Annals of the South African Museum. characters in common with V. globator, and that it is intermediate between that species and V. Rousseletit. Before proceeding to describe the forms we have found, we think it advisable to enumerate the features on which Shaw laid stress in describing new species and genera from Manila.* In addition to the recognised significance of dimensions, presence or absence of connecting protoplasmic strands, and variations in the oospore wall, he considered of importance the size attained by the gonidia (the asexual reproductive cells) before dividing, and whether or not they become segmented before the birth of the young coenobia in which they occur. Another point of importance, in his view, is whether all the daughter colonies are born through one opening in the wall of the mother colony, or whether each daughter is born through a separate opening ; a study of this in preserved material is, of course, very unsatisfactory. Shaw also laid stress on the migration of the gonidia from the outside to the inside of the developing embryo, but the subsequent or almost simultaneous work of Kuschakewitsch ft has shown that on completion of cell-division of the developing gonidium the young coenobium (embryo) undergoes complete inver- sion, whereby the inner surface comes to lie outside.{ Shaw further pointed out that it was important to make observations during the night or late afternoon, when he believed many of the changes in the life-history usually occur. We have acted on this hint, but so far our results do not confirm Shaw’s opinion—it is possibly correct as regards actual cell-division, but all other stages have been observed during the day as well as at night ; in fact, inversion appears to take place more commonly during the day, which perhaps explains how it came about that Shaw during his extensive studies on Volvox over- looked this phenomenon. In the following account we deal with the forms of Volvox we have examined from different localities, placing each under the species to which it belongs. * Shaw, W. R., op. cit., vol. xv, 1919; vol. xx, 1922; vol. xxi, 1922; vol. xxii, 1923. + Kuschakewitsch, Bull. Acad. Sc. de !Oukraine, vol. i, 1922, p. 1. t Both Kuschakewitsch and Zimmermann (Naturwissenschaften, 1925, p. 401) thought that inversion preceded cilia formation. Recent observations of the South African forms, however, show that the cilia are being formed during the process, and that probably the mechanism of inversion is intimately connected with cilia formation. Observations on the Genus Volvox in Africa. 439 A. Volvox aureus Ehrenb. Present in Eliazer Pan (Potchefstroom) and Weltevreden West (Ermelo), Transvaal, vleis in the N’gamo district (S. Rhodesia) and along the Linyanti River (N. Bechuanaland Protectorate). 1. From Potchefstroom. (Collected from Eliazer Pan, where it is associated with V. Rousseletii (Plate IX, fig. A).) The colonies are spherical or sub-spherical, from about 300 to 500 in diameter, and usually contain one type of reproductive cell only ; mixed colonies occur very rarely. The protoplasts of the somatic cells are rounded, with slight indications in surface view of angles where the protoplasmic connecting strands are attached. They measure from 6 to 9 yz across, and are from 1 to 4 diameters apart ; they are rather more closely packed * at the posterior pole (this is true of all three types of colonies). The protoplasmic connecting strands are very fine and thread-like; they do not preserve well, but usually show fairly clearly when stained with methylene blue. The number of daughter colonies varies from 4 to 13 (Plate X, fig. A), 8 being a frequent number; the size on escape is not less than 180 w; the gonidia at birth are most often undivided, or else are at an early stage (4- to 8-celled) of division. The sexual colonies are mostly dioecious, rarely monoecious, a little smaller than the asexual. The male coenobia are similar to those figured by Janet (Le Volvox, 1912, fig. 15, J) ; they are shown in the microphotographs (Plate IX, figs. B and D, and Plate XI, fig. A). In all these photographs the male colonies are somewhat young, with a large number of antheridia in various stages of development. They measure about 400 to 450 yz in diameter. The number of antheridia is very large (in two cases 403 and 605 respectively were counted), scattered thickly over three-quarters of the colony, leaving the anterior quarter free; somatic cells occur among them. The sperm bundles (Plate X, fig. C) are saucer-shaped, measuring from 14 to 20 » across, and each is composed of 32 sper- matozoids. The female coenobia (Plate IX, figs. B and C, and Plate X, fig. D) are slightly larger than the male, diameter 450 to 520 p, and usually contain about 8 oospores, but the number varies from 4 to 11. The * We have noticed closer packing at the posterior pole in English specimens of V. aureus, though no reference is made to this in the text-books. Shaw mentions it in connection with V. aureus from California (Op. cit., vol. xx, May 1922, p. 484). 440 Annals of the South African Museum. oospore measures from 64 to 73 yw in diameter, with smooth, thick outer wall (Plate X, fig. E). A comparison was made with V. aureus forma madagascariensis (Professor Fritsch’s material), but we find it agrees more closely with the type than with this form. 2. From Weltevreden West (only asexual colonies were collected). 3. From N’gamo and Old N’gamo (S. Rhodesia), and from vleis along the Linyanti River (Bechuanaland Protectorate), particularly at Kazungula and Kabulabula (Plate XI, figs. B, C). V. aureus is evidently widespread in this region of Africa. The collections were made in July, 7.e. midwinter, when the vleis left by the rains of the previous wet season (ending March or April) were still extensive. It was usually associated with V. Rousseletii and Pleodorina californica (Plate XI, fig. B), and a rich phytoplankton of desmids, and other unicellular, colonial, and filamentous algae. Except for its greater size it is similar to the V. aureus found at Potchefstroom, and is characterised by the small degree of differentia- tion reached by the third generation at the birth of the daughter colonies, and the marked rarity of mixed colonies, the great majority producing either only gonidia, only antheridia, or only oogonia. The number of cells is usually between 1000 and 3000, the protoplasts are from 6 to 9 in diameter. The greatest size noted for asexual colonies was 782 x 816 jz; female colonies up to 604 x 625 uw, and males up to 539 x 560 w, were measured (these measurements were all made free on the slide, not under the pressure of a cover slip). The smallest free colony measured was 182 x 194 yw, but daughters in situ with a diameter of 243 w were seen; the largest number of daughters per coenobium observed was 11. Platelets, always of 32 spermatozoids, measured 20 uw; oospores (up to 11 per colony) from 65 to 84 yw in diameter. B. Volvox africanus G.S. West. Plate XII, figs. A, B. Late in our investigations we found this species in material from a pool near the wagon road leading from Kazungula to Kasani, on the Linyanti River. The pool was long and shallow, about 1 foot deep, and fairly clear. The Volvoz association was rich—large colonies of V. capensis predominated, with V. aureus, V. africanus, and Pleodorina californica mixed with them. The colonies of V. africanus were comparatively few, and all asexual; in most, the daughter colonies numbered 4—the 2 larger opposite one another in the Observations on the Genus Volvox in Africa. 441 equatorial plane, the 2 smaller near the posterior pole—but the number varied from 1 to 6. In most cases where an unpaired daughter occurred, the other member of the pair was represented by an abortive gonidium. So far as could be seen the daughters themselves were all asexual. The smallest colony seen with mature daughters measured 345 x 431 uw, and contained a single well-developed daughter; the largest, 603 x 646 x, with 4 daughters one of which measured 218 x 258 x; each of these 4 daughters bore 2+2 gonidia in early stages (2- to 4-celled) of division. This Volvo agrees very closely with the description given by West,* except that no trace whatever of protoplasmic connections between the cells in the adult colonies, or even in young colonies, could be made out. (This is in accord with recent observations made by Professor Lyengar on a small form of V. africanus collected in Southern India.t See also Shaw’s remarks quoted above.) In unborn daughter colonies, however, after treatment with methylene blue, a connection between the cells is here and there observable; this resembles the state of things described by Meyer for JV. tertius.t C. Volvox capensis sp. nov. Volvox monoicus; coenobiis sub-globosis vel late ellipsoideis ; coenoblis asexualibus majoribus usque ad 1-3 mm. latis, cellulis ca. 9000-23,000, plerumque 18,000—20,000, in parte posteriore densiore ageregatis ; coenobus infantibus 3-20, plerumque 6-10, saepe 8, in planis duobus alternantibus, 170-430 yw, plerumque 200-300 p latis ; coenobiis sexualibus usque ad 1 mm. latis, bisexualibus protandreis ; androgonidiis 4-23, plerumque 6-10, similiter coenobiis infantibus dispositis ; spermatozoidis 512 rarissime 256, in globoidis depressis 40-44 pw latis; oosporis 60-160, plerumque ca. 120, diametro sine spinis 37-49 y, in parte tertia anteriore coenobii absentibus ; spinis cavis conicis usque ad 5-8 yp longis obtectis. This species is intermediate between V. globator and V. Rousseletii. From the former it is distinguished by (1) greater size; (2) larger number of somatic cells ; (3) antheridia more numerous, developing uniformly into depressed hollow globoids of 512 (rarely 256) spermato- zoids, never into platelets; (4) much larger number of oospores, averaging 120 per coenobium as compared with 30 in V. globator ; (5) spines of exine longer and stouter, less numerous as seen in optical * West, G.S., op. cit., 1918. + In the press. + Meyer, Bot. Zeit., vol. liv, p. 200. 442 Annals of the South African Museum. section. From the latter it is distinguished by (1) smaller size of coenobia, which are subglobose to broadly ellipsoid instead of ovoid ; (2) slightly smaller number of somatic cells ; (3) sexual colonies always monoecious and protandrous; (4) antheridia much fewer—at most 23 or 26 as against several hundreds in V. Rousseletii—and developing differently ; * (5) oospores less numerous, as large or larger, but spines usually shorter. The gonidia are differentiated before completion of cell-division, and can be clearly seen in daughter colonies which are about to invert. They remain undivided until after the birth of the daughter colony, but are easily distinguished by their greater size (diameter 11 to 15 pu as compared with 4 to 5 w in the case of the somatic cells). The sexual reproductive cells may also be distinguished before birth, but are smaller than the gonidia (diameter 6 to 8 4). Under normal conditions the coenobia are markedly globose ; the polar axis in young colonies is a little longer than the equatorial diameter, becoming more so as the colony ages, particularly in the sexual colonies, but less markedly so than in V. Rousseletii. Usually the two poles are similar in form (see Plate XII, figs. C, D, and Plate XIII, fig. A), whereas in the latter species the tendency is for one or other pole to become broader (cf. Plate XVIII, figs. A-C, Plate XX, fig. F, and text-fig. 2, G). In both species during cold weather the anterior pole sometimes protrudes in a peculiar and characteristic manner (see some of the coenobia in Plate XVIII, fig. D). In this species we include the Volvox occurring in many vleis on the Cape Flats, and which we regard as typical, and that collected in Rhodesia (N’gamo, Old N’gamo) and the Bechuanaland Protectorate (backwaters and vleis of the Linyanti or Chobe River). It was at first thought that the latter formed a distinct variety, but after obsery- ing V. capensis on the Cape Flats during an unusually late season— November 1931—at temperatures much higher than are usual for Volvor in these parts, we had to modify this opinion. The chief features considered characteristic of the Rhodesian form—.e. large size, unusually large number of daughter colonies and of sperm bundles per coenobium—were found appearing, though to a less marked degree, in the Flats Volvox, in those vleis which still retained water (e.g. Ottery Road, and Belvedere Road Vlei III). The Rhodesian and Bechuanaland material is therefore described merely as a form of V. capensis and not as a variety. * Pocock, M. A., this volume, Part 4. Observations on the Genus Volvox in Africa. 443 1. From the Cape Flats. The “ Cape Flats ” is the name given to the strip of low-lying land forming the isthmus which joins the Cape Peninsula to the mainland. The greater part is covered by blown sand, but near Table Mountain are fairly extensive stretches of alluvial soil (Plate XXIV, fig. D). During the winter months shallow pools, some of considerable extent, form on the flats, and these yield a rich algal flora. In several of those on the alluvial soil Volvox has been found ; all dry up very soon after the south-east winds of early summer begin, so that as a rule Volvox is found only in winter and spring, only very rarely is it found as late as November, as in 1931. In October 1927 Volvox was collected from two of a series of vleis near Belvedere Road * (between Keurboom and St. Michael Roads), Claremont, and from Scanlan’s Vlei, Wetton Road, a few miles distant ; observations were started on the living material, but the supply ran short just when further investigations were desired. The following season (1928) was a particularly good one for Volvor, whereas that of 1929 was very unfavourable—the winter rains accompanied by cold weather were late, the late rains were poor, and many vleis did not fill up at all. At first a fair amount of good Volvox material (all asexual) was collected, but later on in the season repeated efforts to obtain sexual material were entirely unsuccessful. In one case a promising pool in a ditch near the farm Vaderlandsche Rietvlei yielded on a second visit plenty of coenobia, but in every case they were inhabited by a Rotifer busily eating away daughter colonies and parent cells,t and no sexual colonies were present (Plate XIV, figs. B-D). Other pools previously full of Volvox had now scarcely any water in them and consequently no algal growth. The winter of 1930 was exceptionally dry, particularly on the Cape Flats—one farm on Princess Vlei recorded a total rainfall of only 4 inches for the year (considerably less than the Kalahari)—and in most parts the rain simply soaked into the ground, no pools being formed. The Wetton Road Vlei, however, was an exception to the general rule ; * Of these, one of the best has since become part of Ackerman’s Sports Ground, and is in process of being filled up with rubbish and sand. + A photograph of this Rotifer has been sent to Mr. C. C. A. Monro, Assistant Keeper in the Department of Zoology, British Museum (Natural History), who has been kind enough to examine it. He says the photograph appears to corre- spond most nearly to Ascomorpha volvocicola Plate. Several species of Rotifer are known to live in Volvoz. Many individuals and their eggs were often observed inside a single coenobium. Even when half the cells and the daughter colonies had been devoured the colonies were still active and strongly phototactic. VOL. XVI, PART 3. 29 444 Annals of the South African Museum. it lies near Wynberg, a part which has more rain than the rest of the district, was fairly full of water and yielded abundant Volvox in very good condition. Later in the season the Volvox in the main vlei (Scanlan’s) was badly attacked by Rotifers, but a ditch nearby con- tained particularly large and healthy colonies for some time longer. In 1931 good rains in May, followed by warm sunny weather, resulted in a fine crop of Volvox; the Volvor supply continued good right through the winter—now in one vlei, then in another—and well into the summer, some of the most interesting observations being made from the Ottery Road Vlei, Wynberg, and Belvedere Road Vlei as late as November. It was found that Volvox reached its optimum development in the unusually high temperatures of a week of hot weather in November (water temperature over 90° F.). The study of this living material has thrown light on several per- plexing points, and has enabled the second author of this paper to write on the subject much more extensively. The coenobium is usually a prolate spheroid, with the polar axis a little longer than the equatorial, but the shape varies considerably, and the equatorial diameter may equal, or sometimes in mature asexual colonies even exceed, the polar axis. As a rule, the sexual colonies are more elongated than the asexual. The number of cells in mature colonies varies from about 9000 to 23,000, most often lying between 18,000 and 20,000 in a vigorous mature strain. The number drops in colonies produced in cold weather and as the strain ages. The protoplasts are more closely aggregated towards the posterior pole. As seen in side view they are somewhat pear-shaped (Plate XV, figs. Aand B). The apices of the cells do not quite reach the outer membrane. When young the protoplasts appear star-shaped in surface view (Plate XIII, fig. B) ; the angles are drawn out to form relatively thick protoplasmic connections into which the chloroplast extends, so that the whole coenobium appears covered with a green network. This shows up clearly even when the colony is rotating ; it is seen most beautifully if the colonies are examined with a fairly low objective against a black background. As the colony matures, the protoplasts become further apart and smaller, and the connecting strands more finely drawn out, so that the star-shaped form is more or less lost; the chloroplasts too are withdrawn from the strands. There is a single chloroplast filling the base of the cell and extending up one side to the cell apex at the points of insertion of the cilia ; in the massive bowl-shaped basal portion is embedded one large pyrenoid. Observations on the Genus Volvox in Africa. 445 The two long cilia are inserted separately at opposite corners of the cell apex as in V. Rousseleti (see text-fig. 3, A); at the point of insertion of each is a highly refractive granule. The basal quarter of the cilium projects from the enveloping membrane at right angles to the surface, and is rigid and immobile, with the remaining three- quarters lashing vigorously with a rotary motion in a spheroidal plane parallel to the outer wall of the coenobium. In the anterior cells the large lenticular orange-red eyespot or stigma lies in the upward extension of the chloroplast, a little below the apex, and symmetrically between the points of insertion of the cilia always facing towards the anterior pole (Plate XV, fig. A). At the anterior pole it is about 2 in diameter, becomes smaller towards the equator, and disappears entirely as a pigmented spot in the cells about the posterior pole. In the latter its place is occupied by a highly refractive granule, just as described by Janet * for V. globator and V. aureus. From two to four contractile vacuoles with a periodicity of from 15 to 30 seconds, more or less, occur in the angles of the base of the protoplast near the connecting strands, and probably one or two smaller vacuoles are situated in the colourless throat of the proto- plast near the apex ; the latter were, however, seldom distinguishable. Occasionally contractile vacuoles may occur in the connecting strands themselves. The number of daughter colonies early in the life of a strain is usually 8, arranged alternately in two planes (Plate XIII, fig. D). Plate XII, fig. C, shows a group of colonies from a very young strain, collected about a week after rain had fallen and formed a pool in the vlei. It includes several “juveniles,” + formed on germination of oospores, each with one or two daughter colonies. The latter may give rise to from 4 to 6 daughters, and the third generation to 8. If conditions are favourable—in particular if the weather be warm—the number of daughters increases ; in a vigorous strain there are usually from 9 to 12 daughters, or even more. During the succeeding sexual phase (Plate XII, fig. D) the number may drop, but this is not always the case (Plate XIII, fig. A, and Plate XIV, fig. A). The latter figure shows a group of colonies in which the number of daughters was abnormally high—as many as 20 were seen, while from 15 to 18 were common. This case, from Belvedere Road Vlei No. 3, has already been referred to, and it seems certain that the high number of daughters was directly connected with the unusually high temperature. * Janet, C., Le Volvox, 1912, p. 48. + These juvenile forms are described fully in paper No. 3. 446 Annals of the South African Museum. The daughter colony rotates for some time within its vesicle (the enlarged gonidium wall), finally escaping outwards through a pore which it bores in the wall of the coenobium. Normally, in a healthy colony each daughter escapes thus through its own pore ; the daughter colonies do not rotate freely within the hollow of the parent; each vesicle remains intact until broken by the escaping daughters in the region where it forms part of the outer membrane of the colony. The sexual colonies are usually globose (Plate XIII, fig. C) or broadly ellipsoidal, somewhat smaller than the asexual colonies, monoecious and strongly protandrous, with well-marked polarity, the anterior end always being clear of reproductive cells. Unless the strain is an old one nearing the end of sexual activity it is most unusual to find daughter colonies in a sexual coenobium. The antheridia vary from about 4 to 20 in number, usually from about 6 to 10. Asin the case of the gonidia, there appear to be 8 primary antheridia arranged similarly, 7.e. alternately in two planes. More than 13 were found only in the hot weather form from Belvedere Road. The mature sperm bundle is a depressed hollow globoid of 512 sper- matozoids, the central hollow appearing circular in surface view, narrowly elliptical in side view (Plate XV, figs. C and D). The cilia are very long. Like the daughter colony, the sperm globoid usually escapes outwards; occasionally, however, it escapes inwards, and may disintegrate within the hollow of the parent. The spermatozoid is similar to that of V. Rousseleti (text-fig. 3, D). When the sperm globoids are fully developed, the oospheres are still immature ; only very rarely are sperm globoids and oospores to be found in the same colony. Gaps in the cell network show where the globoids have developed and escaped (Plate XIII, fig. C). The oosphere (Plate XII, fig. E) is deep green, very broadly flask- shaped, the broad apex a little below the outer surface of the parent. After fertilisation it rounds off, lying below the level of the inner membrane, enclosed within the delicate vesicle formed from the much enlarged wall of the oogonium, which in its early stages develops similarly to the young antheridium. Within this vesicle the oospore develops. At first it is bounded by a thin wall; the outline becomes irregular, and an outer hyaline wall of spirally arranged conical spines is formed, the green protoplast extending into the spines. When this exospore has been secreted, the green protoplast is withdrawn from the spines, rounds off once more, and the inner wall is laid down (text- fig. 4, C, and Plate XV, fig. F). This again consists of two layers—a thick mesospore and a very delicate endospore bounding the protoplast. Observations on the Genus Volvox in Africa. 447 The colour of the protoplast changes from green to orange-red, and the oospore is ready for liberation, which is brought about by the disinte- gration of the parent. This takes place very soon after the change in colour, so that on the Cape Flats colonies containing ripe oospores are remarkably rare (Plate XIII, fig. E, and Plate XV, fig. E). Oospores in various stages of development are shown in text-fig. 4, C,. 1-3. Dimensions, etc., of V. capensis from the Cape Flats. Mature asexual colony, 667 x 754 wz to 1250 x 1292 w; average about 800 x 850 pu. Mature sexual colony, 495 x 560 w to 928 x 1034 w; average about 625 x 660 p. Mature daughter colony, 129 x 139 w to 431 x 495 w; average about 200 x 230 p. Somatic cell: anterior pole 12 to 18 2; posterior pole 11 to 13 p. Somatic protoplast: anterior pole 6 to 11 w; posterior pole 5 to 9p; average 5 to7 p. Somatic protoplast in daughter colony before birth, 4 to 5 p. Gonidium protoplast in daughter colony before birth, 11 to 13 p. Sexual cell protoplast in daughter colony before birth, 6 to 7 p. Number of somatic cells, 9000 to 23,000, most often 18,000 to 20,000. Number of daughter colonies, 4 to 12 ; in very warm weather up to 20. Number of sperm globoids, 5 to 13; in very warm weather up to 19. Number of oospores, 60 to 160; average 120. Size of sperm globoid, 33 to 44 p. Size of oospore with spines, 49 to 62 4%; without spines, 37 to 49 w; spines, 5 to 8 p. 2. V. capensis forma rhodesiensis f. nov. Forma magna, antheridiis numerosioribus, usque ad 35, oosporis majoribus quam in forma typica. From N’gamo and Old N’gamo in Rhodesia, and from vleis and backwaters of the Linyanti, Bechuanaland Protectorate.* (a) From N’gamo and Old N’gamo (Southern Rhodesia). Near N’gamo station there are several small vleis from which Volvox was collected at the beginning of July 1930. The following day the vlei at Old N’gamo, 5 miles north, was visited ; Volvox was collected * Geographically parts of both Northern and Southern Rhodesia and the Bechuanaland Protectorate are one region. Their boundaries meet on the Linyanti near Kazungula, hence the regional name may appropriately be used for the form. The word N’gamo is probably of Bushman origin, meaning water. 448 Annals of the South African Museum. in several parts, but chiefly in the deep channel to the east of the railway embankment. V. awreus predominated, but a fair amount of the monoecious V. capensis was found. There were indications - that the latter species was in an early stage of development, e.g. the presence of one or two “juveniles” from recently germinated oospores, and the rarity of colonies with developing or mature oospores. Both asexual and sexual colonies are globose when young (Plate XVI, fig. A), the polar axis only slightly longer than the equatorial, but the sexual colonies often become considerably elongated during the ripening of the oospores. The asexual colonies resemble those from Rhodesia photographed by West in 1910 (Op. cit., pl. 11, figs. 1-7). The average number of daughters is 8, but the number varies con- siderably. There is also considerable variation in the size of the protoplasts ; often these are similar in size to those of typical adult colonies from the Cape Flats (5 to 7 » in diameter), but other colonies of approximately the same age occur in which the protoplasts are considerably larger (8 to 10 y), giving the cell network a very much coarser appearance. A similar variation was observed on the Cape Flats, but there nearly always correlated with either the age of the strain or the individual, or with external conditions. Here and on the Linyanti, however, the various types occur side by side, and it is impossible to say from the preserved material, collected over a total period of only three weeks for the whole district, whether this variation is characteristic of the form as a whole, or whether different forms are actually present. Plate XVII, fig. A, shows a young sexual colony which was of the large-celled type ; 5 or 6 mature sperm bundles and numerous young oospheres are visible ; gaps in the cell network indicate positions of sperm bundles which have escaped. Plate XVII, fig. B, shows in more detail a small portion of the cell network with a very young oogonium in the early stage of enlargement. This can be distinguished in fig. A as a larger black cell a little below the centre to the left, next to the oosphere directly above the large gap in the network seen near the posterior pole of the colony. These photographs and Plate XVI, figs. B-E, show what beautiful preservation may be obtained by using Powers’ Iodine method of fixation.* In Plate XVII, fig. B, the star-shaped cells are practically undistorted ; the protoplasmic con- nections are clear, but since they dip down below the point where they unite with the protoplast they appear discontinuous. The sexual colonies are large in proportion to the asexual, but few * See Chamberlain, Methods in Plant Histology, 4th ed., pp. 178-179. Observations on the Genus Volvox in Africa. 449 really mature asexual colonies are present in the preserved material, and they probably reach a considerably larger size during the asexual phase. The number of sperm globoids is generally high ; from 7 to 20 per coenobium were counted, while 15 to 18 are common. They are composed of 512 spermatozoids, and are similar in structure and appearance to those from the Cape Flats but somewhat smaller (Plate XVI, figs. B and C). The oospores are large (Plate XVI, fig. E), of the same type as those of the Cape Flats, but the spines are a little longer and more strongly developed (text-fig. 4, D). Plate XVI, fig. D, shows a nearly mature oosphere which had sperms swarming round the base. The broad apex is sunk slightly below the neighbouring somatic cells, with which it is still in protoplasmic connection ; the broader rounded base, enclosed in the vesicle outside which are the sperms, projects into the cavity of the parent colony. The rounded central body is the nucleolus, the rest of the large nucleus being obscured by the massive chloroplast. Dimensions of V. capensis f. rhodesiensis from N’gamo and Old N’gamo. Asexual colonies (immature), 800 x 858 px to 1034 x 1142 p. Sexual colonies, 603 x 624 px to 1292 x 1335 p. Daughter colonies, 345 x 388 p. Somatic protoplasts, 5 to 9 p. Sperm globoid, 30 to 35 w; 7 to 20 in number. Oospore with spines, 64 to 69 jz; without spines, 40 to 46 yw; spines, 9 to 13 p. Number of oospores, 50 to 220. (b) From the Linyanti River (Plate XVII, figs. C, D). On the south bank of the Linyanti or Chobe River, near its junction with the Zambesi River, a number of pans and vleis of varying extent form during the rainy season, while farther west there is extensive swamp-land, also bordered by a series of similar vleis. In most of these pans and vleis, and in many quiet backwaters of the river itself, Volvor was found. The district is obviously very rich in Volvoz, and would probably well repay a prolonged and intensive study of the group on the spot. In the majority of cases V. awreus and a form of V. capensis were present in varying proportions ; in one vlei near Kasani V. africanus was found, and in a few cases, e.g. at Sedudu, the dioecious V. Rousseletii was also collected. The collections are, however, too limited—usually not more than one from each locality— 450 Annals of the South African Museum. to arrive at any clear idea as to the conditions governing the appear- ance of any one species. As a whole, the material of V. capensis is very like that described above from N’gamo, but there are a few points which may be noticed here. The variation in the size of the protoplasts is even more marked than in the N’gamo material. The number of sperm bundles is greater, as many as 35 having been counted in a single large coenobium. The oospores, on the other hand, although very large, have as a rule relatively shorter spines, and are more like the Flats form than are those from N’gamo; in some localities, however, the longer-spined form also occurs. Dimensions of V. capensis f. rhodesiensis from the Linyanti River. Asexual colonies, 625 x 667 ys to 1323 x 1357 wp. Sexual colonies, 539 x 646 yw to 1252 x 1313 p. Daughter colonies, 322 x 344 «; 5 to 19 in number. Somatic protoplast, 5 to 10 p. Sperm globoids, 32 to 38 w; (?) 7 to 35 in number. Oospore with spines, 58 to 70 3; without spines, 43 to 53 p; spines, 7 to 11 p. Before concluding our notes on V. capensis we wish to draw atten- tion to the resemblance between it and V. barbert Shaw,* which, however, appears to have a somewhat different cell structure. We have had the opportunity of examining some of Shaw’s Philippine material, and we find that the spines which he gives as about 4 wu high really measure from 5 to 7 ps, or even up to 9 x, numbers which agree well with those found by us for V. capensis. In general appearance, however, V. barberi is more like V. globator. D. Volvox Rousseletii G. S. West. West placed in this species the Volvox whose asexual + and sexual colonies he obtained from the Ussangu Desert. The asexual colonies were presumably considered by him to be identical with those he had * Shaw, W. R., loc. cit., vol. xx, No. 5, May 1922, p. 496. + West did not distinctly state, but left it to be inferred, that asexual colonies from Ussangu were examined by him. Thanks to the kindness of Professor Stiles, we have been able to examine West’s slides of this material, and we find that asexual colonies are present. (We have also been able to examine Mr. Rousselet’s slides of the Volvoz from Rhodesia, thanks to Mr. Scourfield, to whom the slides were lent by Mr. Bryce.) Observations on the Genus Volvox in Africa. 451 previously described from Southern Rhodesia in material collected by Mr. Rousselet, material which, as we stated above, was destitute of sexual colonies. As we have examined West’s slides, we give first a supplement to his description of this Volvor, and after that we describe the similar forms we have obtained from Egypt and Rietfontein. 1. From the Ussangu Desert. (Preserved material only. Associated with V. africanus.) The sexual colonies have already been described (see pp. 434, 435 of the present paper). To that description we make the following additions: The asexual colonies are broadly ovoid in shape, the posterior end being usually more pointed than the anterior. Length from 850 to 1200 yw, width from 810 to 1040 x. The somatic cells are more densely packed at the posterior end, distance between centres here being 9 uz, while at the anterior end it is 11 uy. The daughter colonies vary in number from 6 to 10 (7 and 8 being the most usual numbers), and they reach a diameter of over 300 y while still in the parent. Of the sexual colonies there are four good photographs by West (Op. cit., 1918); they are dioecious, more pointed at the posterior end, and a little smaller than the asexual colonies. The sperm bundles are very numerous, and are crowded into the posterior four-fifths of the male coenobium ; they measure from 29 to 32 » in diameter, and are similar to those which we shall describe from Rietfontein. A drawing of a mature oospore is given in text-fig. 4, A. 2. From Egypt. (Preserved material only ; rather scanty.) This Volvox was collected in April 1924 by Mr. Douglas Simpson, who says that it appeared in rock-pools, after rains, in the Wady Khafur, Egypt. The asexual colonies are subglobose or ellipsoidal (see text-fig. 2, A), and measure up to 1000 p in diameter. The protoplasts of the somatic cells are somewhat angular (text-fig. 2, B, C, and D), and measure from 4 to 6 px. The protoplasmic connections are relatively short and stout. The number of daughter colonies observed was either 2 or 4, but it is extremely likely that had more coenobia been present a greater number of daughters in some of them would have been noticed ; this seems to be a very variable feature in one and the same species of Volvor, depending apparently not only on inherent characters, but also to a large extent on external conditions. The cells of the daughters show very little differentiation before escape. The sexual coenobia are dioecious, broadly ovoid in shape, 452 Annals of the South African Museum. Fia. 2.—V. Rousseletii, from Egypt. A, Asexual coenobium (somatic cells omitted), x 230. B, C, D, Somatic cells in surface view. 1000. (B, stained with methylene blue ; C, with Delafield’s haematoxylin.) E, Oospore (not fully mature). x 400. F, Portion of oospore wall more highly magnified. G, A male coenobium, 170. Observations on the Genus Volvox in Africa. 453 and the more rounded (anterior) end is free from sexual elements. The male colonies (text-fig. 2, G) are from 300 to 950 pw in diameter ; they sometimes contain daughter colonies, and appear to be very similar to the photograph taken by West from the Ussangu material.* The sperm bundles are numerous, and measure from 35 to 42 yz across (usually 36 yx). Two of the female coenobia measured respectively 850 x 1000 p, and 1100 x 1200 p. The oospores are numerous, up to nearly 200, and are thickly covered with spines which are broadly conical in shape (text-fig. 2, E and F), their bases almost touching. The oospores (not perfectly mature) measure 36 mw in diameter ; length of spines 5 to 6 p. It appears then that this Volvox differs only from the Ussangu one in the following respects :— (1) The number of daughter colonies is less. (2) The spines on the oospore are shorter (probably less mature material). 3. From Rietfontein (Transvaal). (Preserved and living material abundant.) In January 1928 Mrs. Hutchinson collected Volvor from Riet- fontein ; in February 1930 Mr. Steer (of Sea Point) obtained vigorous cultures of Volvor from dried mud from the same pool. From his original culture it was transferred to a large tub, whence it spread to other tubs in the garden and to his fishpond, the water of which (4 to 3 feet deep) was green with it for three or four months of 1931. Altogether Volvox continued in abundance in Mr. Steer’s garden for over eighteen months. One of us was able with this very plentiful material to obtain a series of microphotographs showing details of structure and development, the inversion of the sperm globoids, and the subsequent escape of the spermatozoids. It was found that after cold weather and rain the amount of Volvor dimin- ished and that the coenobia were smaller. Usually the first, purely asexual, stage was followed by a sexual stage, and this in turn was succeeded by a second asexual one, the Volvor then as a rule dis- appearing altogether to reappear after some weeks. The Volvoxr in culture remained similar to that collected at Rietfontein. When young the coenobia are globose, very like those of V. capensis, but as they develop they become markedly ovoid, particularly in the case of the sexual colonies ; usually the anterior pole is broader than the posterior, but this is not always so. The number of cells is from * West, G.S., op. cit., t. 29, 1918, fig. 1. 454 Annals of the South African Museum. 15,000 to 42,000 (average 20,000 to 23,000). The mature asexual colonies in a vigorous strain (Plate XX, fig. A) range from 690 x 754 ju to 2058 x 2100 4; very large colonies were, however, rare, and were only seen during the early stages of the original culture; but at the height of development of any one culture colonies about 1300 x 1500 ju or even larger were common. The number of daughter colonies is most often 8, but varies from 4 to 16 (see Plate XVIII, fig. A). The 8 primary gonidia are arranged alternately in two planes, one approximately equatorial, the other half-way between equator and posterior pole. The gonidia are differentiated before birth—probably either before or immediately on completion of cell-division (Plate XVIII, fig. E); they are at any rate distinguishable from the somatic cells very soon after inversion of the daughter colony by their greater size and by the possession of more than one pyrenoid. They do not begin to divide until after birth. At birth their diameter is about double that of the somatic cells (gonidia 11 to 12 px, somatic protoplasts 4 to 5 w). The daughters at birth are usually large (up to 473 x 539 jx), but there is a wide range in size. The smallest free colony measured was 182 x 204 yp, while the largest in situ was 630 x 672 yw, but this was exceptional. The number of daughters depends partly on the phase reached, partly on external conditions, mainly temperature and supply of nutritive material. The groups shown in the microphotographs (Plate XVIII, figs. A—D) illustrate excellently the characters seen in successive phases (A, C, and D were from one culture, B from a later one, as the corresponding phase had been missed in the earlier series) :— A, 22nd February 1930. Mr. Steer’s original culture at the height of the first asexual phase. Daughter colonies up to 13 in number. Colonies large—at about this time the largest colonies (over 2 mm. in diameter) were seen. B. 11th January 1931. From a culture derived from the above: sexual phase at its height, male colonies large, the majority of the sperm globoids having already escaped ; female colonies with ripening oospores predominating. Asexual colonies with few daughters. C. 23rd March 1930. Original culture—end of sexual phase ; male colonies very few, with scattered male globoids. Asexual colonies with 2 to 5 daughters. D. 28th March 1930. Second asexual phase. Daughter colonies very few—l1 to 3, most often 2. Soon after this date Volvor practi- cally disappeared from the culture for some weeks. Observations on the Genus Volvox in Africa. 455 The vegetative cell is similar in structure to that of V. capensis. Plate XIX, fig. A, shows the appearance of cells at the anterior pole of a mature colony (7.e. one with nearly mature daughters). The position of the eyespot relative to the pole is well seen. The outer membrane shows clearly, the inner only faintly. These cells were living, the outer portions of the cilia lashing vigorously, and appear- ing as a blurred zone outside the limiting membrane; the basal quarter, which does not move, shows clearly, as does the mode of insertion, the bases of the two cilia being wide apart from one another. The four other microphotographs in this plate show the difference in the arrangement of the cells at the anterior and posterior poles of a colony, as seen in surface view. B and C are portions of a young middle-aged colony treated with iodine. In B, the connecting strands and cell walls show clearly ; in C, from near the posterior pole the cells are closely packed, and several show contractile vacuoles ; near the bottom is an undivided gonidium. D and E are from an older colony, living and unstained, slightly more enlarged. The protoplasts are further apart and are becoming smaller, particularly in the posterior zone, fig. E, where many strands are seen to be anastomosing, and several contractile vacuoles in diastole are visible. The walls show faintly in D. Text-figs. 3, A-C, show the structure of cells from the anterior and posterior poles in greater detail. The sexual colonies are dioecious, and the anterior region is free from reproductive elements. The male colonies (Plate XX, fig. B) are usually elongated, with numerous antheridia which develop at varying rates. Thus any one coenobium may contain all stages, from antheridial cells which are still in early stages of enlargement, up to mature sperm globoids, and in addition may show gaps where sperm globoids have formed and escaped. The colony shown in fig. B illustrates this particularly well. The number of antheridia varies from about 120 (probably more) to over 200. The sperm globoids are even more depressed than in V’. capensis, and slightly smaller, with the 512 spermatozoids closely packed (Plate X XI, figs. Aand B). In general the male globoid escapes outwards, as in V. capensis, and moves freely in the water for a time before the sperms escape. The delicate surrounding vesicle is left when the globoid escapes, and numerous empty vesicles can be seen in any male colony from which globoids have been escaping. The spermatozoid (text-fig. 3, D) is elongated, 10 to 13 » long, 2 x wide, with thickened base sometimes drawn out into a short point, has an elongated chloroplast with one Fia. 3.—V. Rousseletii. A-C, Details of cell structure. A, Four cells at the anterior pole. B, Cells from region of posterior pole. C, Single cell (side view) from equatorial region. D, Group of spermatozoids. s. =eyespot ; c.v.=contractile vacuole ; b.g. = basal granule ; n.=nucleus ; nl. =nucleolus ; f.v.=food vacuole ; p.=pyrenoid ; g.=glycogen. A-C x 2000 ca. =D x 2500. Observations on the Genus Volvox in Africa. 457 pyrenoid, central elongated nucleus, red eyespot, and 2 contractile vacuoles near the anterior end, below the base of the elongated hyaline beak. At the apex of the beak are inserted the two long backwardly directed cilia (up to 23 pw in length). The spermatozoids swim actively in the water, or they may be seen inside the coenobium, where they move in an amoeboid manner, the cilia remaining motionless. The female coenobia are usually slightly larger than the male; they produce from 60 to over 200 oogonia, each of which develops a single large flask-shaped oosphere (Plate X XI, fig. D), the apex of which les a short way below the surface of the colony and is con- nected by protoplasmic strands with neighbouring cells. Later, pre- sumably after fertilisation, the oospore rounds off and lies entirely within the peripheral zone of the parent (Plate X XI, fig. E). In the young colony shown in optical section in Plate XX, fig. D, most of the oospores have sunk inwards, but a few flask-shaped oospheres are still visible round the wall. The oospore develops as in V. capensis ; the shape of the conical spines, however, is different—they are longer, shghtly narrower, and usually more numerous in an optical section (Plate XXI, fig. C, and text-fig. 4, B). The size of the oospore, including spines, is about the same as in V. capensis; hence since the spines are longer, the spore body in V. Rousseletii is actually slightly smaller. Plate XX, figs. E and F, show two large female colonies with nearly mature oospores. The absence of gaps in the cell network is noticeable. Occasionally mixed colonies containing asexual daughters and either sperm globoids (Plate XX, fig. C) or oospores are seen. Accord- ing to our experience these mixed colonies may occur in any of our forms as the strain advances in age, hence usually towards the end of the sexual phase, never in the early stages of development. In V. Rousseletii as a rule the sexes are quite distinct, but very rarely a male globoid occurs in an otherwise purely female coenobium. Such cases occurred so rarely during the time (18 months) that the Riet- fontein cultures were under observation, that they may be regarded as exceptions to the general rule. Dimensions of V. Rousseletii. (Mr. Steer’s Rietfontein culture.) Mature asexual colony, 690 x 754 pw to 2058 x 2100 p. Mature sexual colony, male, 646 x 753 wz to 1184 x 1378 x; average 700 x 770 p. Mature sexual colony, female, 646 x 711 yz to 1058 x 1292 w; average 840 x 960 p. 458 Annals of the South African Museum. Fia. 4.—Volvox oospores. A, B, V. Rousseletii. A, from Ussangu. B, from Rietfontein. C, V. capensis, from the Cape Flats. D, V. capensis f. rhodesiensis, from Old N’gamo. E, V. amboensis. C and D show different stages of development. A, B, C;, D;, and D, are mature oospores with fully developed exospore. D, and D, are from the same coenobium. All figures x 600. Observations on the Genus Volvox in Africa. 459 Mature daughter colony, 182x204 uw to 630x672 w; average diameter 300 to 400 p. Somatic cell: anterior pole 11 to 17 ~; posterior pole 6 to 11 p. Somatic protoplast: anterior pole 5 to 8; posterior pole 3 to 6 p. Somatic protoplast in daughter colony at birth, 4 to 6 pw. Gonidium protoplast in daughter colony at birth, 13 to 15 p. Number of somatic cells, 14,000 to 42,000; average 20,000 to 23,000. Number of daughter colonies, 1 to 16, usually 6 to 10. Number of sperm globoids, 108 to 214, probably more. Number of oospores, 60 to 201; average 160 to 170. Size of sperm globoid, 35 to 43 p. Size of oospore with spines, 44 to 59 2; without spines, 35 to 44 yw; spines, 5 to 10 p. 4. From Eliazer Pan (Potchefstroom) and Brakpan, both in the Transvaal. At Potchefstroom a similar Volvox with both male and female colonies was found associated with V. aureus (Plate IX, fig. A). From Brakpan no sexual colonies were obtained. 5. From Sedudu, Linyanti River, Bechuanaland Protectorate. In most of the vleis, etc., examined in this district the monoe- cious V. capensis was present. At Sedudu, however, in a small pan V. Rousseletit with typical male colonies containing numerous an- theridia (between 200 and 300) was found (Plate XXII, fig. A) associated with ovoid asexual colonies; the cell network appears finer than in V. capensis. V. Rousseletii was also found in some of the backwaters of the Linyanti River itself. 6. From Grahamstown, Cape Province. (Collected by Miss L. Britten.) The material was nearly all asexual, somewhat badly attacked by Rotifers. From the ovoid form of the large asexual colonies, and from the single young sexual colony with large oospheres and no sign of male elements, this Volvor is probably V. Rousseletiz, but further sexual material is necessary before it can be fully identified. The asexual colonies are large—of those measured, all immature, the dimensions ranged from 948 x 990 px to 1120 x 1184 w, with from 8 to 12 daughters, of which most had still not inverted. The single sexual colony measure 495 x 559 yx, and contained about 40 large bodies, apparently eggs which had rounded off but had not yet laid down their walls. VOle XVis PART 3. 30 460 Annals of the South African Museum. Protoplasts range in size from 5 to 8 y. The record is interesting, since it is the first from the Eastern Province. From the foregoing the Volvox which now follows appears to differ 5 AS ea — b \O AD ease: SS | Une | - Z an aN Fia. 5.—V. Rousseletii f. nov. kaokoensis. A, Asexual coenobium with 8 daughters. x 60. B, C, Somatic cells and gonidia (g) in surface view. x 1000. D, Sperm globoids and peripheral cells, (a) antheridial cell. x 900. in certain respects, chiefly in the form of the oospore, but we have examined only two samples (preserved). The material was mixed with Lemna paucicostata and a large Oedogonium, and was somewhat scanty. It was collected on two separate occasions from one locality. In the absence of fuller material we classify this tentatively as a form of the above. Observations on the Genus Volvox in Africa. 461 7. V. Rousseletii forma kaokoensis f. nov. Forma oosporis paullum minoribus, spinis conicis brevioribus quam in forma typica. From the Kaokoveld, South-West Africa. Collected by Dr. R. F. Lawrence in 1925, and by Dr. K. H. Barnard in February 1926, from a small rain-water pool on granite at Kamanyab in the Kaokoveld. The asexual colony is subglobose to ellipsoidal, 800 to 1600 yp in diameter, with the broadest part often a little anterior to the equator. The distance between the centres of the somatic cells is from 9 to 15 p ; protoplasmic connections are distinct (Plate X XIII, fig. D), and appear to be commonly four (sometimes five) in number (text-fig. 5, B) ; thus there may be a difference between these cells and those of the other forms described. Some markedly ellipsoidal daughters were observed (text-fig. 5, A), but these are the exception, the shape usually being subspherical (Plate XXII, figs. B and C). The number of daughters varies from 2 to 9, though 7 is the commonest; the size on escape is from 300 to 500 p, that is, rather large. The sexual colonies are dioecious. The male colony (Plate XXII, fig. D) is ovoid, with 120 to 150 or more antheridia which develop into typical depressed globoids of 512 sperms, 30 to 43 w in diameter (text-fig. 5, D, and Plate XXIII, figs. H and F). The female colony (Plate XXII, fig. H, and Plate XXIII, fig. A) is broadly rounded at the anterior end. From 120 to 227 oospores have been counted. The oospore wall is covered with conical spines, which may be straight or curved (Plate XXIII, figs. B and C). This form of Volvox differs from the preceding in the occasional elongated shape of its daughter colonies, possibly in the nature of its somatic protoplasts, and in the form of the ripe oospore. Dimensions of V. Rousseletia forma kaokoensis. Mature asexual colony, 525 x 600 to 1240 x 1260 (to 1600) p. Mature sexual colony, male, 657 x 730 to 876 x 905 pw. Mature sexual colony, female, 700 x 755 to 714 x 777 pw (few seen). Mature daughter colony, 294 x 315 to 379 x 481 wp. Somatic cell, 9 to 18 p. Somatic protoplast, 5 to 7 p. Somatic protoplast in daughter at birth, 4 to 5 w. Gonidium protoplast in daughter at birth, 9 to 13 p. Number of somatic cells, 21,000 to 30,000. Number of daughter colonies, 2 to 9, most often 7. Number of sperm globoids, 120 to 150 or more. 462 Annals of the South African Museum. Number of oospores, 130 to 227. Size of sperm globoid, 30 to 43 p. Size of oospore with spines, 44 to 46 4; without spines, 35 to 38 p; spines, 5 to 7 p. E. Volvox amboensis sp. nov. We proceed next to describe a Volvox we obtained from Ovam- boland ; this, though resembling V. Rousseletii in many ways, differs from it rather markedly, and we have decided to institute for it a new specific name. We have been able to observe it only in preserved material, and from one collection, and we have not been able to obtain satisfactory photographs, as the large coenobia are very fragile and readily collapse. The name amboensis is proposed to indicate its occurrence in the land occupied by the Ovambo tribes. Coenobiis asexualibus late ovoideis, 220-1300 uw (plerumque 800- 1000 j) latis, cellulis ca. 30,000-50,000, in parte posteriore densiore ageregatis ; coenobiis infantibus 1-14 (vulgo 8), interdum in coenobiis sexualibus inclusis. Coenobiis sexualibus majoribus, usque ad 2 mm. latis, bisexualibus ; ovoideis; antheridiis 31-33 y latis; oosporis numerosis (ad 700) arcte congestis, in ca. tertiam partem coenobiis carentibus (interdum in utroque polo absentibus), sine spinis 30-34 w latis, spinis cavis, conicis, acutis, ad 5-7 p longis obtectis. Collected by Dr. K. H. Barnard in February 1921, and by Dr. Barnard and Dr. R. F. Lawrence in February 1923, from Ongka in Ovamboland, South-West Africa. Dr. Barnard states that he visited the water-hole in which it occurred on two separate occasions, and that both times the water was green with the Volvox (no filamentous algae were present). The locality is about 15 miles north of Ondongua. This is a large and very variable Volvox with an astonishingly large number of oospores. The asexual coenobia are broadly ellipsoidal in shape and vary in size, a large one measuring 1200 x 1300 pp. The number of cells has been roughly computed as from 30,000 to 50,000 ; they are densely packed at the posterior end. The somatic proto- plasts are pear-shaped in side view (text-fig. 6, C). The protoplasmic processes are difficult to see (the materia] is preserved in alcohol) ; stained with gentian violet they appear as fine strands, sometimes anastomosing (text-fig. 6, B). The two cilia are inserted near together at the pointed end, which reaches nearly to the surface of the common envelope. There is one pyrenoid in each cell, large and nearly filling 2 6 e ge,8 & # x2 a8, 6° 8) 6 5S eine /%°,6 @®&e Fa © Fag verte ee R e792 0%" 86 o%® F @ 4 3 tat We ahert Seo @,°9 © 44°, a 2 isos 2 ag? ? s , “ace oo 5 7a." 8 oe @eSo oie ons teTet sews] Qo 52@o®8 get ae ® 5 OF Ce A We viene «6 2 Oye) ge 5) a5) e ote ame a Oe @ eo ® act GBa@s es © @ a Bee” 6% aoe @Qwe aR wre * SB @) yaa 8 Fic. 6.—V. amboensis sp. nov. A, Unusually small asexual coenobium with 5 daughters. x 90. B-D, Somatic cells: Band D in surface view, C in side view. B x1000; C, D x 700. E, Female coenobium. x 40. F, Female coenobium containing 4 young daughter colonies. x 30. G, An irregular coenobium containing 19 oospores, 3 sperm bundles, and 1 daughter colony. x70. 464 Annals of the South African Museum. the protoplast. The cells vary very much in size, not only according to age, but also according to the position in the coenobium. They are more closely packed in asexual than in sexual colonies. The distance between their centres may be twice as great at the anterior as at the posterior end (of a sexual colony)—20 to 10». The gonidia are late in developing ; a free coenobium measuring 640 x 680 2 was noticed with a few cells only just beginning to be differentiated. The daughter colonies vary in number; sometimes only 1 is present, though 8 is a common number. Several colonies were found with 9 daughters, three with 10, three with 12, and one with 14. On escape they measure from 135 to 200 1. Sometimes daughter colonies are found in coenobia containing male and female elements (text-fig. 6, G). The sexual colonies (text-fig. 6, E and F) are apparently monoecious and protandrous ; they may contain daughter colonies. The anterior end is sometimes more pointed than the other. They are ovoid in shape, often with only a slight difference in the length of the two axes (Plate XXIV, figs. Aand B). They are larger than the asexual colonies, some attaining a diameter of 2 mm., though others are quite small. The material to hand was over-ripe and male elements were rare, but a few scattered sperm bundles in the midst of ripe oospores were observed, 31 to 33 win diameter. Asa rule the oospores are crowded into about two-thirds of the coenobium ; sometimes a free space is observed at both ends (text-fig. 6, F). The number of oospores is very variable ; occasionally in an unusually small coenobium there are only 20—mostly there are several hundreds, even up to 700. This is a larger number than has been recorded for any previously described Volvox: in V. globator there are generally 30, though the number may be 64; V. Rousseletii has 200, V. Barberi about 224, V. perglobator from 300 to 400.* The oospore is densely covered with conical pointed spines, spirally arranged, about 16 or 17 visible round the periphery (text-fig. 4, E, and Plate XXIV, fig. C). The diameter without spines is from 30 to 36 4; the spines vary in length from 5 to 7 w; frequently the measurements are 30 ~ without spines, 40 ~ with spines. The oospores, it will be seen, are rather small. The spines are hollow, and are rather more pointed than in the forms above described (recalling those of V. Merrilli Shaw,t but shorter), the tip * Powers, J. H., Trans. Amer. Micr. Soc., vol. xxviii, 1906, p. 162. V. perglobator is comparable to the Ovamboland Volvox in size and in its large number of oospores, but is dioecious, and the somatic protoplasts form a reticulum, the cell-body becoming hardly noticeable. + Shaw, W. R., loc. cit., vol. xx, No. 5, 1922, pl. i, fig. 3. _ Observations on the Genus Volvox in Africa. 465 is often inclined a little to one side, the bases are narrow. Two oospores from one coenobium are shown in text-fig. 4, KE, drawn on the same scale as the oospores of V. Rouwsseletii and V. capensis to facilitate comparison. Dimensions of V. amboensis sp. nov. Mature asexual colony, largest seen 1200 x 1300 p. Mature sexual colony, 1100 x 1500 »—over 2 mm. Mature daughter colony, 135 x 200 p. Somatic cell: anterior pole 20 w; posterior pole 10 p. Somatic protoplast, 5 to 7 p. Number of somatic cells, 30,000 to 50,000. Number of daughter colonies, 1 to 14, most often 7 to 8. Number of sperm globoids—not known. Number of oospores, 20 to over 400. Size of sperm globoid, 30 to 33 p. Size of oospore with spines, 40 to 50 4; without spines, 30 to 37 2; spines, 5 to 7 p. This Volvoz is specially noticeable on account of its variability, the large size of the sexual coenobia, the mixed colonies, and the number of rather small oospores with pointed spines. In order to compare the species of Volvox referred to in this paper, some of their characteristics have been set out in parallel columns in the table on p. 466. SUMMARY. In the foregoing paper a short history of Volvox in Africa has been given. The occurence of V. aureus Ehrenb. has been noted in several localities from which it was not previously known, viz. the Transvaal (Kliazer Pan, Potchefstroom, and Weltevreden West), Rhodesia (N’gamo and Old N’gamo), and the Bechuanaland Protectorate (vleis and backwaters of the Linyanti River). V. africanus, G. S. West, which had been recorded previously only from Albert Nyanza, the Ussangu Desert, and the Philippines, has been described from pools in the neighbourhood of Kazungula on the Linyanti River (asexual only). A Volvox from the Cape Flats has been described as a new species, Volvox capensis, intermediate in character between V. globator and V. Rousseletti. A slightly divergent form of this species has been 71-9 W1-G “sourds jo qsua'T Autds W 1g-08| 00L<— Autds W ge-gg| 133-081 Aurds 103-09 Aurds F008 W FF-9E 9g Autds OST-06L Aurds 033-08 1 OF-98 W §c-OF Autds 7 6F-LE| 09T-09 Aurds F3S-G9 *Auids Og qnoqe AyTyensn W gg-GF| ‘F9-ZT WE qjoows W Gt-SF| 08-31 qjoous 1 $8-¢9] TI-L qyoous 7 ¢1-49| TI-> qjo0ouIs 701-09] &I-T ‘souids |*so10dsoo qnoyyt) Jo od44 alodsoo}| pure jo azig | Jaquiny | pure uL10,7 *‘jenxese | ‘papunol W ee-o¢ ueqy Ayyenba Sploqo[y) é May PCyer cia § eoue yqog|2 snorsaouoyy| eaoqge sy |7 QOZ-O9T| FI-1 aaoqe sy | OOgT-008) * ¢ * sasuaoquip + ATION194 *yenxase -sod 1! eF-08 ‘a10UL ueyy pequtod 7 (O09T-) SPIOqOTH}] 10 O : - _ may a 7 - a - ir - a - : Ppa mn : : . _ . 7 : My 7 —t i ve , oe 7 \ a . 7 a4 : a i > aaa F ol al ’ ie i, _ er : * i : ! nt) ie 7 7 . - ae - a ‘i =. on 7 —_ SOA iis : : - : ar WL = el Vital) Waass is 7 Ad ay uid eat - q i ; ; a Ah, i = a | i. 2 nf te! 0 ; : a ah “) ite =" cf : ; : : ' f i . , ¥ oe: 7 a a ‘ 7 . Ml “t, 7 * moat 0 n Plate IX. Ann. S. Afr. Mus., Vol. XVI. AFRICAN VOLVOX. SOUTH Neill & Co., Ltd. Es A, Photos M. Ann. 8S. Afr. Mus., Vol. XVI. Plate X. ‘S . * & E>» . few e ay _ & ~~ -~ ° ‘«: = @ @ 6 % 0 ® SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd, Plate XI. Afr. Mus., Vol. XVI. Ann. 8. VOLVOX. AFRICAN SOUTH Neill & Co., Ltd. PR. A. Photos M. Plate XII. Ann. S. Afr. Mus., Vol. XVI. a Erte S) ‘e oF 4 ey ‘@ > “my ‘ Ke nr ‘ SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd. £ ba. a8 ae. a Aun. S. Afr. Mus., Vol. XVI Plate XVII. SOUTH AFRICAN VOLVOX., Photos M. A. P Neill & Co., Ltd. Ann. 8. Afr. Mus., Vol. XVI. Plate XVIII. SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd, Plate XIX, Vol. XVI. Ann. S. Afr. Mus., SOUTH AFRICAN VOLVOX. Ann. 8. Afr. Mus., Vol. XVI. Plate XX. A SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate X XI. SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd. ey > : n ‘ he 1 1 + i = y ‘ ‘ ‘ his o = ‘ i 0 mt . i a ‘ 2 es, ~ os Ann. S. Afr. Mus., Vol. XVI. Plate XXII. SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate XXIII. SOUTH AFRICAN VOLVOX. Photos M. A. P. Neill & Co., Ltd. Plate XXIV. Ann. 8. Afr. Mus., Vol. XVI. AFRICAN VOLVOX. SOUTH Neill & Co., Ltd. A and By Prof. Iyengar. Photos: ae ( 473 ) 4. Volvox and Associated Algae from Kimberley.—By M. A. Pocock, Ph.D., F.L.S. With Field Notes by J. H. Powsr, F.Z.S., F.R.S.S. Africa. (With Plates XXV-XXXVII and 7 Text-figures.) CONTENTS. PAGE Part I. Introductory . : 5 : : ; : : . 473 Part Il. Description of the Bpecies é ° : : : : . 479 Eudorina elegans Ehrenb. : : : F : 5) cory) Volvox gigas sp.nov. . . ; : : : . 480 Volvox africanus G. 8. West . : F ; , j . 492 Volvox Rousseletii G. S. West forma griquaensis f. nov. . : ; é : . 498 Sphaerocystis Powerit sp. nov. : 3 : : : . 501 Kirchneriella africana sp. nov. : : : : 5 . 506 Part III. Field Observations and Records F ; : é : 2» 509 Summary. 5 ; ; : : : : : ; . 513 Discussion and Gonciusion : : . ; : 6 : : . 514 Bibliography : : : : : : : . ; : a {ellis Description of Plates. c : : : : : é é Os Part J.—INTRODUCTORY. Towarps the end of the year 1931, while examining pools in the neighbourhood of Kimberley for fresh-water crustacea, Mr. J. H. Power found Volvoz in abundance in a small quarry pool. The material collected was sent, with the crustacea, to the South African Museum and handed to the writer for investigation. The Volvox proved to be a large form, apparently of the type of Volvox Rousseletii West, but as it was purely asexual an exact determination could not be made. A request for further material, however, resulted in a second collection made early in December, in which there was abund- ant sexual material of all ages, the sexual colonies if anything out- numbering the asexual. The provisional diagnosis made from the first material was in the main confirmed; but although obviously very closely allied to the material from Rietfontein (Rich and Pocock, 1932, p. 455), which may be regarded as typical, it yet differed from it in some details. It was thought, therefore, that it might be VOL. XVI, PART 3. 31 474 Annals of the South African Museum. instructive to watch its development, and this was fortunately made possible by having on the spot a keen and experienced collector in the person of Mr. Power. When next visited, on the 22nd of December, the pool was com- pletely dry, but this rainy season in the Kimberley district proved to be a somewhat abnormally prolonged one, characterised by a succession of distinct downpours separated from one another by several weeks of dry, hot weather. Consequently the various pools dried up completely and refilled several times during the summer. During the succeeding three months (January to April 1932) Mr. Power kept the first Volvox pool under frequent observation, visiting it at intervals of two or three days whenever it contained water, and the present paper is the result of his observations and collectings. Although a very incomplete record of the seasonal history of the pool, the results are instructive, especially when compared with the facts known about the behaviour of Volvox in other areas, and are par- ticularly noteworthy as illustrating the extraordinarily rapid develop- ment of certain forms of Volvo in a region of intermittent summer rainfall, and the much greater rapidity of development under identical conditions found in some species than in others. Meanwhile search for Volvox was continued in other local pools without success, until late in the season, when it was found in two more small pools, called Pools II and III respectively in this paper. All three pools are formed in quarries from which building material has been removed, and are therefore probably of comparatively recent origin. Pool I is in typical red Kimberley soil, a mixture of clay and very fine-grained sand, drying very hard; owing to the fineness of the particles it forms a thick suspension in the water, which is consequently opaque and of a deep reddish colour throughout the existence of the pool—very unlike the clear water characteristic of Volvox vleis on the Cape Flats. In Pool II, on the other hand, the water is clear, the sides and bottom of the pool being formed of broken stone, while Pool III is of yet a third type, since it is in limestone tufa, of which material the sides and bottom are composed. The water is muddy and has a higher lime-content than usual in the district. A point which may possibly have some bearing on the behaviour of Volvox during its life-cycle, namely the Hydrogen-ion concentration of the water, has not so far been investigated. Keeping in mind the high day temperature, the rapid evaporation of the water, and the enormous quantity of animal life present, it is probable that the pH alters considerably and rapidly, and it is perhaps to this that we must Volvox and Associated Algae from Kimberley. 475 look for the explanation of a phenomenon already observed elsewhere, but as yet nowhere to such a marked extent as in these Kimberley pools, namely, the sudden and complete disappearance of Volvox which often occurs when the water reaches a certain degree of lowness, usually about 4 inches in depth, but long before the pool is nearly dry and while associated animal (mainly crustacean) and plant life continues unabated for some days longer. This disappearance is not heralded by any apparent falling off in numbers or development, but comes with remarkable suddenness, often apparently at the height of development of the strain. A very few colonies continue for a time, but the majority disappear completely in the course of a few hours. High temperature * alone does not seem sufficient explanation, nor can the crustacea be entirely to blame. To give an idea of the sequence of events it is necessary first to describe briefly the nature of the fauna and flora of the pool, or rather, of the plankton, as revealed by Mr. Power’s collections ; the attached algae, of which some at least certainly occur, were not collected. All collecting was done with a small net swept through the surface layers of the water, so that planktonic life alone, and at that only the larger species, was collected. The fauna consisted almost entirely of crustaceans of various kinds, Streptocephalids being particularly abundant. It would be interest- ing to correlate the appearance and numbers of the various species with those of the associated algae, but that is not possible here. In addition, a few Protozoa were noted; Rotifers were scarce as com- pared with other Volvox inhabited pools, but in one collection a few individuals of Proales parasita were found in colonies of V. Rousseletiz. The parasite never obtained a very great hold on the Volvox, however, probably owing to the brief duration of the pools. As regards the flora, besides two species of Volvox in Pool I it con- sisted almost entirely of colonial forms belonging to the Tetrasporales or Protococcales, with some Hudorina, while in Pools I and III Volvox only was found, two species in Pool II, one only in Pool III, the latter being the form of V. Rouwsseletii which is common to all three pools. In the case of the two latter pools, however, too few collectings could be made to establish much beyond the fact of the occurrence in them of certain forms of Volvoz. * A similar phenomenon was observed on the Cape Flats in November 1931, when such a sudden disappearance followed on a day of very great heat, when the water (3-4 inches deep) in the pool reached a temperature of over 32° C. at midday. 476 Annals of the South African Museum. The common Volvox is undoubtedly very near to V. Rousseletia West ; but since, as mentioned above, certain marked characteristics distinguish it from the typical form, it is described here as a form of that species to which the distinguishing regional name Griquaensis is applied. It develops very quickly and is associated in Pool I with a second species much slower in development. The young colonies of the latter were frequent in the third collecting and somewhat suggested V. africanus in appearance ; but whereas V. Rousseletii was already in an advanced sexual phase this second species was obviously still very young, with numerous large but still undivided gonidia. In succeeding phases it never appeared in the first collectings. In the first phase the pool dried up before any further stages could be obtained, but, when next formed, the pool had a longer existence and more advanced stages of the species were obtained in the later collec- tions. These proved truly startling—V. Rousseletii, previously regarded as an unusually large species of Volvox, was completely dwarfed by the maturing asexual colonies of the new species; the diameter of most of the colonies containing embryo colonies was well over 2 mm., while colonies of 3 mm. or over were not unusual. The first microphotograph (Plate XXV, A) shows the relative sizes well ; in the centre are an average-sized male colony and an immature asexual one of V. Rousseletii forma griquaensis, with a young colony, such as was first observed, of the giant, and round these are four maturing colonies of the latter, with fully formed daughters and undivided gonidia. The group, as in the case of the other low-power photographs reproduced, was floating in water in a glass dish and not compressed by its own weight on a slide. When placed on the slide such colonies flatten out enormously, appearing among associated V. Rousseletii colonies as large, clear, nearly colourless circular patches. It is considered that the name Volvox gigas can with justice be given to this magnificent species.* No trace of sexual organs could be found in this material nor in the succeeding collection during this phase of the pool’s existence ; but during the next phase, from the third collecting onward, Volvox gigas became more and more abundant until on 16th March it was the dominant organism among the plants, showing all stages in sexual reproduction. Thus by repeated collec- tions made at frequent intervals very rich material has been obtained and consequently as full a description can be given as is perhaps * A specific name in honour of its collector was barred by the previous existence of Volvox Powersii (= Besseyosphaera Powersi Shaw), Printz. See p. 492. Volvox and Associated Algae from Kimberley. 477 possible from the examination of preserved material only. Observa- tion of living material is of course needed to fill in details of habit and general behaviour, such as reaction to light, and of some phases of development, as, for instance, certain stages in the development of the embryo colonies, and the mode of escape on maturity. As regards response to light, since the eyespots in the anterior part of the colony are particularly large and well developed, there 1s no doubt that this species, like the other members of the Volvocinae, is strongly phototactie. Soon after the appearance of young colonies of Volvox gigas in the pool non-motile gelatinous colonial algae begin to be prominent in the phytoplankton. As development in the pool proceeds and the water grows lower, these increase enormously in amount, while the Volvox decreases, until finally it disappears altogether and the flora is composed entirely of these gelatinous forms. Thus during the existence of the pool there are several distinct stages, each character- ised by the appearance or dominance of a different alga or group of algae: first, Volvox Rousseletii develops in enormous quantities, then Volvox gigas appears, and sometimes, but not always, becomes domin- ant; about the same time the other colonial algae begin to get numerous, subsequently increasing greatly, until finally they alone survive. These colonial algae included at least four different forms, all of which at first appeared to belong to different genera. On further examination, however, three proved to be forms of one extremely polymorphic species which has been identified as belonging to the genus Sphaerocystis Chodat (1897a, p. 119). In its first mature form this is a striking and very beautiful alga, very unlike anything hitherto seen (Plate XXXVI, B). The colony is usually an almost perfect sphere with 16 or 32 (occasionally 8) ellipsoidal groups of large green spherical cells placed symmetrically within the periphery of the sphere. The regularity of the arrangement of these groups within the common envelope, the uniformity of the number and size of the cells composing the groups in any one colony, and their symmetrical spacing distinguish it from anything as yet described (Chodat, 18976, p- 293, emphasises the irregularity often seen in his species, Sph. Schroeteri), and it is therefore regarded as a new species to which the name Sphaerocystis Poweri, in honour of its collector, is given (Plates XXXVI and XXXVIL.) But in addition to this striking spherical form other forms appear, and, since there are transition stages present, all evidently belong to 478 Annals of the South African Museum. the same species. The second conspicuous form, which eventually becomes dominant, consists of irregularly rounded, lobed or elongated, sometimes branched masses composed of large numbers of rounded cells in a gelatinous matrix, very similar to a Tetraspora, but showing no signs of pseudocilia. As there has in the past been considerable difference of opinion as to whether similar masses found associated with Sphaerocystis Schroetert were actually a phase in the life of that alga or a species of Tetraspora (West, W. & G. S., 1912, p. 413) it was particularly interesting to find the two forms here, linked up by every stage in transition from one to the other. Finally, a third striking form, comparatively rare and only seen during fairly early stages of development of the pool, consists of botryoidal clusters of rounded aggregates of groups of very small cells, similar to very young spheres of the first stage. These have no firm exterior lamella to the gelatinous envelope, the presence of which cannot be detected until after treatment with some stain, e.g. iodine or methylene blue. When first seen such colonies suggested a Dictyosphaerium or Westella in general form, but no signs of con- necting strands could be distinguished. It is apparently inter- mediate between the spherical form and the Tetraspora-like phase ; the gelatinous envelope of a sphere in which the cells of the groups have already divided loses its firm outline, and the daughter spheres, instead of escaping separately, remain in contact with one another, without developing outer membranes. Continued division soon results in an amorphous gelatinous mass containing numberless cells such as is characteristic of the palmelloid Tetraspora-phase. Obviously a study of living material of this very interesting and somewhat puzzling alga is highly desirable. Mixed with the various forms of Sphaerocystis is yet one more arresting alga, a very large species of Kirchneriella Schmidle. It is characterised by the large size of the colonies, often over 2 mm., sometimes as much as 4 mm. long, and the great number of cells composing them. It is usually much less abundant than the Tetra- spora-form of Sphaerocystis which it much resembles in general form and appearance ; but in fairly early stages of the pool’s development (e.g. 8th February, see Table I) it may far outnumber the latter, and is then usually associated with comparatively numerous colonies of the botryoidal form of that alga. These gelatinous colonial algae can evidently thrive in quite shallow water in which the temperature at midday must be high, continuing to do so for some days after conditions have become Volvox and Associated Algae from Kimberley. 479 impossible for Volvor. Eventually, as the water dried they formed a compact gelatinous mass at one end of the pool. Further, if the pool refilled even after a very short period (in one case only three or four days) of dryness, V. Rousseletii developed again with great rapidity (cf. Phase 3), yet if rain refills it towards the end of a phase when the Volvox has already disappeared, but before the pool has quite dried up, this alga does not reappear. This is seen at the end of this same phase, when on 18th March light rain prolonged the life of the pool. Although collections were made for twelve days after that date, in all of which the non-motile algae were most abundant, only occasional colonies of V. gigas were found and no V. Rousseletit. This seems to indicate the need for at least a brief period of dryness to mature the oospores of the latter form of Volvor before they can germinate. Part II.—DEscRIPTION OF THE SPECIES. A. VOLVOCALES. I. Genus Euporina Ehrenberg. Eudorina elegans Ehrenb. Coen.. : ; . 53-215 pw x 62-258 pp Cellos e . : iy Coen. c. embryon. . 366 x 430 p Embryon. . ; . 44x53 pw Occasional, among Volvox, etc. in Pool I, particularly during the earlier stages. Except for its somewhat unusually large size this material seems to conform to typical #. elegans. Mostly large 32-celled colonies, often with daughter colonies fully formed, were seen, small colonies being very rare. In some, all the 32 cells had formed daughter colonies ; in others, several cells had remained undivided. It is probable that this species is actually far more abundant than the number of colonies noted would indicate, since the collecting was done mainly for Volvox and not for the finer plankton. 480 Annals of the South African Museum. II. Genus Votyox (Linnaeus) Ehrenberg. Section MrRRILLOSPHAERA (Shaw) Printz. 1. Volvox gigas sp. nov. (Text-figures 1-3, Plates XX V-XXX.) Volvox maximus, coenobiis asexualibus mixtisque sub-globosis, usque ad circa 3 mm. latis; cellulis rotundis, circa 1000-3000, sine connexionibus ; coenobiis infantibus permultis usque ad ca. 82, plerumque 30-60, interdum cum numerosis oosporis mixtis ; coeno- biis prorsus sexualibus minoribus quam asexualibus vel mixtis ; coenobiis masculis oviformibus usque ad circa 1 mm. latis, 1-3 mm. longis, vel rotundis, minoribus; antheridiis numerosis usque ad circa 600-700 in parte quarta anteriore absentibus, spermatozoidis longis, 16,32,64, vel 256,512 coalitis in globulis 18-22 yp, vel 40-49 p latis ; coenobiis femineis globosis usque ad 2-25 mm. latis, oosporis numerosis, globosis, membrana duplici, episporio levi, aliquanto crassi, endosporio tenue. Coen. asex. et mix.. . 1615-2800 p x 1700-3135 pw Coen. masc. . : . 132-1141 p x 797-1292 p Coen. masc. min. . . 246- % px242- 2% pw Coen. fem. . : . 1679-2154 p x 1723-2206 p Coen. infant. . ; . 108-388 p x 129-431 pe Cell. veg. : : . 26-84 p Pplast. veg. . : . 9-13 p; Stigma, 3-4 p Gonidium : . 35-49 p Glob. sperm. . . 18-22 pw; vel 40-49 p Spermatozoid. : . 24 pxI11-17 p; cilia, 14-19 p Cell. num... : . 1000-38000 Coen. infant. num. . . ad 82; pler. 30-60 Glob. sperm. num. . . ad ca. 600 vel 700 Oospor. num. : . ad ca. 300 vel 400 (?) Found in Pool I only. This remarkable member of the Volvox group is outstanding on account of : (1) its size, completely dwarfing the form of V. Rousseletiz —a fairly large one—with which it is associated ; (2) the large number of daughter colonies, as many as 82 having been counted in one coenobium, while 30 to 60 per coenobium, often associated with numerous oOospores, are common; (3) the slow development; the Volvox and Associated Algae from Kimberley. 481 gonidia are quite clearly differentiated from the somatic cells at a very early stage, certainly very shortly after inversion of the embryo colony, and continue to enlarge with the somatic cells within the parent and after birth, but do not begin to divide until the colony has reached a large size, usually over 1 mm., when the gonidia may be as much as 44 to 49 yw in diameter ; and (4) the structure and dis- tribution of the sperm bundles; in the posterior part of the male colony all the cells become antheridium mother-cells, 7.e. they are not mixed with somatic cells; the size of the antheridium mother-cell, and consequently the size of the resultant sperm bundle and the number of spermatozoids composing it, all vary greatly. Membranes and Cell Structure.—The membranes (fig. 1) are similar in arrangement to those of V. tertius Meyer and V. africanus West. In the recently inverted embryo the polygonal (usually hexagonal) protoplasts are in contact, but begin at once to separate from one another. At birth and for some time after, treatment with methylene blue shows the cell membranes as more or less hexagonal in outline (fig. 1, B) ; as development proceeds and the cells become further apart, each protoplast is seen to be enclosed in a wide, lenticular membrane, which is in contact with the outer limiting membrane over a circular or elliptical area above the protoplast (fig. 1, A, D). The membranes of adjacent cells are at first in contact, but in adult colonies they may be completely isolated from one another (fig. 1, E, K). Even in this stage, however, it is sometimes possible to distinguish faint hexagonal markings on the outer membrane of the colony (fig. 1, J). The spaces between the cell membranes stain faintly purple, while the membranes themselves are deep reddish purple. The appearance after treatment with methylene blue, as seen in surface view at different focal depths, is shown in fig. 1, F—H, while J and K show corresponding views from an older colony. In young colonies the peripheral gelatinous layer is wide, and more so in the posterior than in the anterior region of the colony, about 60 to 70 w at the anterior, 70 to 80 x at the posterior pole (fig. 1, A) ; as the colony gets older it becomes narrower and less regular (fig. 1, C), varying considerably in one and the same colony. It is bounded on the inside by a purple-staining membrane, similar to but more delicate than the outer membrane. The fluid filling the central hollow appears to be slightly mucilaginous, but shows no definite structure. The daughter colonies develop within the much enlarged gonidium walls, lying between the inner and outer membranes ; hence, prob- ably, the elongated shape of the young embryos (fig. 1, E). Fie. 1.—Volvox gigas, membranes. A, Young asexual colony with undivided gonidia showing inner and other envelopes and membranes of individual cells ; B, part of peripheral layer of very young colony at about time of birth; C, the same of a mature asexual colony, one embryo beginning to invert; D, E, portions of A and C respectively on a larger scale ; F—K, cells in surface view; F, top, G, middle, and H, bottom, focus of cells from colony A; J, K, top and middle focus of cells from colony C. A x40; C x100; B, D-K x 200; all approximately. Volvox and Associated Algae from Kimberley. 483 The somatic cells (fig. 2, A, B) are practically spherical, sometimes showing a very slight anterior projection ; they lie below the surface membrane suspended in the non-staining vesicle enclosed by the lenticular cell membrane. The extensive lobed chloroplast lines the whole wall except for a small area at the apex; there is a large basal pyrenoid from which several smaller pyrenoids later separate off, and a central nucleus; with methylene blue the whole proto- plast stains blue, the nucleus and pyrenoids more deeply than the cytoplasm. In the anterior part of the colony each cell contains a large lenticular orange-red eyespot (3 to 4 uw in diameter), which, as in all other colonial members of the Volvocinae as yet examined, occupies a definitely orientated position in the cell, 7.e. it lies on that side of the cell farthest from the anterior pole. The eyespot is usually rather low down, about midway in the cell (fig. 2, A). The cilia are inserted separately, diverging slightly before they reach the surface membrane. No indication whatever has been seen of protoplasmic connections in mature coenobia, although in colonies of V. Rousseletii in the same material they showed clearly. In young unborn daughters, however, here, as in V. africanus, occasionally a connection may be seen between adjacent cells. Such connections are rare and at the best of times obscure, at any rate in preserved material (but see later, in connection with inversion). Reproduction.—The reproductive cells are variously distributed ; a coenobium may produce only gonidia which develop into embryo colonies (Plate XX VI, C, D, E), only male or female organs (Plate XXV, B 1 and 2), or all three. As a rule, however, the sexes are separated, sperm bundles rarely occurring except in exclusively male coenobia, although mixed colonies producing daughter colonies and oospheres are common (in one case over 70 embryos and about 43 oospheres were counted in a single colony), and reach a large size (Plate XX VI, B, and Plate XXX, B). Purely female colonies, on the other hand, are little larger than the male colonies, but usually rounder in shape. The four types of colonies, 7.e. asexual, male, female, and mixed, and their relative shapes and sizes, are well illustrated in the group figured on Plate XXV, D. In this group, too, a few very young colonies are included. (a) Asexual.—As already stated, the gonidia are differentiated early, but undergo a prolonged period of enlargement before division begins (cf. Plate XXVI, A). Colonies of over 1 mm. diameter in which the gonidia show no sign of division are often seen. Structure and Development of the Gonidium.—When young the 484 Annals of the South African Museum. gonidium is distinguished from the somatic cells by its larger size and more numerous pyrenoids; the apex of the chloroplast is lobed, so that the colourless anterior area appears more or less star-shaped - Fia. 2.—Volvox gigas. A, Somatic cell from anterior pole; B, young gonidium and two somatic cells ; C—O, development of the embryo ; C, gonidium viewed from above ;_ D, first division; E, F, four-celled stage seen from above and in side view; G, eight- celled stage ; H—O, inversion of fully formed embryo ; H and J, preparatory stages in optical section ; K, general view ; L—N, actual inversion in process ; O, inversion complete, cilia appearing (L—O semi-diagrammatic). A,B x500 ca.; C-F x250ca.; H-O x350 ca. when seen from above (fig. 2, C). The cilia are lost early in the period of enlargement. As is the case throughout the group, the first two divisions of the gonidium are longitudinal, in two radial planes at right angles to one another. The separation of the apices of the resultant cells is apparent even in the 2-celled stage (fig. 2, D), while the hollowing to form a Volvox and Associated Algae from Kimberley. 485 bowl-shaped structure is well defined in the 4- and 8-celled stages (fig. 2, E-G), becoming more and more pronounced as development proceeds. Once division has started it proceeds rapidly to completion, so far as could be ascertained following the scheme of division already described by Janet (1923, p. 109 et sec., Plates VII to XVII) for Volvox aureus (Janetosphaera aurea) ; as in that species, the number of cells is comparatively small, despite the much greater size of V. gigas. Usually the number of cells is about 1000 (21°= 1024), z.e. there are probably at least 10 successive divisions in normal embryos, and not more than 12, since the highest number of cells counted was about 3000 per coenobium. Inversion.—Completion of cell division is followed by inversion of the embryo, resulting in the characteristic football-shaped daughter colony lying between the inner and outer membranes of the parent. In proportion to the number of embryos, the number of inversion figures seen in the material was extraordinarily small. This may be due to the fact that the process of inversion is a rapid one, or in part at any rate to the time of collecting, since the evidence in other species points to an optimum time for inversion about midday (Pocock, seq. p- 579). Inversion figures are commoner in some tubes than in others, though never very abundant. In spite of their comparative rarity, however, sufficient were seen to indicate the method of inversion which, as was to be expected, isa simple one, on the whole following the lines described for V. tertius Meyer by Kuschakewitsch (1922 and1931),* but in some ways more like that seen in sperm bundles of V. capensis. * Kuschakewitsch (1931, p. 328) states that the stages in inversion in V. awreus are so like those in V. tertius that a separate description for the former species was not necessary in his short paper. Janet (1923, p. 127, and pl. xviii) agrees with Kuschakewitsch, giving a single description for the process in the two species. Zimmermann (1925, p. 51), however, working with V. awreus, found considerable difference in detail in that species from the description given by Kuschakewitsch for V. tertius. Thanks to the kindness of Mr. Scourfield, samples of both V. aureus and V. tertius from Europe were available for comparison. These contain a few good inversion figures of which those in V. awrews include an early stage and the stages figured by Zimmermann as fig. 2, g and h, while in JV. tertiws stages corre- sponding to Kuschakewitsch, Taf. 20, B and E, and Janet, pl. xviii, C and D, were seen. Hence it seems probable that the descriptions and figures given by Kuschakewitsch and Janet must be regarded as applying to JV. tertius only, while those given by Zimmermann are correct for V. aureus. The fundamental difference in the two methods is that in the JV. tertius type of inversion actual inversion starts by the lip of the phialopore turning outwards, while in V. aureus inversion begins behind the lip, which remains doubled inwards until a late stage, when it finally straightens out. A comparison of fig. 2, H-N, with fig. 5, BF, will make this clear, since in V. africanus inversion is of a similar type to that in V. aureus. 486 Annals of the South African Museum. A colony with undivided gonidia and developing embryos, including several in various stages of inversion, is shown in fig. B, Plate XX VI. The earliest stage seen (fig. 2, H) shows a widening 4-sided phialopore with 4-lobed lip, depressed and curled inwards, while the whole embryo shows some denting and distortion. In the next stage (fig. 2, J, K) the denting disappears, a slight constriction forms, and the lip begins to straighten out and rise. The phialopore continues to widen and the lobes turn outwards, becoming very distinct in the process ; the cells at the edge are as a consequence in a state of tension, slightly separated from one another, and at this stage protoplasmic con- nections between the cells could be clearly seen (fig. 2, L, and Plate XXVII, A). These connections continue to show distinctly in the lobes as they fold back (fig. 2, M, and Plate XX VII, B, C). Character- istic ‘‘ hat” stages follow (fig. 2, N, Plate XX VI, F, and Plate XX VII, D), and finally the edges of the phialoporic lobes meet and inversion is complete (Plate XX VII, E). At first the phialoporic side is flatter than the anterior pole, the ends of the elongated cells are rounded and separated from one another, and the very short developing cilia are visible (fig. 2, O, and Plate XXVII, F). The cilia lengthen rapidly, the cells alter in shape, becoming shorter, wider, and more compact, and the common envelope is formed (Plate XX VII, G) ; enlargement follows, and soon the gonidia are distinguishable from the somatic cells (Plate XXX, D). The whole process of inversion is very like that observed in the germ colony formed on germination of the oospore of Volvox Rousseletii (ef. succeeding paper, p. 622); there, on completion of cell division the hollow globoid hes freely in a fairly wide vesicle much as does the young embryo of V. gigas, which, owing to the absence of intercellular protoplasmic strands, has no connection with the neighbouring somatic cells. The surrounding vesicle in the latter shows clearly in several of the microphotographs, e.g. Plate XXVII, C, D. This type of inversion would appear to be the primitive one characteristic of the simpler multicellular members of the Volvocales, in which the cell-number is several hundreds, but in general not more than 1000 to 2000. A very interesting point emerged from the study of the inversion figures in this species. In his description of V. tertiws, Meyer (1896, p- 200) states that although in the adult coenobium there are no proto- plasmic connections between neighbouring cells, yet in living unborn embryo colonies such connections are readily visible. In both V. gigas and V. africanus protoplasmic connections can occasionally be seen Volvox and Associated Algae from Kimberley. 487 in embryo colonies before escape, but are entirely absent in the adult. At certain stages of inversion of the embryo in V. gigas, as described above, relatively stout protoplasmic connections between the cells became obvious in those parts of the embryo which are being stretched, and are certainly present throughout, although owing to the small size of the embryo and the close packing of the cells, except at the edges, they are not distinguishable. Now it must be remembered that the cells of the young embryo are not enclosed in a common membrane, but lie in close contact with one another. During inversion the cells act in concert with one another, but each is at the same time altering its shape and must to some extent act independently ; while some are being compressed, others are in a state of great tension, tending to separate from one another. Obviously, since the globoid inverts as a whole, the connection between the component cells must be very intimate. It would appear that this is obtained by the presence of protoplasmic connecting strands, which therefore play an important role in the process of inversion, on completion of which they become functionless and are withdrawn, one only surviving here and there to a later stage in development. A further advance in complexity of structure is reached by those species which retain the protoplasmic connections throughout their life, e.g. V. aureus, while the members of the globator group have progressed yet a step further, since in these the connections develop as the colonies mature, reaching their maxi- mum thickness when the reproductive cells are developing. Development of the Embryo subsequent to Inversion.—The embryos are very small at the time of inversion (e.g. 58 x 66 uw, 71 x53 yw) and enlarge considerably before birth (Plate XX VI, E). The component cells very soon begin to separate from one another, enlarging as they do so, while the gonidia now become very distinct from the somatic cells (Plate XXVI, E, and Plate XXX, D), their diameter at this stage being nearly double that of the latter ; for example, in an embryo measuring 77x87 the vegetative cells measured 3 to 4, the gonidia 5 to 6 yz. As the embryo develops the proportion between the two types of cells remains approximately constant, until the somatic cells reach their maximum size, after which the gonidia still continue to grow until finally they are three times the diameter of the vegetative cells, or even more. Young oospheres are considerably smaller and increase in size more slowly, while the majority of the antheridium mother-cells do not grow much larger than the somatic cells until late in development (Plate XXVIII, C), although in some male colonies a few here and there enlarge similarly to the gonidia 488 Annals of the South African Museum. (Plate XXVIII, A, B). The relative sizes of recently inverted and mature embryo colonies can be seen in Plate X XVI, E. In addition to the typical ellipsoidal embryos composed of over a thousand cells there are often scattered among them others, usually rounded, composed of far fewer cells, in some cases as few as 64, but more often 128, 256, etc. ; in these the majority of the cells enlarge proportionately more than in typical embryos, e.g. in one such colony, already free, measuring 58 pw in diameter and composed of 128 cells, the majority of the cells measured 10 , while a few at one end were smaller. Such “ dwarf” colonies appear to be male colonies in which nearly all the cells function as antheridium mother-cells ; the develop- ment of the sperm bundles, however, does not take place until after escape, and such small males are rare compared with the larger egg- shaped male colonies. The small colony shown in Plate XXV, C, contains young male colonies of this type and a few oospores. (b) Sexual.—(1) The male reproductive cells are nearly always found in exclusively male colonies, which are slightly smaller than the purely female colonies and which often become markedly egg- shaped as the sperm bundles develop (Plate XXV, B), or more rarely they may be such rounded dwarf colonies as are described above. Occasionally, however, sperm bundles have been found in female or mixed colonies, where one or two sperm bundles were scattered among the developing oospheres. As, however, such bundles are small and easily mistaken at low powers of magnification for oospheres just beginning to enlarge, it is quite possible that they may be of more common occurrence than would appear to be the case. In the male colonies all the cells in the posterior half become antheridium mother-cells (Plate X XIX, A), a small anterior portion remains somatic, while between the two lies a zone where male initials and somatic cells are mixed (Plate XXVIII, C). Thus some three- quarters or more of the total number of cells in the male colony may develop into sperm bundles. The antheridium mother-cell varies enormously in size—it may divide on reaching a diameter of 18 to 20 yp, or division may be delayed until it is more than double that size, 7.e. as large or nearly so as a mature gonidium. The number of divisions undergone depends on the initial size of the cell; most often it is 5, resulting in a bundle of 32 sperms, but bundles of 16 and 64 occur, while in many male colonies scattered among these smaller sperm bundles are others very much larger, containing a much greater number of spermatozoids—256, 512, or possibly even more. The latter are not unlike sperm bundles of Volvox and Associated Algae from Kimberley. 489 V. Rousseletti, in which closure has not been complete and which have not yet undergone subsequent compression (fig. 3, H, J, and Plate XXIX,E). Inside view such bundles, like the smaller ones, are bowl- shaped, almost globular, but flattened at the open side. Thus in one and the same colony sperm bundles of several different sizes may occur (Plates XXIX, A, C, and XXX, A). Stages in the development of sperm bundles are shown in fig. 3. Division of the antheridium mother-cell (fig. 3, A, B) follows the usual lines, resulting in the formation of a more or less rounded hollow bowl or globe according to the number of divisions undergone ; where the divisions are few the resultant structure is more open and bowl- shaped, or, even in the case of the smallest bundles, saucer-shaped, while the very large type is almost globular (fig. 3, C). On completion of cell division inversion takes place; although few stages could be made out, sufficient were seen to indicate the type of inversion. In the case of the smaller bundles it is similar to that seen in Hudorina, while in the larger, many-celled bundles it is naturally rather more elaborate, approaching that seen in the case of the asexually formed embryos. Fig. 3, D, shows a medium-sized bundle in which the phialopore is opening preparatory to inversion, while fig. 3, E, shows a recently inverted small bundle with elongating cilia. Fig. 3, F and G, show optical sections of bundles of 64 and 128 spermatozoids re- spectively, while the largest-sized bundles are shown in different views in fig. 3, Hand J. An early stage in the development of such a bundle is shown in Plate XXIX, fig. B, in which the second cell division is not quite complete ; the separation of the cell apices is particularly clear, while the surrounding vesicle is faintly discernible. The adjoining microphotograph (EK) shows the appearance of the mature bundle. Examples of male colonies are illustrated on Plate XXVIII. Of the four male colonies depicted, one (C) contains only small and medium-sized sperm bundles, while the other three all contain a number of the largest sized as well. These appear in the photographs as very dark, rounded bodies. Of the smaller bundles many are in early stages of division, others are inverting. Colonies A, B and C are of the characteristic egg-shape; D is more nearly spherical. Plate XXIX, figs. A, C, and D, show portions of male colonies further enlarged to show details of structure and development of sperm bundles of various sizes. The spermatozoids (fig. 3, K) are found in great numbers swarming on or within the peripheral layer of colonies containing oospheres VOL. XVI, PART 3. 32 490) Annals of the South African Museum. Fia. 3.—Volvox gigas. A—E, development of Androgonidium. A, two-celled, and B, four-celled stage; C, cell division complete; D, bundle inverting (phialoporic view); EE, inversion complete, cilia lengthening ; F, G, optical sections of bundles of 32 and 128 spermatozoids respectively ; H, largest-sized bundle (512 spermatozoids) viewed somewhat obliquely from below ; J, phialoporic view, optical section ; K, spermatozoids found swarm- ing in female colony ; L, M, spermatozoids swarming round oospores; N, ripe oospore. 8, eyespot ; n, nucleus; c¢.v., contractive vacuole. K x 2000; other figures x 500; all approximately. (Plate XXX, C, D). They are long (11-17 2) and of rather distinctive form; as a rule there appears to be a central wider region (3-4 p) with an elongated nucleus and narrower anterior and posterior Volvox and Associated Algae from Kimberley. 49] portions. The colourless beak ends in a blunt point, the cilia, usually a little longer than the spermatozoid itself (14-19 jx), being inserted at opposite sides of the apex and directed sideways and forwards, not backwards as in the globator and aureus types; in this respect they resemble the spermatozoids of Hudorina elegans. At the point of insertion is a refractive granule. At the apex of the chloroplast and near the base of the beak is a small red eyespot and adjacent to it are two contractile vacuoles. Most often only one vacuole could be distinguished, hence their periodicity is probably more or less alter- nating. The narrow posterior part may be prolonged into a delicate point or it may be rounded ; probably the posterior process may be withdrawn or prolonged at will during the movement of the sper- matozoid, as is the case in Hudorina. From the contorted shape of many of the spermatozoids the movement is evidently a combined amoeboid and ciliary one. In addition to the normal spermatozoids numerous long-drawn out, structureless ones, particularly common in older colonies, had evidently died before the material was fixed (Plate XXX, C). In developing and mature sperm bundles starch is present, but none could be detected in the free spermatozoids. They are rather longer than usual in Volvoz. (2) The female reproductive cells develop directly into oospheres ; at first these can be distinguished from neighbouring somatic cells only by the presence of more numerous pyrenoids, but the size con- tinues to increase and the cilia are soon lost; mature oospheres vary a good deal in size, and are practically spherical in shape. Spermatozoids swarming round the oospheres are common (fig. 3, L, M). The oospores are comparatively small (31-44 w), with smooth fairly thick exospore, within which, when ripe, the protoplast, enclosed in the delicate endospore, les somewhat eccentrically (fig. 3, N; Plate XXX, C). When ripe they are reddish gold in colour—even in preserved material the large colonies with ripe oospores glitter with a golden sheen when viewed by reflected light. The number per colony is large, certainly several hundred, but probably not so great as the number of androgonidia in the male colonies. In mixed colonies the number varies enormously. Purely female colonies are in general smaller than the mixed ones, which attain nearly, if not quite, as great a size as the purely asexual ones (Plate XXV, D). From the above description it will be obvious that Volvox gigas is a very distinctive, if somewhat primitive, species. The only form as yet recorded which approaches it is that described by Powers 492 Annals of the South African Museum. (1907, p. 123) * without name, as a “ second form” of Volvox, from Nebraska. Subsequently Shaw (1916, p. 253) named it Besseyosphaera Powersi (= Volvox Powersii Printz, p. 59), the asexual colonies alone having been observed by Powers; apparently, since there are no further references to it in the literature, it has never again been recorded. It resembles the present species in (1) the great size of the colonies, often over 2 mm. and up to 2-5 mm. in diameter, and (2) the large number of daughter colonies, up to 78 being recorded. Further, judging from Powers’ remarks it is slow in development. But, according to Shaw, the gonidia are not differentiated until after birth, and on this feature, combined with the absence of protoplasmic connecting strands, Shaw founded his new genus Besseyosphaera, which he regarded as intermediate between Pleodorina Shaw and Volvoz. Powers found the asexual colonies in a pond which was deeper than those in which Volvor usually occurs in Nebraska, and “ found nothing resembling it ’’ subsequently (Powers, 1908, p. 142). It seems possible that on a fuller examination it might prove to be identical with the Kimberley material described here as Volvox gigas. On the existing descriptions, however, it must be regarded as distinct. The resemblance of V. gigas to Pleodorina is certainly very remote ; it is unmistakably much more nearly allied to Volvox than to Pleo- dorina, and seems to be a somewhat primitive form belonging to the section Merrillosphaera (according to Printz, genus according to Shaw). Powers’ photographs of other species of Volvox are so excellent that it is most unfortunate that he apparently published none of either this ‘“‘ second form ” described in his first paper or of another outstanding species which he named Volvox perglobator and which closely resembles V. Rousseletii G. S. West. The field observations given below (pp. 38 to 39) show how much slower in development V. gigas is than the associated V. Rousseletiv. 2. Volvox africanus G. 8. West (1910, p. 102, and 1918, p. 1). (Text-figures 4 and 5, Plates XXX (fig. E) to XX XIII.) Asexual colonies in all stages of development. No sexual colonies seen. Mature colonies large, over 1 mm. in length and nearly 1 mm. in width. * Unfortunately, Powers’ original paper has not been available for reference ; the substance has had to be gathered from references by other authors, particularly Shaw, and from Powers’ own remarks in his subsequent paper (1908, p. 142). Volvox and Associated Algae from Kimberley. 493 Daughter colonies most often eight, arranged symmetrically in pairs, two pairs nearly equatorial, two nearer the posterior pole. The pair most nearly equatorial develops first, other pairs develop successively in order of their position, the most posterior developing last. Anterior pairs may have inverted and may contain developing gonidia before cell division in the posterior pair is complete. In some of the young coenobia the four central daughters were in early stages of division, while two pairs of undivided gonidia still in the stage of enlargement and 35 « and 28 pw respectively in diameter occupied positions near the posterior end. Sometimes fewer, rarely more than eight, daughters develop ; colonies with six, five or even fewer embryos were observed, while in one case there were nine, an extra one having developed near the equatorial plane. If an odd number develops, an abortive gonidium representing the missing member of the pair is usually, but not invariably, traceable. The protoplasts of the somatic cells are rounded, without proto- plasmic connections. In the developing embryo, however, inter- cellular protoplasmic strands are present until inversion is achieved, and occasionally one here and there survives and can be seen in the unborn embryo. Coen. asex. . : : . 9366-926 pw x 387-1077 pw Embryo, inverting . : BT SY cs) 7 Embryo, mature. : . 366x451 pe Gonidium, mature . : . 28-44 p Gonidium in inverting embryo. 9-16 pw Somatic cell . : - . 16-20 p Somatic protoplast. : . OTp No. of cells. : . 3000-7000 Found in Pool II only ; rare, associated with V. Rousseletit. This species was found in Pool II late in the season, after the last rains. It was present in very small quantities mixed with numerous very large, vigorous, but young asexual colonies of V. Rousseletii ; not a single sexual colony of either species could be found, although the pool was visited several times until it was completely dry. It is to be hoped that next season it will be possible to keep this pool under observation for a longer period and that sexual material will be obtained so that the description of the species may be amplified, since West did not obtain satisfactory male material, and even Shaw, in his extended description (1923, p. 197), does not describe the male reproductive organs fully, and differs from West in some details. 494 Annals of the South African Museum. The species is very distinctive, in some ways the most beautiful of all the African species of Volvox (Plate XX XI, A-D). Its fairly large ellipsoidal colonies with symmetrically arranged pairs of embryo colonies are quite unmistakable, even in the absence of sexual reproductive organs. The two poles are usually equally rounded, though occasionally one, usually the anterior, but sometimes the posterior pole, may become slightly broader than the other. The Kimberley material includes colonies considerably larger than any yet described, and, in spite of its rarity, there was sufficient material to add several details to our knowledge of the species. It is therefore described here in some detail. Membranes.—Diagrams showing the form of the membranes in this species have already been published (Rich and Pocock, loc. cit., Fic. 4.—V. africanus, treated with dilute methylene blue to show the membranes. A, Young colony, optical section, polar view; B, older colony after escape of equatorial embryos. Side view. x about 45. p. 432, fig. 1, B). As, however, no detailed description of them was given they are briefly described here. As in Volvox gigas and Volvox tertius, each cell is enclosed in a lenticular membrane which is united with the outer common envelope above the protoplast. The adjacent cell membranes are often sepa- rated from one another, but never to so great an extent as in the mature colonies of V. gigas. The inner limiting membrane is fairly distinct and regular in young coenobia, the developing embryos in their surrounding vesicles lying in the peripheral layer between the inner and outer membranes (fig. 4, A). As the embryos develop, however, they grow very large and push aside the inner membrane, displacing it considerably Volvox and Associated Algae from Kimberley. 495 (fig. 4, B). The microphotograph reproduced in Plate XX XI, C, shows a mature colony from which the equatorial embryos have already escaped, the respective pores of escape of two of them showing clearly at the two sides near the equator. The posterior embryos, of which three are seen, have continued to grow after the others had escaped and nearly fill the posterior half of the parent. The membranes surrounding the embryo colonies show clearly, and the inner mem- brane of the parent can just be distinguished above the latter and in the anterior part of the coenobium. The colony had been treated with methylene blue and was photographed free on the slide to avoid undue distortion. The same colony, unstained, appears in the group (Plate XXXI, A) just above. The position of the embryos between the two membranes in J. africanus, as in V. gigas, probably accounts for the lateral compression and elongated shape often characteristic of young embryos; but in the former species they attain a very much greater size in proportion to the parent than in the latter, and soon become wider than the peripheral zone, hence the great distortion of the inner membrane. Inversion of Embryo Colonies.—Inversion figures are not abundant in the material, but in proportion to the number of embyro coenobia examined they are far more numerous than in V. gigas, and what there are are excellent, particularly in the iodine-fixed material, showing no appreciable distortion. Probably the process is a much lengthier and more complicated one than in that species; this was to be expected from the fact that the cells are far more numerous in proportion to the size, and further, that the gonidia are already large before inversion takes place. On the completion of cell division, the gonidia, particularly the two equatorial pairs, are large and project far inwards from the peripheral layer of somatic cells. In one inverting embryo the size of the gonidia varied from 9 yp at the posterior (phialoporic) end to 16 yx at the equator. So far as could be made out from the available material the stages in inversion are as follows :-— There is a well-marked preparatory stage characterised by the infolding of the four lobes of the phialopore and more or less extensive denting such as is seen in V. Rousseletii and V. capensis, etc. (fig. 5, A, B). At this stage the vesicle sometimes has a curious dumb-bell shaped appearance. The phialopore enlarges considerably without the lip becoming straightened out, and a constriction appears some- where between the equatorial and the posterior gonidia; the upper (outermost) part, still double, folds back (fig. 5, C), and the “ hat ” 496 Annals of the South African Museum. Fic. 5.—V. africanus. Inversion of the embryo (diagrammatic). A, B, Preparatory stages, showing denting and incurving of lip of phialopore, and position of gonidia before inversion; C, early stage of inversion, infolded lip curving back; D, side view of the same colony, slightly tilted, on a smaller scale; H, a later stage, the infolded portion opening out; F, G, inversion nearly complete, but phialopore not yet closed; F, as seen from the side, G from above; three pairs of gonidia of different sizes can be seen projecting from the surface of the inverted embryo; H, detail of relation of gonidium to the somatic cells (1) directly after inversion, (2) later when in-sinking of the gonidium has begun, and (3) in final position. D x150; others x 260, approximately. Volvox and Associated Algae from Kimberley. 497 stage begins (fig. 5, D, Plate XX XIII, A, B). In Plate XXXITI, in A a gonidium shows at the angle of the fold on the right, B is of the same coenobium seen with slightly different focus, while C shows the phialoporic view; the four-sided opening is clear, with projecting gonidia visible on two sides, while a larger one shows in the fold at the top on the right. The infolded lip now begins to straighten out, forming a wide fringe round the centre of the embryo, nearly filling up the lenticular vesicle. Plate XXXII, E, shows this stage as seen from above, fig. 5, EK, and Plate XXXII, D, as seen from the side. Again, the four-angled open- ing and large gonidia can be seen. The other member of this pair of embryos, nearly at the same stage of development, is shown in the microphotograph C, Plate XXXII. In the next stage found inversion was almost complete and the phialopore beginning to close (fig. 5, F, G, and Plate XXXIII, E). The opening is irregularly lobed, the four lobes being of somewhat different sizes. The embryo is now a very striking and peculiar looking object ; in polar view it is four-sided, with rounded angles. From the centres of the four sides the relatively huge gonidia project, the narrower inner end being sunk in the peripheral layer of cells, while the broad end now projects outside the embryo (Plate XXX, D, and Plate XX XIII, D). The position of the gonidia is readjusted by the gradual insinking of the gonidia until the anterior ends are on a level with those of the somatic cells (fig. 5, H, 1 to 3, and Plate XXXIII, F). In one large colony (Plate XXXII, A and B) the two posterior daughters were inverting (Plate XXXII, C and D), of the next pair one was in the stage of projecting gonidia (Plate XXX, D, and Plate XXXII, F), while in the other the gonidia had already sunk into their final position. In the four equatorial embryos the central gonidia had already started dividing, while the posterior pairs were still enlarging. Microphotographs showing several corresponding stages, notably ‘hats’? and recently inverted daughters with projecting gonidia, are given by Powers (1908, pl. xxiv, figs. 25, 28 to 32) for Volvor Weismannia (Merrillosphaera Carteri (Stein) Shaw, var. Wersmannia). Good as these few inversion figures in V. africanus are, a study of living material is essential for a full account of the process of inversion in this species ; it seems, on the whole, to approximate to the type of inversion described for V. aureus by Zimmermann (1925, p. 51; see note above, p. 485), but even these few stages in preserved material 498 Annals of the South African Museum. are sufficient to show that there are several differences in detail, and the whole process of inversion is complicated by the large size of the gonidia. The two new records for V. africanus, i.e. from the Linyanti River (Rich and Pocock, loc. cit., p. 441) and the present one from Kimberley, are interesting, extending its range in Africa considerably. It is evidently widespread in the warmer regions of the world, having now been recorded from Albert Nyanza, Tanganyika, Rhodesia, and Kimberley in Africa, the Philippines, and 8. India (Iyengar, 1933). So far, it seems to be nowhere abundant. Possibly, like V. gigas, it is slow in developing and undergoes a prolonged asexual phase followed by a relatively brief period of sexual activity. It will be especially interesting if it proves possible to study it in life, since so far, apparently, all descriptions, even Shaw’s, have been made from preserved material only. The Kimberley material is outstanding both from its larger size and the very regular development of four pairs of embryos; in the earlier collectings two and four have been the usual numbers of embryos, with six occasionally, although Shaw, 1923, p. 197, found in the Manila material a small proportion of colonies containing eight. Section Eu-Votvox Printz. 3. Volvox Rousseletii G. S. West. Forma griquaensis f. nov. (Plates XXXIV to XXXV, A to E.) Forma coenobiis sexualibus majoribus quam in forma typica, antheridus oosporisque permultis. Coen. asex. ; . 710-1809 pw x 904-1900 pw Coen. masc. . 471-1443 p x 559-1550 p Coen. fem. : : . 600-1464 pw x 644-1636 pw Coen. infant. . . $494 x ? -517p Cell. veg. . : ; 3 C-l5u Pplast. veg. ; : =e One Glob. sperm. . ; . 385-41 px 18-22 p Oospor.. ; ; . 40-53 p Oospor. sine sp.. . 935-40 pw Spin. long. : , . 5Dp Cell. num. ; : . ad ca. 43,000 Volvox and Associated Algae from Kimberley. 499 Coen. infant. num. . . 419 Glob. sperm. num. . . #800 Oospor. num. . : . %655 Coen. juvenil. . ; . 215-322 p x 258-366 uw Coen. juvenil.embryon. . 129-215 uw x 139-240 p Coen. juvenil. cell. num. . ca. 330-418 Found in all stages in Pool I; asexual only in Pools IT and III. This form combines characteristics of various forms of Volvox Rousseletii and its allies—the very large number of sperm globoids (Plate XXXV, A, B) suggests the material from Ussangu (West, 1918, p. 2), while in the unusually high number of oospores (Plate XXXV, D, E), often several hundreds, it resembles Volvox amboensis (Rich and Pocock, loc. cit., p. 462); on the other hand, the sexual colonies, although on the whole larger than usual, do not reach anything like the size attained in the Ovamboland material, and very large asexual colonies are on the whole rare and slightly below the maximum size observed in the Rietfontein cultures. It is typically dioecious, the male colonies (Plate XX XV, A, B, D) being better developed and more numerous than in any other form as yet examined; but an occasional sperm-globoid in otherwise female colonies is not unusual, this again suggesting a resemblance to V. amboensis. The oospores (Plate XXXV, F) are usually slightly smaller than in the Rietfontein material, with shorter spines, and are intermediate between that form and the form from the Kaokoveld (forma kaokoensis, Rich and Pocock, p. 460). The asexual colonies are often markedly egg-shaped, particularly in the purely asexual material from Pools II and II which is in- cluded here, although sexual colonies are lacking. In the more egg- shaped coenobia the embryo colonies are nearer the posterior pole than usual, even the most anterior being definitely posterior to the equator. In purely asexual material the number of embryo colonies is high, usually over eight, as many as 19 having been counted, while 12 to 14 are common. During the sexual phases the asexual colonies are smaller and contain fewer daughters, usually under eight (Plate XXXIV, A, B). The Juvenile form (Plate XXXIV, G, E) was collected in great numbers on one occasion, and is similar to that of V. Rousseleti and of V. capensis, but is slightly larger, composed of rather more cells, and usually produces two or three embryos (Plate XXXIV, C), while as many as five were seen. 500 Annals of the South African Museum. The most striking characteristic of the form, however, is the great rapidity of development; this is brought out clearly in the field notes given below (p. 509), but the most marked and at the same time most complete record may be summarised here :— The pool had dried up completely by 24th February 1932. Three days later heavy thunder-showers during the night of Saturday, 27th February, filled the pool once more. The following Wednesday (2nd March) a fine crop of Juveniles, with a few very young second- or third-generation colonies, was obtained. Two days later (7.e. in less than a week after the rain) among large asexual colonies containing up to 10 embryos were a number of sexual colonies, mostly young (Plate XXXIV, A), while by 7th March it had reached the height of the sexual phase, and a week later had almost disappeared, although the pool did not dry up until some time later. Obviously it is a form well adapted to the local conditions, which are such that any pools formed can have only a short existence, although they may form several times during a single season. From observations made on the Cape Flats during last year one is forced to the conclusion that a relatively high temperature is beneficial to both the rate and degree of development, provided the water does not fall below a certain depth. The behaviour of Volvor in the neighbourhood of Kimberley is very suggestive of the notes made by Powers (1908, p. 141) in Nebraska, where conditions appear to be very similar to those prevailing in some parts of Africa where there is a summer rainfall. V. Rousseletii was found in all stages in Pool I each time it filled, and after the last rainfall in asexual stages only in Pools II and III. The identification of the latter is therefore only provisional; at present, however, there is no reason to regard the latter material as at all distinct from that from Pool I. Other pools in the neighbour- hood were examined for Volvox but so far without success. Volvox and Associated Algae from Kimberley. 501 B. TETRASPORALES, PALMELLACEAE. Genus SPHAEROCYSTIS Chodat.* Sphaerocystis Powert, sp. nov. (Fig. 6, and Plates XXXVI, XXXVILI.) Familiae formae pervariantes, aut (1) familiae plerumque conspicuo perfecte globosae in tegumento gelatinoso hyalino vel sub-flavo involutae ; cellulis 32, raro 8 vel 16, regulariter ad peripheriam dispositis, plerumque omnibus extemplo partitis in sub-familiis similis ellipsoideis 8- (raro 4-), vel 16-, vel 32-cellularum compositis, ad peripheriam aequabilitissime dispositis ; aut (2) familiae magnae amorphae, interdum fere rotundae, plerumque elongatae irregularitae, fere teretes; cellulis permultis, rotundatis, parvis, in tegumento gelatinoso hyalino ; zoosporis utrobique parvis vel paullum majoribus, ovatis, interdum leviter obliquis; aut raro (3) familiae botryoidae, sub-familis pluribus, cellulis minimis in tegumento gelatinoso hyalino irregulariter lobato consociatibus. 1. Fam. glob. diam.. . 140-1615 p Cell. diam. . ; . 6-16 w; plerumque, 16-20 p Subfam. 22210 we xD Ie OB at Zoospor. . ; - 4-5 xi1-9 ye; vel Su x10 ay 2. Fam. palmell. long. . ad 2-5 mm. Cell. diam. . ; . 4-10 (-13) p Zoospor., Vv. supra. Found in Pool I only, most abundant in the later stages of each phase, finally forming the bulk of the phytoplankton as the water dried up. In the early stages of the pool’s existence, among masses of Volvox Rousseletii_ colonies, small rounded Hudorina-like gelatinous colonies containing usually 32, occasionally 8 or 16, cells are fairly common (fig. 6, A). They are distinguished from the Hudorina colonies which are present in the same material by (1) absence of cilia, (2) * Wille (1903) and Lemmermann (1915) sink this genus in Gloeococcus A. Braun, but Chodat (1904) contested this, and West (1916, p. 186) accepts Chodat’s genus as distinct. As it does not appear that any fresh facts have been added to the knowledge of the two species constituting Braun’s genus (cf. Lemmermann, loc. cit., p. 32) which are still described as “‘incompletely known,” and in which there is apparently nothing whatever to correspond to the very characteristic spherical form described by Chodat, West is followed here, and Chodat’s genus accepted (see p. 505). Oltmanns (1923, vol. i, p. 243, etc.) also accepts Chodat’s genus. 502 Annals of the South African Museum. thicker but less regular outer membrane with numerous attached granules of foreign matter, (3) more closely adpressed cell membrane round the constituent cells, and (4) different behaviour when treated Fic. 6.—Sphaerocystis Powerit, sp. nov. A, Eudorina-form ; B, typical spherical colony ; C, botyroidal form ; D, transition from spherical to Tetraspora-form ; E, Tetraspora-like colony ; F, cell from spherical form; G—H, motile stage of either spherical or palmelloid form ; G, division of a cell to form two zoospores; H, mature zoospores—a, b, smaller type seen from the side and from the back respectively; c, larger, more rounded type. A x100; B,D x35; C xl50; E x30; F—H x1000: with methylene blue, the mucilaginous envelope and matrix reacting differently to the stain. The small Hudorina-like colonies appear very soon after the pool has formed; for example, in the third phase they occur with the Volvox “‘ Juveniles”’ collected on 2nd March (Plate XXXVI, A), and may be either spherical or slightly elongated. One measured Volvox and Associated Algae from Kimberley. 503 141 x 163 w and contained 32 cells 19 yz in diameter, of which one had already divided once. In this material practically all contained 32 cells; the colonies varied from 49x57 yw to 215x236 p. It appears to be a very early developmental stage, and during the later development of the plankton is very rare. The individual cells soon divide up into groups; and the resultant colonies, containing usually 16 or 32 groups of 8, 16, or 32 cells each, are the characteristic spherical colonies which form such a striking constituent of the phytoplankton (cf. Plate XXVI, B). The groups are placed at regular intervals within the periphery of the sphere; the cells of each group move apart and are themselves arranged regularly within the much enlarged wall of the parent cell, the resultant group being elliptical in radial section (fig. 6, B, and Plate XXXVI, B). Later, each cell of the group divides similarly, the group becomes spherical, the surrounding membrane more clearly defined, the parent envelope alters, and the daughter colonies escape, with their constituent groups already fully formed (Plate XXXVI, C to E). Throughout this phase there is extraordinary regularity in the numbers of cells and their spacing, division evidently going on more or less simultaneously throughout the colony, both before and after birth. Occasionally spheres in which the full number of groups had not formed were seen, but, as a rule, the number is very regularly 16 or 32, and in the majority of cases, if one group contained 16 cells, all the others had a like number. In colonies where division was going on the constituent groups may not all be exactly at the same stage of development (Plate XXXVI, C, D), but, as a rule, all divide very nearly simultaneously (Plate XX XVII, A). The cells (fig. 6, F) are spherical or very nearly so; each is sur- rounded by a delicate wall, usually only apparent on staining, has a massive bowl-shaped chloroplast, a single very large basal pyrenoid, and a central nucleus with a large nucleolus (or “inner body’). Much starch is formed, the whole chloroplast becoming packed with it. The vesicle surrounding each group, 7.e. the enlarged wall of the parent cell, is hardly visible without staining in young colonies, but in older colonies it may show fairly clearly (Plate XX XVII, A). Apparently reproduction by the formation of daughter colonies, as described above, may continue for some time, each daughter colony escaping and developing separately. Or the gelatinous envelopes may soften, and the daughter colonies, instead of escaping separately, each enclosed in a firm membrane, remain in contact with one another, forming an irregular botryoidal mass (fig. 6, C, 504 Annals of the South African Museum. and Plate XX XVII, B), in which no limiting membranes can be seen. Even the gelatinous matrix is invisible until stained. With methy- lene blue it turns purple, and it is clear that there is no differentiation of a surface membrane. Apparently this botryoidal form only arises from very young colonies, since the spheres and their con- stituent cells are always very small and delicate. This form was the rarest, and occurred most abundantly in material which contained fairly advanced stages of Volvox gigas. Finally, yet a third form, the Tetraspora-like form (fig. 6, E), occurs at first in small quantities (Plate XXV, D, top left-hand corner), but rapidly increases until it forms nearly the whole of the phytoplankton (Plate XXXVI, H, and Plate XXXVII, EK). Appar- ently it may arise either from the botryoidal form or from the spherical. In the former case the constituent cells divide repeatedly, the gelatinous matrix extends and its composition alters slightly, the outer layers becoming denser. Transition stages from one form to the other are occasionally seen, though they are not common. When a colony of the spherical type is passing over into the Tetraspora- type (fig. 6, D, and Plate XXXVII, D, E), the cells of the groups divide as if to form daughter colonies, but the resultant cells continue to divide more or less irregularly (Plate XX XVII, D), some at any rate forming biciliate zoospores which move through the gelatinous matrix; this has meanwhile been losing its firm outline, often becoming irregularly lobed (fig. 6, D). Eventually all trace of grouping of the cells is lost, and an irregular gelatinous mass, packed with small rounded cells (Plate XX XVII, C) similar in structure to, but smaller than, those of the spherical form of the alga, is formed. There is no trace of pseudocilia, and here, as in the spherical form, zoospores of two sizes are often seen, some ranging from 4x7 py to 5x9 p, others from 8 x 10 w to 9 or 10x11; the cilia are equal in length to, or slightly longer than, the protoplast, the chloroplast cup-shaped with nearly central pyrenoid. The zoospore is elongated, slightly oblique, with a very small apical protuberance and central pyrenoid (fig. 6, H). Thus the characteristic Tetraspora-like palmelloid stage consists of irregularly shaped gelatinous masses containing very large numbers of rounded cells without ciliary appendages of any form, except where the round cells have divided to form small elongated bi-ciliate zoospores. The colonies may be more or less rounded, but are more often elongated, irregularly lobed or branched, often 2 mm. or more in length. The cells, similar in form and structure to those of the Volvox and Associated Algae from Kimberley. 505 spherical phase, but smaller, usually show no grouping, but some- times grouping in fours or even eights is very marked. This probably depends on the rate at which division is proceeding—unless the rate is high the daughter cells separate before the next division takes place, and all trace of grouping is lost. In some colonies the number of zoospores among the round cells is very great. The four stages are so very distinctive that were it not for the intermediate stages which occur they might easily be taken as belonging to different genera. The entire absence from all stages except that of the motile zoo- spores of any trace of cilia or pseudocilia distinguishes this genus not only from the Tetrasporaceae, but also from Gloeococcus, as defined by Braun. It is difficult to understand the view taken by Wille (1903) and Lemmermann (1915, p. 31) in regarding Sphaerocystis Chodat as identical with Braun’s genus, since Braun in his definition of the genus (as quoted by Rabenhorst, 1868, ili, p. 36) explicitly describes the presence of cilia: “ Cellulae ovales, virides, in parte antica ciliis vibratoriis binis . . . instructae”’ (cf. also Chodat, 1904, p. 233), and in the following description amplifies this, stating that : “ The cells in all the succeeding generations which take place during the formation of these families except the transitory generations (in the case of repeated division) are provided with two very long per- sistent moving cilia which only disappear at the beginning of division”’ (Braun, 1851, quoted by Chodat), whereas in Sphaerocystis cilia are present only in the zoospores. In the Kimberley material the palmelloid colonies may reach a fair size, up to 3 or 4 mm. in length, but this is a very different matter from families the size of an apple or a hazel-nut which characterise the two species of Gloeococcus on which Braun founded his genus. W. and G. 8S. West (1912, p. 413) found that Sphaerocystis Schroeteri formed a large part of the phytoplankton of Loch Lomond during the warmer months only (July to October), and was not seen during the colder periods of the year. Although even then the temperature of the water (59° F. or 15° C.) was very much lower than that of the Kimberley pool (25° to 30° C.), yet the observation is suggestive. No doubt, at least, a part of the associated palmelloid material which these workers diagnosed as Tetraspora lacustris was actually the corresponding form of Sphaerocystis, since it is evident that pseudo- cilia were not seen in this material (West, 1916, p. 186). The spherical form of Sphaerocystis Power: differs from Sphaerocystis Schroeteri Chod. (1897a, p. 119, and 1897, p. 292), in the very great VOL. XVI, PART 3. 39 506 Annals of the South African Museum. regularity of development and arrangement of the constituent groups of cells; the almost perfect spheres with their symmetrically placed groups of equal numbers of equally spaced similar cells are very beautiful objects, and very different from the spheres containing groups at all stages of development characteristic of Chodat’s species ; this irregularity of development is well shown in the figures accom- panying his description (1897), pl. ix, particularly figs. 1, 5, and 7), and was so striking that Chodat made special mention of it as a characteristic feature of the species. The following examples are chosen from the many examined to give an idea of the spherical colonies :— No. of cells or Colony. groups: Constitution of groups. 49x ST p 32 cells Cells rather close together. (very young) 160 x 170 p | 16 cells (20 p) | Cells arranged peripherally. 215 x 236 pw 32 cells 1 cell dividing. 500 x 500 pw 8 groups Each containing 16 cells. S90R<1990) Lose = 5 Gua 990i 990in 1673s Each containing 16 cells, each already divided into 8. 1012 x 1012 pw 32s Each containing 8 cells (13 p). 1012 x 1012 pz O25 ie - OR (LOU): 1500 x 1500 x a2) es a ss 32 ,, (15-18 p). 1615 x 1615 pu a2 Bs . 32 ,, (20-26 p). 1000 x 1000 yu 1G ee Each containing 16 cells, some already div ided. | C. PROTOCOCCALES, SCENEDESMACEAE. Genus KIRCHNERIELLA Schmidle. Kirchneriella africana sp. nov. (Fig. 7, and Plate XXXV, G, H.) Familiae magnae, usque ad 2-4 mm. longis, juvenae fere rotundae, maturae plerumque fere teretes saepe lobatae vel ramosae; cellulis permultis, in tegumento gelatinoso hyalino sine regula dispositis, magnitudine variis; cellula trilobata, basi rotunda, apice profunde emarginato, lobis lateralibus divergentibus, interdum sub-recurvis; Volvox and Associated Algae from Kimberley. 507 chromatophora late patenti, parietali, lobata, pyrenoide singulo in parte basal. Fam., 322 x 430 pw, 1290 x 2350 pw, 1400 x 2910 pw, 862 x 4000 p, ete. Cell., 4x5 p-16 x 13 p. Cellre.g. 9x9 pu, V2x Tl pe, 16x13 pp, 13x14 p. Found in Pool I associated with Sphaerocystis Poweri, Volvox Rousseletit forma griquaensis and Volvox gigas. Sarineeity Si Lys mae VA ” Fic. 7.—Kirchneriella africana, sp. nov. A, B, Typical colonies. A, young, rounded colony ; B, older, much lobed, elongated colony of large size; C, 1-7 cells of various shapes and sizes from a single colony ; D, group of four cells formed by division of a single large cell, the empty wall of which lies adjacent to the group; cells show the characteristic shape even at this stage; E, two very young cells, one nearly straight, the other three-lobed ; F, large cell with nucleus already divided and pyrenoid dividing. A x70; B x30; C-F x1500. A very large form; the gelatinous colonies are more or less rounded when young (fig. 7, A), later becoming elongated and irregularly lobed or branched (fig. 7, B), often 1 to 2 mm. in length, while the largest colony as yet observed measured 4 mm. in length by nearly 1 mm. in the widest part. The cells are very numerous, embedded without definite orientation within the colourless gelatinous matrix, very variable in size and form in one and the same colony (fig. 7,C; Plate XX XV, H), usually tending 508 Annals of the South African Museum. to be three-lobed; basal lobe massive, rounded, apical lobes usually widely divergent, separated by a wide and deep cleft somewhat V-shaped, but with rounded base. Typically this cleft is about one- third the total length of the cell in depth, more rarely it may be quite shallow (fig. 7, C 3). The apices of the lobes are more or less narrowly rounded, sometimes slightly recurved. The parietal chloroplast is extensive, slightly lobed, lhning most of the cell except a small part beneath the cleft; in this the nucleus lies (7.e. approximately in the centre of the cell). The single large pyrenoid lies in the massive basal part of the chloroplast. The cells are rich in starch, and contain, in addition, refrangent globules, probably volutin, generally most conspicuous in the apical lobes. The wall is delicate, closely adpressed to the protoplast, only becoming apparent on staining, e.g., with iodine or methylene blue. With the latter the gelatinous matrix stains faintly bluish purple, the protoplast deep blue. The cells divide to form 4- or 8-celled groups (fig. 7, D) ; the empty wall of the parent cell is often visible near such groups of young cells. The three-lobed structure is apparent at a very early stage (fig. 7, D), but sometimes the young cells appear almost straight (fig. 7, E). As the cells thus formed grow they diverge from one another, and all trace of grouping is soon lost. This species appears to be nearest K. Malmeana (Bohlin) Wille, (cf. Brunnthaler, 1915, p. 181, fig. 266), but the cells are typically much more deeply incised and more distinctly three-lobed, with rounded apices, than in that species, although some cells occur in which the anterior end is only slightly concave (fig. 7, C, 3). The mature cells here are larger than in most species and the colonies very much larger than anything as yet recorded for the genus; the number of com- ponent cells, too, is exceptionally great. Large cells in which the nucleus or pyrenoid had already divided, presumably preparatory to cell division, were occasionally seen (Caney, 1, I) On and inside many of the colonies biciliate motile cells were often found, but, as these are identical with the zoospores of Sphaerocystis Poweri, it is almost certain that they belong to that species, and have penetrated the matrix of the Kirchneriella colonies. In the same way they may enter older Volvox colonies of either of the associated species (cf. Plate XXX, B). Volvox and Associated Algae from Kimberley. 509 Part IJJ.—FIELD OBSERVATIONS AND RECORDS. Pool I.—The pool where Volvox was first found lies in an excava- tion roughly 40 yards in diameter and about 3} feet deep in the deepest part. It was made some eight years ago by the removal of red clay for building purposes, and the bottom and sides are com- posed of this red earth. Any water which lodges in it is very muddy and of a reddish colour. It lies about a mile to the S.E. of Kimberley, just off the Samaria Road and close to the Bultfontein Floors. Volvoz was first collected here in November and again early in December 1931. The pool dried up about the middle of December, but was filled again by a thunderstorm during the afternoon of Saturday, 23rd January, and remained in existence for about four weeks, the last collection being made on 18th February. By 24th February it was completely dry. The third phase was started by a thunderstorm during the night of Saturday, 27th February, and this time the pool had a rather longer existence, another storm on 19th March partially refilling it before it had quite dried, so that it continued some days longer; but it was dry again by 2nd April, and with that date the observations on this pool finish. Volvox disappeared from it about the 19th of March, and did not reappear. Pool II.—No other Volvox pools were found until after the last rains, when Volvox was collected in two more quarry pools. The second one lies about 4 miles 8.W. of the town. The sides and bottom are of broken stone and the water clear. The phytoplankton consisted entirely of asexual material of Volvox Rousseletii, with a very small proportion of Volvox africanus mixed with it. Streptocephalids, abundant in all the pools, were particularly so here. Pool II1I.—The third pool is of yet a third type, though it also is in a quarry. This quarry, however, is in limestone tufa which forms the sides and bottom of the excavation; the latter is covered with mud and the water is very muddy. The pool lies about 1} miles N.E. of Kimberley ; it was fast drying up when Volvor was first found in it and was only visited twice. The Volvox was very scarce, and consisted entirely of asexual colonies, mostly immature but large, markedly egg-shaped, and containing numerous embryo colonies. On analysis, the water was found to contain -92 per cent. lime and “Ww 29 X eg ‘T]TeUS awos = ‘[VUuoIsvo0Q *SaTLO[OO Jay qSNVp YY auLos ‘asivy = ‘[VUOISvIDO *[BUOIStI0Q, ‘supDbaja Dursopny oIVY *ATOATQOV Sur -plAIp sjjao ‘asaey] ‘quepunqe = Ape, *ULUL @. X “UU 9+ <— ‘anIe] “uOUIULO) APALE “puvor4{D DIA 4UY ILI ‘EGIL WadY ANZ OL [EG UAANAAON ‘J 100d AO NOLUNVIAOLAHG ANV AUOLSIY IVNOSVAS “May APOAIQVAVAULOD S]]A9 [RUOIS -e000 (p) $— (9) ‘que -punqe Ayarey (9) + — (n) *S91U0]09 LOQ]O A plo opisut Suruaeas solodsooz ‘[[BUS S[[e0 ‘aIRL (p) ‘ uourm10d Aparey (0) Syeuorsvo00 (q) paleo -7e GE ‘peuotsvoo0o AdaA (VD) *[BUOISBDDQ () ‘aITI (p) Syeuontsuvsy = Moy e Sasavy ‘[euotsBvo0g (q) “ulloy Blodsvsqay, =(p) ‘Teprosaqog =(9) *jwormay dg = (4) ‘uLIO} VULIOpNY =(”) “wanog sushoo1anydy “suBaoRysnio JUBpUNGe OD AoA YQIA poxtur ‘areyy ‘sazodsoo odir Auvur $ [pews ‘4uepunqe s,P *“SATMOTOO JapTo Ut uvy9 sunod Ul snotewnu aioul ‘4, <— sodrquia ‘padeys-s50 AtoA Sunos ‘quvurmopaid [T@nxasy *OINYVUL Moy @ Ssunod ATJSOUL quepunqge Ape *"SoTVoudd YIM poysejut aemog ‘sno -dUINU a1OUL [VNxXeSse ING ‘Quvurmopeid [[Qs [vnxes “Wl Zg VIPIUOS ‘A OGZLXGPIL <— ‘sunof = ‘[vuoIsvo0Q, ‘1 SOIT XZ1OL < 3 ‘1 TLGLX9GOL < P “W OEh X 99g <— SoAIqUIA ‘r 6LOT X LOGL <— [enxesy ‘OL <— 9nq ‘2 > Apsour sodrquaipy *Auojoo aad viprmosoipuy puv ssa Jo laquinu as1v] AdaA ‘yenxese ueyg gue -punqe o10ul TONUL [BNxeG “1! SPL X B63 Ayensn ‘¢— ‘soLIquia Moy TIA [enxosy “‘qurulmopedd penxeg *savaads yuDLULO(T *[euoIsvo00 AIBA SoTTMOAN(* ‘area AIOA 4 ‘orngeut AweUr ‘Guepunde P *Auojoo aad JT <— g IaAO Apisout ‘sokaquia asavy YQIM ‘[Tenxese quepunqy *e-T sodaquia ‘WH T@XELp <— soptueane “uolIyeIoues pug Jo sarm0]09 pue ,, SeTIWeANS: ,, *x02]04 JO USIS ON ‘hap ood ‘alquad aBau sjods poqvjost ur daap soqout % [ood “ATuavd polpyor ‘Aap Apwou joog ‘daap satpout p ‘deop soyoul g [ood “qand ysadaop ut 4OOJ TT JHOQB [OO “pat[yort [00d “Aap Ajoje,dur09 “qqata suyanp ured aul0g “qystu sulnp ured aul0g “VYs10 Sunp UIvI Aavay “ adv puz “4908 ee e ISLS eS SUSI G “ uqt ss Seeger “ure OF OT GEGL “BIN UPL “Ue O86 GGL AV Ub “Ure O86 Ss “IBIN PUG “ “ 166 SE6L “G9a WILG ITT aspyds SEG “G9T UPS 512 Annals of the South African Museum. ‘004 per cent. sodium chloride, 7.e. about twice and three times respectively the normal amount in Kimberley tap-water. Most of the collecting was done from Pool I—when the pool filled and the water was comparatively deep it was from the surface layers only; for instance, on 3lst January: “‘ The water was about 6 inches to 1 foot deep where the collecting was done, and the net was never immersed more than 3 inches. The Volvox [i.e. V. Rousseletii] was very plentiful, one sweep of the net being sufficient to collect quite a quantity.” It is possible that the non-motile algae are as a rule not actually at the surface but floating in a somewhat lower stratum, and that therefore the collecting at the beginning of each phase does not present the phytoplankton in its entirety ; at the same time it seems improbable that this is the whole explanation of the increasing proportion of these algae as the water grows lower. There is certainly a continual increase in their number and bulk as the pool ages; but possibly, if deeper collectings were made in the early stages of the pool’s existence, the number of individuals of these species would be found to be more numerous, and a more complete picture of their development could be given. As it is they are a sufficiently striking constituent of the vegetation of the pool. On the last occasion that a collection was made from this pool (2nd April) the gelatinous algae are described as forming “a solid mass at the N.W. side of the pool, making the water there a vivid green. Eventually, when the pool dried up, they formed a thick green layer in several small depressions in the mud. The temperature of the water varied from 25° to 30° C. during the whole period of collecting. The collecting was usually done between the hours of 9.30 a.m. and 12 noon.” The amount of Volvox often varied in different parts of the pool according to the prevailing wind; when a strong wind was blowing the Volvox tended to collect at the leeward end of the pool. The table on pp. 510, 511 shows the succeeding changes in the pool and its vegetation during the time it was under observation, the latter as shown by the material collected and preserved at each visit to the pool. It must be emphasised that this does not pretend to be the full history of the phytoplankton, since, as the collecting was done entirely with a view to the study of Volvox, no attempt was made to obtain its finer constituents apart from those incidentally collected with Volvoz. Volvox and Associated Algae from Kimberley. 513 SuMMARY. 1. Volvox has been found in three temporary rain-water pools in quarries near Kimberley ; these pools may fill and dry up several times during a single season, Volvox disappearing before the pool is dry and reappearing soon after it refills. 2. Of these three pools the first one found—Pool I—was visited repeatedly from November 1931 to the beginning of April 1932, during which time it came into existence on three separate occasions, the duration of the pool varying from three to five weeks, alternating with periods of dryness varying from a few days to several weeks. On each visit collections were made and preserved in formalin or iodine. The other two pools were not discovered until late in the season, after the last heavy rain, and the collections from these are consequently few and scanty. 3. Three species of Volvox were found: one, a local variant of the widespread V. Rousseletii, was common to all three pools; the second, a startlingly large one, sometimes as much as 3 mm. in diameter, is diagnosed as a new species to which the name Volvox gigas is given. It was found in Pool I only. The third is a very large form of Volvox africanus G. 8. West, usually containing four pairs of embryos; it occurred in small quantities, entirely asexual, in Pool IT only. 4. In Pool I three other algae are associated with the two species of Volvor, namely: (a) Eudorina elegans, mainly large colonies, occurring in small quantities in the tubes, probably very much more numerous in reality ; (b) An interesting, very polymorphic gelatinous non-motile alga, described as a new species, Sphaerocystis Poweri, one form of which eventually becomes the dominant alga in the pool ; and (c) A large, very many celled species of Kirchneriella, described as K. africana sp. nov. 5. The phytoplankton of Pool I shows an interesting succession, of which the following are the main phases :— (a) Volvox Rousseletii dominant, Sphaerocystis present in small quantities, and, towards the end, young colonies of V. gigas. (b) Volvow Rousseletii becoming less numerous, V. gigas increasing, sometimes becoming dominant, Sphaerocystis and Kurch- neriella increasing ; 514 Annals of the South African Museum. (c) Volvox scarce or absent, Sphaerocystis dominant, Kirchneriella fairly numerous; (d) Vegetation consisting almost entirely of the palmelloid form of Sphaerocystis Powerit ; pool almost dry. 6. A summary of the seasonal history from November 1931 to April 1932 of Pool I and its vegetation is given in tabular form. Discussion AND CONCLUSION. Occurrence in a region of summer rainfall appears to be peculiarly favourable to the development of Volvox—all the three species found in the Kimberley district are outstanding either for their size, e.g. V. gigas and V. africanus, or for the extreme richness of development, e.g. the reproductive cells of V. Rousseletii, particularly the sexual cells. With this may be compared V. amboensis, Rich and Pocock (loc. cit., p. 462), from a region which has very similar climatic conditions to Kimberley. That a comparatively high temperature is favourable for development in Volvox was also suggested by observations in 1931 of V. capensis on the Cape Flats. The same is true of the associated algae, all of which are nese of their kind—even the Hudorina colonies are unusually large, the Kirchneriella is larger than anything previously recorded for the genus, while Sphaerocystis Poweri shows great luxuriance of growth. Of recent years much work has been done on the algae of South Africa, but mainly from the systematic standpoint, consisting in the identification and description of species and records of occurrence as shown by more or less isolated collections. As yet the ecological aspect has hardly been touched; quite recently an ecological study of the plankton of various pans and other inland waters has been published (Hutchinson, Pickford, and Schuurman, 1932), but even this very interesting piece of work deals with seasonal observations in two only of the localities under review. A more detailed study of the seasonal changes in planktonic life, extending over a period of two years, was made in the case of Florida Lake, Johannesburg, by Miss Schuurman (1932). Both these papers deal only with the plankton. As yet no study of the seasonal succession of the algal vegetation as a whole has been published for any piece of water in South Africa. Further, in most cases the pans or vleis concerned in these two papers are either permanent (e.g. Florida Lake), semi- permanent, or extensive artificial bodies of water very different from the small temporary pools near Kimberley. Such rain-water pools, either as in this case in artificially made hollows in the surface of the Volvor and Associated Algae from Kimberley. 515 ground or in natural depressions, are very characteristic of large areas in Africa. If the rains are good they may fill up yearly or several times each year; on the other hand they may remain dry for several years in succession. When they do fill, their duration is often very brief, often only a week or two, yet in that time an extraordinarily rich algal flora often develops. Obviously such pools would often form most favourable subjects for intensive observation. The Kim- berley pool is apparently of comparatively recent origin, and it seems probable that far richer results could be obtained by a systematic study of some of the many small temporary pools which form in natural hollows in high veld regions one wet season after another, and have probably done so from time immemorial. The present account is as far as is known the first attempt as yet made to describe the seasonal changes in a part at least of the flora of such a transitory pool, and since it is compiled from a series of observa- tions and collections made with one object only, namely the study of the behaviour of the single genus Volvoz, it will be obvious that its value as a general seasonal history of the pool is small. It may, however, serve to illustrate the interest and value of a detailed study of the algal vegetation of individual pools of this kind, and it seems time that such seasonal studies should be initiated to supplement our growing knowledge of the systematic side of freshwater algology in South Africa, and to correlate it with the occurrence of the various con- stituents of the rich and varied algal flora of this country under varying local and seasonal conditions. RONDEBOSCH, CAPE PROVINCE, July 1932. ; BIBLIOGRAPHY, Most of the papers named here are referred to in the text ; a few are not speciti- cally alluded to but were consulted during the work. Those marked * have not been seen in the originals. *Bravn, A., 1851.—Verjung. i. d. Natur., p. 170, fig. 43, b-k. Brawn, A., 1855.—Algarum unicell. gen. nova et minus cog., ete. Lipsiae. *Bravn, A., 1886.—‘* Uber Chytridium, etc.,” Abh. der konigl. Akad. d. Wissensch. zu Berlin. BRUNNTHALER, J., 1915, in Pascher, Siissw. Flora, vol. v, p. 180. Cuopart, R., 1897.—(a) ‘“‘ Algues pélagique nouvelle,” Bull. de ? Herb. Boiss., vol. v, p. 119. Cxopat, R., 1897.—(b) ‘‘ Etudes de Biologie lacustre,” ibid., p. 289. Cuopat, R., 1904.—‘‘ Quelques points de nomen. algolog: 1. Sphaerocystis Chodat, ou Gloeococcus A. Braun,” ibid., vol. iv (2me. ser.), p. 233. 516 Annals of the South African Museum. Hvrcuinson, PickrorD, and ScuuuRMAN, 1932.—‘‘ A Contribution to the Hydro- biology of Pans and other Inland Waters of South Africa,” Archiv fur Hydrobiologie, Bd. xxiv, 8. 1-154. IYENGAR, 1933.—‘‘Contribution to our Knowledge of the Colonial Volvocales of S. India.”’ (In the Press.) JANET, C., 1923.—Le Volvox: III. Limoges. *K USCHAKEWITSCH, S., 1922.—‘‘ Zur Kenntn. d. Entwicklungsgesch. von Volvox,” Bull. de ’ Acad. de Sci. de ?Oukraine, vol. i, p. 31. KuscHakeEwItscu, S., 1931.—‘‘ Zur Kenntn. d. Entwicklungsgesch. von Volvox,” Archiv f. Protistenk, Bd. 73, p. 323. LEMMERMANN, E., 1915, in Pascher, Siissw. Flora, Heft v. Meyer, A., 1895.—‘ Uber d. Bau v. V. aureus Ehrenb. and V. globator Ehren.,” Bot. Centralbl., vol. Lxiii. *Mryer, A., 1896.—‘* Die Plasmaverb. u. Membranen v. Volvox, etc.,” Bot. Zeitung, vol. liv, pp. 187-217. Ottrmanns, F., 1923.—Morph. u. Biol. d. Algen., Aufl. ii, vols. i and iii. Pascuer, A., 1927.—Siissw. Fl. Deutschlands, Osterreichs u. d. Schweiz., Heft iv, Volvocales, etc., 1. *Powers, J. H., 1907.—‘‘ New Forms of Volvox,” Trans. Am. Micro. Soc., vol. XXVi, p. 31. Powers, J. H., 1908.—‘‘ Further Studies in Volvox,” ibid., vol. xxviil, p. 141. Printz, H., 1927, in Engl. Prantl, Pflanz. Fam., Aufl. ii, Heft 3. RaBenuorst, L., 1868.—Flora Europaea Algarum, vol. iii (Gloeococcus A. Braun, pp. 6, 36). Ricu, F., 1932.—‘‘ Phytoplankton from §8.A. Pans and Vleis,” Trans. Roy. Soc. SeeAtres vOle xexe 949), Ricu, F., and Pocock, M. A., 1932.—‘‘ Observations on the Genus Volvox in Africa,” Ann. 8. Afr. Mus., vol. xvi, p. 427. ScnuuRMAN, J. F. M., 1932.—‘‘ Seasonal Study of the Microflora and Microfauna of Florida Lake,” Trans. Roy. Soc. 8. Afr., vol. xx, p. 333. Saw, W., 1894.—‘‘ Pleodorina: A New Genus of the Volvocinae,” Bot. Gaz., VO leeexdx. Suaw, W., 1916.—‘t Besseyosphaera: A New Genus of the Volvocacae,” ibid., vol. Ixi, p. 253. SHaw, W., 1922.—* Merrillosphaera,”’ Phil. Journ. Sc. Manila, vol. xxi, p. 87. SHaw, W., 1923.—‘‘ M. africana at Manila,” ibid., vol. xxii, p. 185. West, G. S., 1910.—‘‘ Some New African Species of Volvox,” Journ. Quek. Mier. Club, Ser. 2, vol. xi, p. 99. West, G. S., 1918.—‘‘ A further Contribution to our Knowledge of the Two African Species of Volvox,” ibid., vol. xiii, p. 425. West, G. 8., 1916.—Algae, vol. i (Cambridge). West, W., and West, G. S., 1912.—‘*‘ On the Periodicity of the Phytoplankton of some British Lakes (Loch Lomond),”’ Journ. Linn. Soc. Bot., vol. xl. (May). *WittE, N., 1903.—Algolog. Not. in Nyt Mag. f. Naturvidensk Christiania, pp. 90-176. WitLE, N., 1909.—Chloroph. u. Conj. in Engl. Prantl, Nat. Pfl. Nachtr. zu Teil i, Abt. 2. ZIMMERMANN, W., 1925.—Die Ungeschl. Entw. v. Volvox. Die Naturwissen- schaften, vol. xix, p. 397. Volvox and Associated Algae from Kimberley. 517 DESCRIPTION OF PLATES. The number in brackets at the end of each description denotes the reference number of the original negative. PLATE XXV. Volvox gigas sp. nov. FIG. A. Group, four large asexual colonies with young embryos and gonidia, and a very young colony; in the centre an asexual and a mature male colony of V. Rousseletii forma griquaensis (602). Dimensions: V. gigas (1) 2000 mu, (2) 2148 x 2590 yw, (3) 2222 x 2660 nu, (4) 2814 x31ll yu. V. Rousseletii: asexual 1111 4, male 740 x 880 yu. B. Male and female colonies, the latter with large scattered oospheres ; the male colony in B2 typically egg-shaped. In both males all the cells in the posterior part of the colony are developing into sperm bundles; spaces among the cells show where sperm bundles have already escaped (627). C. Small mature colony with few-celled embryos, probably males (compare with Plate X XVI, E), and a few oospores (694). D. Group, showing very young and nearly mature asexual colonies (one with between 50 and 60 embryos), mixed colonies with eggs and embryos, and pure male and female colonies ; at the top left-hand corner are two spherical and three or four palmelloid colonies of Sphaerocystis Poweri, while near the bottom right-hand corner is a small spherical colony beginning to show transition to the palmelloid state (639). AL x 13:5) Bx 25.) C385 EDP > @. VOLVOX GIGAS. Photos: M. A. P. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate X XIX. VOLVOX GIGAS. Photos: M. A. P. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate XXX. VOLVOX GIGAS anp AFRICANUS. Photos: M. A. P. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate XX XI. VOLVOX AFRICANUS. Photos: M. A. P. Neill & Co., Ltd. Ann. 8. Afr. Mus., Vol. XVI. Plate XXXII. VOLVOX AFRICANUS. Photos: M. A. P. Neill & Co., Lid Ann. S. Afr. Mus., Vol. XVI. Plate XX XIII. VOLVOX AFRICANUS. Photos: M. A. PB. Neill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate XXXIV. VOLVOX ROUSSELETII Forma GRIQUAENSIS. Photos: M. A. P. Veill & Co., Ltd. Ann. S. Afr. Mus., Vol. XVI. Plate XXXV. VOLVOX ROUSSELETII rorma GRIQUAENSIS ann KIRCHNERIELLA AFRICANA. Photos: M. A. P. Neill & Co., Ltd. Ann. 8. Afr. Mus., Vol. XVI. late XXXVI. ge € 40m Oe eg “ oe “ase? & es Photos: M. A. P. SPHAEROCYSTIS POWERI, Neili & Co., Lid. Ann. S. Afr.-Mus., Vol. XVI. Plate XX XVII. “4, ‘ | wee: ts bree os (Betis e os ; me B+: SPHAEROCYSTIS POWERI. Photos: M. A. P. Neill & Co., Ltd. F igeaas . —— er a o , eo t = _ : - 2 i — a 1 “=e Fr - - ‘y 5. Volvox in South Africa—By M. A. Pocock, Ph.D., F.L.S. (With Plates XX XVIII-XLIX and 10 Text-figures.) CONTENTS. PAGE Introductory ; : 0 : : : A : . 525 Aim and Results of lnivestication : : é C C : . 526 Methods . : : : : ‘ F : : . 527 Occurrence of Volvox in Weiton : : ; 3 - : : . 529 Habitat of Volvox in Africa. é : 5 : : : : . 530 V. Rousseletii and V. capensis. I. The Adult Form . : : : ool Distinguishing Features. : ; : 5 : : : 5 Geil Parasites of Volvox . : : : : : : . : » 533 Macroscopic Observations . ; : : 5 : : 3 . O34 Movement ; Phototactism . : ‘ ‘ : : i . 535 Sequence of Asexual and Sexual Pisses : : : 9 : . 538 Observations on the Cape Flats . : : : : é : . 539 Constitution and Size of the Coenobium : : : : : . 543 The Somatic Cell : : ‘ : : : ; : é . 546 Size of Cell : : : : : : ; ; , . 547 . Cell Membranes . ‘ : : ‘ : : : . 548 3. Protoplast, Form and Stricture : : : : : . - 590 (i) Nucleus . é E : ; : ; ; : soo (ii) Chloroplast. : : : c : c : . 652 (iii) Pyrenoid. : F P ; ‘ : j f . 552 (iv) Cilia : , : . : : : : ; - 503 (v) Eyespot . : : : : - . . : . 554 (vi) Protoplasmic Connections . : : ‘ : : - 556 (vii) Contractile Vacuoles. : c : : : . 558 Entity and Interdependence of the Somatic Cells ; 2 : : . 559 Change in Appearance as Coenobium Ages . é : A . 559 Cell-structure compared with that of V. globator_ . : : , - 560 Reproduction . : 2 : : - 560 Number and Anraneciiont of Rapiodicee Cells : ? ; - 562 A. Asexual. : P : : : : : : : 2) O02 B. Sexual, Male . : ‘ A ; , : ; : 7) O04 C. 33 Female : : ; : ; ; : 5 - 565 Development of Reproductive Cells . é : : c : - 566 A. Gonidium é : ; 3 3 : : 5 : - 566 Enlargement : : : : . : : : - 566 Division . c 5 : : . : : 5 - 567 Period of Re-orientation : : : B , . = eiz/il (a) Preparation : ; : : : : : - O12 (6) Inversion . ‘ ; < . : ° : . 574 524 Annals of the South African Museum. Reproduction—continued. Development of Reproductive Cells—continued. PAGE A. Gonidium—continwed—Results of Inversion : : : 5 OT Time of Inversion 6 : : : : : : ors Discovery of Inversion. , : : . : c OS Daughter Colony. : : : : : : > 582 (a) Development after inversion c . ; 5 : . 582 (6) Birth . ; : : : ‘ : ; , - 583 B. Male Initial Cell. ; : 3 ‘ : : : . 585 1. V. Rousseletit . : : ‘ ; : : ; . 587 Enlargement ; . : : : ‘ : . 587 Shifting of Polar Axis . : ; ‘ ‘ : : ool Division : ; : : : é : . 588 Preparation for Inversion : : : : 2 . 590 Inversion é c 592 Sperm Globoid, Maturation : ; é : é . 593 Escape 2 : : : : : : . 594 Life Free in Water : : : : : : . 596 Escape of Sperms. : ° é c : : . 596 Structure of Sperms . ; : : ‘ : . 598 2. V.capensis . 2 : C : : ‘ P . 599 C. Female Initial Cell . ; é : : : : ; > 601 Enlargement é . : : : : : : = oOl Mature Oosphere : : 5 : : : - 603 Homologies of the Reproductive Cells 5 : : : : . 603 Fertilisation . : ; : : : : . 606 Oospore, Development ntver Bertiteation : : : : ; . 609 Size and Appearance in the Two Species : : ; , 5 Oe Liberation . : A , 3 3 ; : : : 5 Il Juvenile Form, Discovery : ; ; . . : : - 612 Oospore Cultures ; Collection of Oospores : ; , , . 614 Culture Methods and Results : ; ‘ : ‘ : . 614 Germination of Oospore ‘ : : ‘ ‘ , : 2 616 Zoospore, Structure. ; : : ; : : : 5 (ale Division. : ; 5 . : 5 : O20 Inversion of the Germ Oslony é ; : ; : : 5 ey Appearance after Inversion . : : : : é . . 624 II. The Juvenile Form, Structure . : : : : : ; 4 625 Development of Second Generation ; : : ; : - 626 Previous Accounts of Germination . : . : : ; : » 627 Discussion. : : : : 0 : ‘ : : . 629 Ie Geonimation : : : : : . 629 2. Form of the Germ Colony aveniion : 5 6 é : . 630 3. The Process of Inversion . ‘ : : : , - 631 4. Variation in Number and Form of Cells : : : : > 631 5. Systematic Aspect . ; 3 : . ; ; : . - 632 Summary . 0 c ‘ : ‘ ‘ i ; Ose ‘Anpendis = Ocspore Galtares : : 5 é ; : é é . 636 Note—Volvox from Grahamstown . : : : : : 3 . 638 Bibliography c : ; 6 : 6 5 : : ; . 639 Description of Plates. 2 : : : é b : c . 642 Volvox in South Africa. 525 INTRODUCTORY. From the time when Volvor was first observed and described by Leeuwenhoek (1719, p. 149) it has been a favourite object for micro- scopic study, and there has sprung up an extensive literature on the genus Volvoz, fruit of the labours of both botanists and zoologists as well as microscopists. Nevertheless, even in such well-known forms as Volvox globator (L.) Ehren. and Volvox aureus Ehren. (=Janetosphaera aurea Shaw) certain points in the life-history remain obscure. Pascher (1927, p. 450, note) points out that his account of the genus can only be regarded as provisional owing to the lack of comparative studies from wide areas. Again, he emphasises the need for further investi- gation by means of careful fixation, sectioning, etc., but above all for numerous studies of living material. That there has been a tendency to neglect the latter is well exem- plified by the fact that the inversion of daughter colonies subsequent to the last cell division in the development of the gonidium was accurately described for the first time as recently as 1919 by Kuschakewitsch (1922), although it had actually been observed by Powers years earlier (1908, p. 158), and this despite the fact that if a small quantity of a well-developed, actively reproducing gathering of Volvox is placed in a watch-glass of water and observed for half an hour or so under a low power of the microscope, some of the daughter colonies are almost certain to show inversion, in living material a most striking and very easily observed phenomenon. In his critique of Kuschakewitsch’s paper, Zimmermann (1923, p. 85) remarks: ‘‘ The discovery (7.e. of inversion) shows perhaps even more markedly than the discovery of the gametophyte generation in Laminariaceae how fundamentally wrong it is to assume, even in such fully investigated organisms as Volvoz, that the earlier in- vestigators fully elucidated all points, even if only in the external course of development.” Volvox has been known from Rhodesia since 1905 (Rousselet, 1906, p. 393 ; and West, 1910, p. 99), from the Cape Flats from about 1915, and more recently from various other localities in South Africa. Little has, however, been done on the South African forms beyond describing them generally (as a rule from preserved material only) and recording their occurrence. No systematic observations of the behaviour and development of Volvor in this country have been 526 Annals of the South African Museum. published, although the Cape Flats form has for some years past been collected fairly regularly during the winter months. In February 1930 Mr. E. J. Steer succeeded in obtaining a rich growth of Volvox from soil brought by Miss E. L. Stephens from Rietfontein, near Johannesburg; the culture was first made in a small glass tank, later established in a large tub, whence it spread to other tubs and fish-ponds in Mr. Steer’s garden at Sea Point. It has now continued to flourish in one or other of its chosen spots for over two years, for a great part of which period it occurred in very large quantities. Thus a continuous, unlimited supply of vigorous living material became available, and it seemed a good opportunity to carry out detailed observations on Volvox, in the hope, first, of clearing up various points as regards the identity of the South African species, and, secondly, of possibly adding to our knowledge of the genus Volvox as a whole. At the same time a careful study of the form of Volvox found in various vleis on the Cape Flats was started, and its structure and behaviour compared with those of the Riet- fontein material. This was made possible in the beginning by the fact that, although the year 1930 was a very bad one for Volvox on the Flats in general, yet in one locality (Scanlan’s Vlei, Wetton Road) Volvox was even more abundant than usual, while the following year (1931) there was an exceptionally good and prolonged season (May to November) for Volvoz in most of the localities near Cape Town in which it is known to occur. Aim of the Investigation. The aim of the investigation, then, is twofold : (1) To carry on a detailed observation of two of the forms of Volvox found in South Africa—Volvox Rousseletii West, as it appeared in cultures of soil from Rietfontein, and Volvox capensis Rich and Pocock (1932, p. 462), from the Cape Flats—more or less on the lines followed by Klein (1889 a and 6, 1890), Janet (1912, 1922, 1923), and other workers in the case of V. globator and V. aureus in Europe, supplementing microscopic observation of the living organism with observations carried on in the field, and more detailed observation of prepared sections of carefully fixed material ; and (2) To supplement, if possible, the facts already known as to the life-history and structural details of Volvox in general. Volvor in South Africa. 527 Results of the Investigation. Mr. Steer’s material proved exceptionally favourable in many respects, and, as a result, not only has it been possible to obtain full details as regards the structure and habits of the two South African species in question, but also to extend our knowledge of the genus Volvox as a whole, filling up several gaps in the life-history. The most important of these are the following :— 1. Changes in the daughter colony and its component cells on completion of cell division preparatory to and during inversion. 2. Inversion of the “ male colony ” formed by division of the male initial cell to form the sperm globoid. 3. Formation of a free, motile zoospore as the result of the germina- tion of the oospore. 4. Development of this zoospore to form a new colony, which inverts on completion of cell division. 5. Form and structure of the resultant young colony, the “juvenile ” form. Methods. As already mentioned, the main investigation has consisted of detailed observation of living material. The author was, un- fortunately, not successful in establishing vigorous permanent Volvox cultures (cf. Klein’s observations of the capriciousness of Volvox in captivity : . cultures, so successful the previous year . failed . . .,”’ 1890, p. 2), hence visited Mr. Steer’s garden at frequent intervals for Volvor Rousseletii, noted the amount and state of Volvoz in the various tubs and pools, made collections of the material and examined it in greater detail as soon after as possible. In cool weather it could be kept healthy indoors for some days or even weeks, but in summer it soon became abnormal and degenerated, cf. again Klein (1889 a, pp. 45-46). The same was done in the case of vleis on the Cape Flats ; in the collections from the latter, however, degeneration was usually more rapid, probably largely owing to the greater wealth of animal life present. In both cases, colonies of various kinds were placed in a watch-glass of water to avoid the distortion due to the weight of the colony resting on the slide (in the case of large colonies this is con- siderable even in a comparatively big drop of water), and measured under the microscope. Many photographs were taken of the colonies as a whole under low powers of magnification ; this necessitated the killing of the colonies to prevent movement, the method used being ee 528 Annals of the South African Museum. usually the addition of a drop of 1 per cent. osmic acid to the water containing the Volvox. By the use of a Davis shutter for attachment to the microscope objective, kindly lent by Mr. Steer, additional focal depth was obtained ; this was most useful in photographing the colonies in the round. Cell detail was photographed under higher powers—in many cases it was found possible to observe and photo- graph successfully, while actually living, single cells or groups of cells, using a 1/12” oil-immersion lens. In this way a surprising amount of detail as nearly free from distortion as possible could be obtained. Further, by the use of panchromatic plates, in some cases, e.g. in the study of oospores, details of structure almost or completely imper- ceptible by the eye are recorded by the camera. In the case of inverting daughter colonies, the pressure of the cover-slip supported on pieces of cover-slip glass keeps the parent coenobium in position without interfering with the process of inver- sion, so that it could be followed photographically by a series of fairly high-power photographs taken at intervals of a few minutes. The male bundles are, of course, very much smaller, and in the study of their development the oil-immersion objective (Bausch and Lomb, 1:9 mm.) was used with a Leitz 3 optical. In the process of photographing with high powers there was often considerable diffi- culty owing partly to the fact that the slightest change in pressure on the cover-slip may cause the object to slide slowly out of the field, but more particularly because here, as in the developing daughter colony, the whole structure and the individual cells composing it are slowly but continuously moving and altering their shape. Thus the range of possible exposure is very limited. In the case of inverting sperm bundles when the fastest available panchromatic plates were used with a powerful projector illuminant, the minimum exposure for good results was 8-10 seconds, and even this usually gave slight blurring in some parts owing to the movement of the cells (cf. Plates XLII and XLIV). This, of course, applies much more to the cilia, of which in living cells only the basal, non-moving parts appear clearly in the photograph, the upper parts showing only as a blurred outer zone parallel to the surface of the colony (see Plate XIX, A).* In addition to the photographs, careful drawings were made throughout. Records were kept as to the state of the Volvor in Mr. Steer’s * For reasons of economy, photographs and diagrams already reproduced in the former paper on Volvox in Africa (Rich and Pocock, 1933) are not repeated here, although referred to in the text. Volvox in South Africa. 529 various pools and in the vleis on the Cape Flats, the latter being correlated as far as possible with weather conditions. Further, material has been fixed in various ways, embedded and microtomed to supplement the details observed in living material, particularly with regard to inversion. Since in some respects it has been possible to follow out the life- history in the case of the Rietfontein Volvox (Volvox Rousseletir) more fully than that of the Cape Flats Volvox (V. capensis) that form will be described fully, the points in which the latter differs from it being considered as they arise. Where nothing to the con- trary is stated, the account given applies equally to both species. Although much of the detail repeats the findings of earlier workers (e.g. Cohn, Klein, Overton, Janet) in the case of V. globator in Europe, it is given fully here, since V. Rousseletii and V. capensis, though undoubtedly very closely allied to V. globator, yet differ from it in many details, and it is important that they should be known in every aspect. Moreover, there is at present no recent comprehensive account of Volvor in English, Shaw’s papers (1916, 1919, 1922, 1923) dealing more particularly with the classification of the group. Occurrence of Volvox in South Africa. Since the occurrence of Volvox in South Africa has already been dealt with (Rich and Pocock, p. 427), it will be sufficient here to give a very brief summary of the records. Shaw (1922) has subdivided the genus Volvox into several new genera, retaining the name Volvox for that section of which V/V. globator is the type. As yet his classification has not been generally accepted in its entirety, but both Pascher (1927, p. 462) and Printz (1927, p. 58) agree that some reclassification is necessary, and the latter, while not accepting Shaw’s new genera as such, uses his nomenclature to denominate subsections of the genus. Of these subgenera or subsections three have so far been recorded from Africa : 1. Merillosphaera, represented by Volvox africanus West (Merril- losphaera africana according to Shaw), apparently restricted in its distribution, since it has been recorded only from Lake Albert Nyanza, the Ussangu desert in Tanganyika (West, 1910, p. 102, and 1918, p. 1), and more recently from the Linyanti, where it was collected by the writer and Miss E. L. Stephens, and from Kimberley, collected by Mr. Power (Pocock, 1933, p. 492). Possibly, like 530 Annals of the South African Museum. V. tertius in Europe, it is more widely distributed than would appear from the records, since it is fairly easily confused, particularly when young, with V. aureus. To this must be added the very interesting giant form also collected by Mr. Power near Kimberley, and named V. gigas (Pocock, loc. cit., p. 480) (= Merrillosphaera gigas according to Shaw). 2. Janetosphaera, the representative of which apparently differs in no essential from the European species V. aureus Ehren. (=J. aurea Shaw), and is fairly widespread in warmer subtropical and tropical Africa (Transvaal, Zambesi, Egypt), but does not appear so far south as the Cape. 3. Volvox, in the more restricted sense in which Shaw uses it, which is by far the most abundant and widespread, ranging from the Cape to the Zambesi, from Tanganyika to Ovamboland, and found also in N. Africa (Egypt). It is represented by several distinct species and forms, including V. Rousseletii West and V. capensis Rich and Pocock. These three sections of the genus are sharply distinguished from one another by the form of the cell membranes. If dilute methylene- blue solution is used as Meyer (1895, p. 229) directs, the membranes and mucilaginous content of the spheroid stain purple, the proto- plasts blue, and the form of the membranes can then be seen with very little difficulty. The three types of membrane found in the African species have already been described and figured (Rich and Pocock, p. 431, and fig. 1; and Pocock, pp. 481, 494, and fig. 1). The present paper is concerned solely with Volvox in the narrower sense and with two-species only of this section, the dioecious Riet- fontein Volvor, V. Rousseletii, which is apparently widespread in regions of summer rainfall, and the monoecious Cape Flats Volvoz, V. capensis, which has been found at the Cape, in Rhodesia at N’gamo, and on the Linyanti, a tributary of the Zambesi. In the two latter localities it differs somewhat from the Cape form (see Rich and Pocock, p. 452). Habitat of Volvox in Africa. Apart from the records of Volvox africanus from Lake Albert Nyanza, and of V. Rousseletii and V. capensis from backwaters of the Linyanti, all the records so far as can be verified are of collections from temporary pools or vleis (7.e. more or less extensive sheets of water formed either by rain-water collected in shallow depressions, or left by receding flood-water near rivers, as on the Linyanti), which Volvox in South Africa. 531 dry up completely for several months of the year, during which they become grass-covered flats. Even in the case of the Linyanti back- waters, it is possible that the water dries up more or less completely, since this river, like all the large tributaries of the Zambesi, is subject to very great fluctuations in water-level throughout the year. Thus the life-cycle is dependent very largely on seasonal rainfall ; moreover, as the vleis dry up and become shallower, Volvox is sub- jected to very much higher temperatures than are the European forms. At the other extreme, however, it is seldom subjected to very low temperatures, rarely if ever having to cope with even a few degrees of frost. The vleis in which Volvox is found may dry up very gradually, in which case it often happens that the life-cycle has been completed, sometimes more than once, and only the resting form remains, the ripened oospores being hidden in the soil at the bottom of the vlei; but very often the water dries up with astounding ‘rapidity, and Volvor may be cut off at the height of its development. This occurs not only in regions of summer rainfall such as Kimberley, where the great heat causes the pools to dry very soon after formation, but also on the Cape Flats, where the beginning of summer is usually _marked by the coming of strong south-east winds, which lick up the water from the remaining pools almost more rapidly. Instances of this sudden drying of the vleis will be described in more detail later. Votvox ROovUSsSELETII AND VOLVOX CAPENSIS—I. THE ADULT FoRM Distinguishing Features. Before describing the observations on the life-history and habits of the two species of Volvor concerned, it will be as well to summarise the outstanding characteristics which distinguish them from one another. A. Volvox Rousseletii West (1910, p. 101, and 1918, p. 1). The colonies are globose when young, with slightly elongated polar axis, usually becoming markedly egg-shaped when adult, the sexual colonies more so than the asexual (Plate XVIII, B, C). The colonies are of three main types: (1) purely asexual, pro- ducing from 1 to 16 daughter colonies, most often 6 to 8 arranged alternately in two planes (Plate XX, A); (2) male colonies, pro- ducing from about one to several hundred sperm globoids; and (3) female colonies, producing from 60 to 300 or 400 oospores, average 532 Annals of the South African Museum. number 180 (Plate XX, B-F). To these must be added occasional “mixed” colonies containing both asexual daughter colonies and either sperm globoids or oospores (Plate XX XVIII, B, C). Such colonies were comparatively rare in the Rietfontein material, but less so in that from Kimberley. Even rarer were female colonies with occasional sperm globoids, often somewhat abnormal, which were very occasionally seen in the latter material. The anterior pole is always free of reproductive cells, these being as a rule confined to the posterior half in the case of the asexual gonidia, to the posterior two-thirds or three-quarters in the case of the sexual cells. At the height of its growth, under favourable conditions it is one of the largest forms of Volvox yet recorded, occasionally reaching a diameter of over 2 mm., though this is exceptional. In a typical healthy, well-grown strain, however, the mature asexual colonies are mostly over 1-5 mm. in equatorial diameter. The sexual colonies are in general smaller than the asexual, but even they are often over 1 mm. in diameter. The only recorded forms which exceed it in size are V. (Besseyosphaera) Powersii (Shaw), Printz (Shaw, 1916, p- 253), 1-8-2-5 mm. diameter, and the recently discovered ‘ Kim- berley Giant,” V. gigas, which may reach the huge size of 3 mm. or over; both these species belong to different sections of the genus. In the ELu-Volvox section the only form which is possibly at times larger is one from 8.W. Africa, V. amboensis, Rich and Pocock (loc. cit., p. 462). West (1918, p. 2) remarks on the marked ovoid shape character- istic of the mature sexual colonies as opposed to the ‘‘ globose ” form of the asexual, but all the latter shown in his photographs are still young. When mature, the asexual colonies, too, may be ovoid, though in general less elongated than the sexual individuals (Plate XVIII, B-D); the external form, however, varies greatly, being influenced markedly by external conditions. The number of cells composing the colony is large, usually 20,000 to 30,000, though as many as 50,000 were estimated in some cases by Rousselet (1914, p. 393) and West (1910, p. 101); the proto- plasts themselves are small, with well-marked protoplasmic connec- tions, larger and more widely spaced at the anterior than at the posterior pole. Reproductive cells are differentiated by size before birth, but undergo no divisions until some time subsequent to birth. The sperms are formed in more or less completely closed globoids, never platelets, while the oospheres ripen into spiny-coated oospores. Volvox in South Africa. 533 B. Volvox capensis, Rich and Pocock (1932, p. 442). The colonies here are globose when young, and generally remain so throughout their life, even in the case of mature sexual colonies ; sometimes, however, older colonies may become ovoid. There are only two main types of coenobia: (1) asexual, with from 2 to 20 daughter colonies, but, as in V. Rouwsseletii, most often 6 to 8, arranged as in that species ; and (2) sexual colonies producing from 4 to 23 sperm bundles and from 60 to 180 oospores (average number 120). The antheridium mother-cells divide very early, and the sperm bundles have usually matured and escaped before any oospheres are ready for fertilisation, so that a colony with ripe oospores appears purely female; gaps in the cell network, however, show where sperm globoids have formed and escaped. The mature colony is somewhat smaller than in V. Rousseletii— adult asexual coenobia in a strong strain average about 1 mm. in diameter-—and is composed of fewer cells (18,000 to 20,000). The sperm globoids are very similar to those of V. Rowsseletii, and the oospores here, too, are enclosed in a spiny exospore, but the spines are slightly shorter and broader, while the protoplast is larger. This species is very closely allied to V. globator, and is intermediate between it and V. Rousseletii. Further divergences in structure and development are referred to in the course of the ensuing investigation. Parasites of Volvox in South Africa. On the Cape Flats there are two very prevalent parasites of Volvoz. One is amoeboid, attacking and ingesting single somatic cells of either parent or daughter colonies, and often assuming strange shapes. It has been observed frequently in various localities, most often during or just after a spell of cold weather; the most marked occurrence as yet noted, however, was in the deep ditch between the road and the Wattle thicket near Dabchick Vlei (5th September 1932), where, although Volvox was present in very large quantities, scarcely a single healthy colony was to be found. The water of the ditch on being tested was found to have a pH concentration of 8-4, while in the neighbouring pools the water was approximately neutral (pH 7-3) and the Volvor particularly abundant, large, and healthy. The second common parasite is the Rotifer best known as attacking Volvor—Proales (Notamata) parasita (Plate XIV, D); it has been found in many vleis, but in some appears to be more prolific than in others. Frequently several individuals and eggs may be seen in 534 Annals of the South African Museum. one coenobium, or adult rotifers may be watched eating their way into a colony from the outside. A large proportion of the somatic cells may be eaten away before the colony ceases to move, while the developing embryos are more or less completely destroyed, or else are malformed and distorted owing to the activities of the destructive little parasites during the course of their development. While in the case of the amoeboid parasite low temperature and high hydrogen-ion concentration seem to be predisposing causes of susceptibility to attack, in the case of the rotifer long-continuance of Volvox in the pool in question would appear to favour the develop- ment of the parasite, which is rarely seen early in the season. During the subsequent appearances of Volvor as the season advances, however, it becomes increasingly abundant. This has been noted in several vleis, particularly Scanlan’s Vlei (Wetton Road) and Vlei III, near Belvedere Road. The almost complete absence of the rotifer from the Kimberley Volvox—one or two individuals were noticed in some of the last collections—is confirmatory of this, since there the duration of the pool after each rain was very brief, and normally the pool dried completely between successive phases. The crustaceans found with Volvox do not appear to have much power of attacking the uninjured colonies, but may apparently feed to some extent on coenobia which have been distorted or torn in some way. Very occasionally fine fungal hyphae were observed in the mucilaginous content of the hollow spheroids. When badly parasitised, Volvox shows the effect macroscopically by the pale yellowish-green colour of the colonies; if this is due to the activity of rotifers, peculiarly dark opaque bodies (the eggs of the rotifer) inside the colonies may be distinguished from the young daughter colonies by the naked eye. Macroscopic Observations. If active living material of Volvox is placed in a clear glass tube and held up to the light, a number of features are distinguishable by the naked eye. In a quantity of a well-advanced strain of V. Rousseletii so examined, the following details may be seen :— The colonies are a bright, clear leaf-green, swimming about actively, rotating about a polar axis which is perceptibly greater than the equatorial diameter in the older colonies, although the smaller ones appear spherical. The pole which is directed forward as the colony moves is lighter in colour than the posterior half, in which in addition Volvox in South Africa. 535 very dark green bodies can in many cases be seen. Usually the axis lies somewhat obliquely during movement, due probably to the greater weight of the posterior half. In the larger colonies the arrangement of the daughter colonies in two planes may sometimes be made out. Other colonies appear bright reddish orange—these contain maturing oospores, the red colour of which masks the green of the vegetative cells. Such reddish colonies, when present, are very conspicuous among the predominating green of the majority. Among the latter closer inspection can differentiate various types, besides those containing daughter colonies. Some are darker green and very finely punctate; these are young female colonies whose eggs have not yet changed colour. Others are lighter in colour than the asexual colonies, and they too appear finely punctate except at the anterior end—these are male colonies with sperm globoids, which are visible among the vegetative cells owing to their greater density and size, not from stronger colouring as in the case of the eggs. Finally, large, clear, nearly colourless colonies may also be seen— asexual colonies from which all the daughter colonies have escaped, or mature male colonies in which few or no sperm globoids remain. Both may continue moving slowly among the younger colonies for some time after the loss of the reproductive bodies, whether asexual daughters or male globoids, before finally degenerating. Movement. With the assistance of a good hand lens the movement of the coenobia may be seen to be twofold, a rotation on the polar axis and a movement forward in the direction of the anterior pole, which is in general directed towards the light. In addition to the colonies which are actively moving, it is usually possible to see others in which the voluntary movement has ceased, and which are as a consequence slowly dropping downwards through the water, the heavier posterior end always lowest. The rotation on the axis is not always in the same sense. In favourable material, the majority of the coenobia can be seen to be rotating in a clockwise, others in a counter-clockwise, direction. If one colony be singled out for observation, it may often be seen to pause momentarily in the rotatory movement and to resume in the opposite sense. Janet (1912, p. 133) and Klein (1889, p. 168) observed the same behaviour in the case of the European JV. globator. If active material be placed in a watch-glass with plenty of water 536 Annals of the South African Museum. and lighted strongly from one side, the colonies can be watched swimming towards the light; in a very short time all the colonies are collected on that side of the glass nearest the light, moving about and jostling one another. This is specially marked if the material has been kept in the dark or badly lighted for some time previously. In half a minute practically all the colonies have passed to the side nearest the light. Reverse the position of the light without moving the watch-glass, and the colonies immediately start in the opposite direction, again collecting on the side nearest the ight. This gives an average rate for the crowd of about 6 to 8 cm. per minute, but some individuals obviously attain a much higher speed (cf. also Klein, 1889, B, p. 47, for V. globator). From a few such simple experiments it becomes obvious that Volvox is strongly phototactic, and that in this pursuit of light the anterior pole is directed towards the light; it would thus appear that this pole is the light-sensitive region of the coenobium. Under natural conditions, of course, the upper surface of the pool is the most strongly lighted, hence during its voluntary move- ment Volvox tends to swim upwards and collect in the upper layers of water. In the watch-glass experiment, however, if the one-sided illumination continues to operate for a lengthened period, there comes a time when the stimulus is no longer effective, the colonies cease to respond, become quiescent, and gradually sink under their own weight, collecting at the bottom of the glass. Similar behaviour may be seen in their natural surroundings ; after remaining motile near the surface for some time, if the light is strong, the coenobia soon become quiescent, and slowly sink. Possibly this may be a question of “ fatigue,” the cells ceasing to respond if the stimulus is long continued. However that may be, the alternate swimming towards and dropping away from the light is probably directly connected with photosynthetic activity—during exposure to light carbohydrates are actively formed, accumulate in the cell until a certain concentration is reached, when active movement ceases and the plants slowly sink under their own weight. Further, it would appear that the stimulus of light is necessary to initiate the move- ment of the cilia, since at night the colonies collect at the bottom of the vessel containing them. Interesting observations on the vertical distribution of Volvox in deep waters at different times of the day were carried out on Lake Monona, Wisconsin (Smith, 1917), but conditions there are, of course, very different from those obtaining on the Cape Flats. Further Volvox in South Africa. 537 observations of the response of Volvox to light, carried out in the field or with large tub cultures imitating as far as possible natural conditions, might prove very instructive in this connection. Con- siderable difficulty, of course, attends such work in the field, and so far experiments on phototropism in Volvor have been mainly con- fined to the laboratory. The most detailed are those carried out by Oltmanns (1917, p. 280), who attributes the sinking during the quiescent period to geotaxis. This certainly seems sufficient explana- tion for this part of the process; even in preserved material, if the tube containing it is shaken up and then left to settle, the colonies can be seen dropping slowly, the axis often slightly oblique, but the posterior pole, which with its closely packed somatic and numerous reproductive cells is undoubtedly the densest part of the coenobium, always on the under side. The specific gravity of the living colony is evidently a little higher than that of the surrounding water. The movement was most beautifully seen on a large scale one morning of February 1931. On that date Mr. Steer’s fishpond * was particularly rich in Volvor—the strain inhabiting it was almost at its optimum development—the clear water in the centre of the pool being green with Volvor. To the north of the pond is a fernery protected by a roof of wooden slats an inch or so apart. The morning sun shining on this roof threw its shadow on the water, making alternate bars of light and shade. In each bar of light was a column of moving green spheroids swimming upwards and moving about near the surface, jostling one another as they moved. The light reflected from the countless moving bodies gave the bars of light a scintillating appearance, like motes in a ray of sunshine, most striking even at a distance, while the bars of shadow, on closer examination, were found to be practically free of Volvor. Among the mass of upward-moving bodies others could be seen slowly sinking without other movement. In the field it is seldom possible to observe the movement so well. To begin with, at the Cape at any rate, there is seldom such a quantity of Volvoxr in so small a space, and, further, conditions are much less favourable for observation. Still, when Volvox has been particularly abundant on the Flats, it has sometimes been possible, by wading out into the middle of the pond and watching carefully, to see similar * A cement pool about 6 by 8 feet in area, 4 to 6 inches deep at one end, 2 to 3 feet at the other, with various water plants, Arrowhead, Cyperus, Limnanthemum, etc., growing round the edge, the centre fairly clear, inhabited by tame frogs, small fish from the Flats, and so on. VOL. XVI, PART 3. 35 538 Annals of the South African Museum. behaviour in the case of V. capensis in both the Ottery and Wetton Road vleis. Henneguy (1876, p. 288) states that in V. aureus the mature female colonies exhibit negative phototropism, actually swimming away from the light. This has not been observed in the case of the African species; indeed, colonies red with ripening oospores have been watched joining in the mass movement, but as they grow older their activity decreases and they tend to collect at the bottom of the water. But when watching a dish of Volvor brought into the light it is obvious that the horizontal movement is not invariably directly towards the light. Some colonies at first swim in other directions, occasionally even directly away from the light ; eventually, however, all become similarly orientated, and move towards the illuminant. Sequence of Asexual and Sexual Phases. In both V. Rousseletii and V. capensis the appearance of the various types of colonies, provided external conditions are favourable, follows the same sequence :— 1. Asexual Phase.—Soon after rain has fallen and pools have formed—if the weather is warm, within four or five days—‘“‘Juv- eniles,” resulting from the germination of oospores, appear (Plates XLVII and XLVIII). They are quickly replaced by normal asexual colonies, which by the third or fourth generation have the full com- plement of cells and sometimes a large number of daughter colonies (cf. Plate XVIII, A). This asexual phase may be prolonged, all the daughters produced being themselves asexual, or it may be short and soon succeeded by the sexual phase, the duration apparently depending to a great extent on the temperature. 2. Sexual Phase.—The sexual phase begins with the appearance among the asexual colonies of occasional sexual individuals; the proportion of sexual to asexual colonies rapidly increases, until the former may actually outnumber the asexual colonies, which are, however, always present (Plate XVIII, B). In V. capensis the number of sexual colonies is relatively smaller than in V. Rousseletia (Plate XII, D). In the latter, at the beginning of the sexual phase, male colonies are often present in large numbers, sometimes out- numbering the female colonies. Usually, however, the latter pre- dominate, while, as the phase advances, the number of male colonies falls rapidly (Plate XVIII, B). It is usually at this stage that mixed colonies occasionally appear; in V. Rousseletii these contain asexual 99 Volvox in South Africa. 539 daughters and either sperm bundles or oogonia (Plate XXXVIII, B, C), while in V. capensis both kinds of sexual organs are usually present. As the sexual phase advances in V. Rousseletii coenobia containing ripe or nearly ripe oospores may actually predominate, whereas in V. capensis the proportion of such colonies is always relatively small; very few colonies red with ripening oospores are, as a rule, found at one time. 3. Second Asexual Phase.—After a longer or shorter time sexual activity begins to wane, the proportion of asexual colonies again increases, and finally all sexual colonies disappear. Sometimes, as this phase advances, e.g. in the Rietfontein cultures, there is a marked diminution in the number of daughter colonies formed in each coenobium ; as few as two, or even one, may be formed. Often, too, the colonies are smaller and composed of fewer cells (Plate XVIII, D). This is, however, not always the case, and is probably affected to some extent by external conditions. Observations on the Cape Flats. From observations made on the Cape Flats in 1931, it was sus- pected that in Volvor capensis at any rate there might be several successive phases during the seasonal existence of any one vlei, but the observations made were not sufficiently detailed to confirm or refute this. In order, therefore, to obtain further evidence on this point, during the season May to November 1932 attention was con- centrated on the chain of vleis running parallel to Belvedere Road, Claremont. These vleis were visited regularly at intervals of a few days throughout the greater part of this period, and the occurrence of Volvor noted in various parts. As a result several interesting facts emerge :— (1) The appearance of Volvox is extraordinarily local—it may be swarming in one part of the vlei and absent from other parts. This is entirely independent of the direction of the wind. Possibly the phanerogamic and larger algal flora may to some extent prevent the spread of Volvox from one part of the vlei to another, but as the same phenomenon was noted in an extensive open stretch of water with scarcely any such vegetation, it seems more probable that the hori- zontal range of Volvor is very limited. If it is found at one particular spot in the vleiit is because it has developed at or very near that spot. (2) In every case it was found that Volvox disappeared and re- appeared several times (most often thrice) during the season. The 540 Annals of the South African Museum. length of the periods during which Volvox was present in any one vlei or part of a vlei varied from two to six weeks, most often lasting for from three to five weeks. The time elapsing between disappear- ance and reappearance at the same spot likewise varied; usually periods of from one to two weeks during which no Volvoxr could be found separated each two successive phases, but in one vlei on one occasion Volvox disappeared for as long as five weeks. In the latter case the prolonged disappearance was both preceded and followed by comparatively lengthy (four and and six weeks respectively) periods of great activity, during which large numbers of oospores were produced. (3) “ Juvenile’ colonies, resulting from the germination of oospores, were collected at the beginning of the season only. Judging from observations made elsewhere (notably near Kimberley), the obvious explanation of the reappearance of Volvox after each total disappear- ance is, that some of the oospores formed during the previous phase ripen quickly and germinate at once without undergoing a period of drying, and it was hoped that this would be demonstrable by the presence of Juveniles at the beginning of the successive phases, but none were found. Nevertheless, the failure to detect Juveniles cannot be regarded as proof of their absence. So far, whenever Juveniles have been collected, it has been soon after the pool has formed (3 to 5 days) and usually while it is still small. Hence, firstly, the many oospores dormant in the soil have all, after a period of dryness, been suddenly re-immersed in water under the same conditions, and it is only to be expected that large numbers would in such circumstances germinate simultaneously, whereas when oospores germinate without previous drying it is probable that germination is sporadic, spread over a longer time, and depending in part on the age of the various oospores. Secondly, the very small size of the Juveniles (about that. of large Eudorina colonies) must be taken into consideration—in larger expanses of water the difficulty of collection and detection is enormously increased unless they are present in very great numbers. The present field observations then can only be taken as showing that several successive growths of Volvox capensis may occur in a given piece of water during the season. The probability is that this is due to the germination of oospores formed during the preceding phase, of which some at least may undoubtedly germinate without. previous drying, but more exact investigation is necessary to prove it, since so far field observations give negative results. The difficulties Volvox in South Africa. 541 of exact observations in the field are obvious, and so far the Flats Volvox, unlike the Rietfontein one, has proved peculiarly intolerant of culture conditions. Incidentally, observations of Volvox in Africa suggest an explana- tion for the non-observance of Juveniles in V. globator—it seems that these are formed in large numbers at one time, only when rain has fallen and refilled the pool after a dry period, and that when oospores germinate without drying, germination is much more sporadic. Con- sequently the number of Juveniles present at any one time is compara- tively small, and their presence proportionately difficult to detect. Observations on external conditions, in particular temperature and hydrogen-ion concentration of the water, were carried on side by side with those on Volvox. Here again more extensive observations are necessary before it is possible to arrive at definite cenclusions. At present, however, the following facts seem to be fairly well established :— 1. Effect of Temperature.—Other conditions being favourable, comparatively high temperatures are conducive to increased rapidity and extent of development. The most luxuriant growths of Volvor on the Cape Flats have always coincided with warm, bright weather. In May 1931 a spell of unusually hot weather at the beginning of winter, immediately following good rains which had caused the for- mation of small pools, resulted in a very rich crop of Volvor. Again, in November of the same year, one of the richest growths of Volvox as yet observed on the Flats occurred in Belvedere Road Vlei III, when the temperature of the water at midday was over 90° F. on several successive days. In 1932 there were no very warm spells during the winter; in the vleis under observation Volvox was never very abundant till well on in the season when, with warmer weather, there came a succession of very rich crops in several pools—e.g. De Klip, Belvedere Road Vlei III and parts of Vlei II. 2. Hydrogen-ion Concentration.—At the Cape the water of those vleis and ditches where Volvoz occurs is always either approximately neutral or slightly on the acid side, the pH ranging from 6-5 to 7-3, most often from 6-7 to 7-0; on the other hand, the pools producing Hydrodictyon are almost always on the alkaline side, pH 7-6 to 9-0, and very rarely are the two algae found in the same vlei. Occa- sionally H. africanum and V. capensis occur together, e.g. in the Wood pool next to Dabchick Vlei, but so far the writer has never found Volvor associated with the second form of Hydrodictyon (H. indicum ?) which occurs on the Flats. On one occasion, in a 542 Annals of the South African Museum. ditch at the side of the road near the farm Vaderlandsche Rietvlei, a very rich growth of the latter Hydrodictyon was found, the pH of the water being 8-3.