rh ttn hey see 72 RA OU Roe eels is , 7 bites Lee OF ; aft iden Hae weitt ti y fos Us Hi SSS See et “ SY a ay ei ahh an " Mire aT ; ey f at LR ey : en th f ee ie % tr Ree 1 j a (ates . Peet fini ; ~~ ees eA ae Th me wed a, Se <= a 3 ye aie ¥ Pine aR TS 4 ay tented nie Goh : tae fi | | . MIA ey ates Ys AC, as EA tk fea nok ett harp | Whe dk ; | | ¥ b : | 2 | | ee ; ae { ay 4 ty erat rt? thet ; ie | : e aie AN Ast | | 4504 ee i ; | | | | Me BN oy taicetee Titers 4 le URED Ney Baninmacan es Pitot eet Weta 2g ( ( : ett Has hurieseseh ise : ) Reale Melero teieeret 4 Pare Chat i} < : Pea RMP iat hc ; : és } é oat Hk Sages ey Ly deF: a ttrake | : | i EMMA oie Led Seer us yaks wit : 5 | | : cat sneha s Alt ace Rit : H . : | | ‘ a loty 554 be hp ta petted Kathe! sts ; eek : ae hence | , | 4 feet ie rly 8 ay if > | Mase ch ie ; iy mith iat ee 4 stat be aed it r Meseseee i, 4 eet ree Sree f niith anes Naas Soe % Hf Ne + - Ai eit of “| . , ¥ 2 is ny mea ‘ thy aS i Gee An ie aH < oe Ae oe ae re ey ase yee x 7 ai eat Ee Pea 1m ae) seniieneans Het a S: abs | O05" 7 ‘ , 7 5 HY : a 1 4 yu cise bo Be, rvs ee ; ' - “ 2 Ss i a - 4 . + J ” . Bulletin of the patorres GENER British Museum (Natural History) Botany series’ Vol 16 1987 British Museum (Natural History) London 1987 Dates of publication of the parts Nol. : ; : : ‘ ; . ‘ : ; 29 January 1987 No2s:, : F : ; ; : ‘ : : : 25 June 1987 No3. : é : : : ‘ : : ‘ : 27 August 1987 No4 . : ; : : : : ; : : 26 November 1987 ISSN 0068-2292 Printed in Great Britain by Henry Ling Ltd., at the Dorset Press, Dorchester, Dorset No 1 No2 No 3 No 4 Contents Botany Volume 16 Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) N. K. B. Robson ; : The lichen genus Ramalina in Australia G.N. Stevens . An annotated list of vascular plants collected in the valleys south of Mt Everest G. Miche. Further genera of the Biddulphiaceae (diatoms) with interlocking linking spines | R. Ross and P. A. Sims Page 225 269 o ei =) Bs Baie ae da Saar Sy eT £ Bulletin of the British Museum (Natural Bio) Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) Norman K. B. Robson Botany series Vol 16 No 1 29 January 1987 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum’s resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and several volumes may appear within a calendar year. Subscriptions may be placed for one or more of the series on either an Annual or Per Volume basis. Prices vary according to the contents of the individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.) © Trustees of the British Museum (Natural History), 1987 The Botany series is edited in the Museum’s Department of Botany Keeper of Botany: Mr J. F. M. Cannon Editor of Bulletin: Mr J. R. Laundon Assistant Editor: Dr A. J. Harrington Editor’s Assistant: Miss M. J. Short << SH MUSED v YBLICA MONS ISSVED Lig if XN ISBN 0 565 08014 8 Hs = ISSN 0068-2292 ies ams ees a * Ts Botany series * Pi Se % Vol16No 1 pp 1-106 British Museum (Natural Hist py WALES OFF a Sy Ss he Cromwell Road oe TURAL His SF0x London SW7 5BD Issued 29 January 1987 Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) Norman K. B. Robson Department of Botany, British Museum (Natural History), Cromwell Road, London SW7 5BD Contents A AEs Desk | DET OANA TOS PRR OC ARPT Ca OAS THERON PETS PET ONT EAE Ee REIET OST ey Ae 1 RS ICRC GI CHOIG rosa sc ran ch Saco eke tene eddy Veoh eve edad e coney ta saueseboe ye ceevans diol Snassesuaeestorrs 2 WIDEB OLOGY 5. cries seer tat esse neers dousso- ca iea carwacsetler shun coene aw on opesseanak Hine eed des the 3 ERP UASCAVES: Mireles h tanks ce scrut oo tata nese ih doane coer or Poe aecdee ane tae ters tetas 5 RU PMEOT SCONCES aa tre ecred seen nccan Usnechtumde noany sas eine cdua sk Vin soe eee b Wey PIG WETS ain SUNS oso ny usa noe sen nee hohe Ca sdreane aint de ating a ica ne itconst 5 ESVCOIOG Ms IY DIIIS o25 esc lays ots suey roo etay cae oes cattechac scree asansetsaitar aie? ol PPISUTIOULIOM AIO VOM, o, 00 sate rss onea rts Sa one onesies nase va cseaweeeadedeadiy bees} it CO) WV IGE CIS IUNICHIONS Cort. onersescccrnts oes nce) car is iis dost cats Soe oa ee tae oad coerce ei (b) Species with leaves persistent or deciduous near the base SDR IS 25 ee aay eee Rie eds IYER A A teed 9 (c) Species with leaves with sheathing leaf-bases or marginal Secrenons(Spps 24226. Ola oid het a vetet eck cmt eer smuaien: 10 (d) Species with incurved-cucullate leaves but no marginal SECTELONSI( SPP 2h SU) ecu cst siies oo sate ee ales ce ee cae ma wma eaannd dee 10 (e) Species with sessile leaves deciduous above the base but notincurved-cucullate (Sop; 31.0: —48 )s.25;.< foc-anwns cos cc tetas sees -kaceen 11 (f) Species with pseudopetiolate leaves deciduous above the base but not incurved-cucullate (Spp. 31p.p., 49-60) «20.0.0... eee ce ec ee eee ee es 12 em Phat LOLOL on URES meee oie Pere Sea ayer e Pon ie in Pee tee aM REET Cate ESSN 12 eRe MNO WICCEEMIONES serit cau tnsy aac esier etn ere ceed As sous cane Teneo au a euta mes toes ea teaaoree 103 EN ong Lo} £1 2h coed SCRE Eee Oe re RNS tae eon AOE PEE ACRE PL RET OEN ES RO ee ee 104 Pic AnsVS CCA UIC MINOR oo ca. ors cassie tere cso see na Ren Ce A artist Rumen ad | ps arean aN ae eam auttae 104 Synopsis A systematic account is given of the 61 shrubby species of Hypericum sect. 29 Brathys, which occur from Belize and the Greater Antilles to Peru and Bolivia. A discussion of the morphology, chromosome number (only one known), distribution, and evolution of the members of the section is included. Sect. 30 Spachium sensu Robson (1977) is shown to comprise three distinct groups: (i) sect. Spachium sensu stricto (Central America, Greater Antilles, eastern U.S.A.), which is included in sect. 29. Brathys; (ii) sect. Knifa (Adanson) N. Robson, stat. nov. (S. American Cordillera, Central and North America, Cuba, tropical and southern Africa, east and south-east Asia, Australia); (iii) sect. Trigynobrathys (Y. Kimura) N. Robson, comb. nov. (S. America south of Amazonia and along the Cordillera, Australasia, south-east Asia). The following other new taxa (sp. nov. or subsp. nov.) are described: H. phellos subsp. oroqueanum N. Robson (Colombia: Norte de Santandér/César), H. acostanum Steyerm. ex N. Robson (Ecuador: Loja, Morona-Santiago), H. irazuense Kuntze ex N. Robson (Costa Rica), H. simonsii N. Robson (Colombia: Magdalena), H. papillosum N. Robson (Colombia: Boyaca), H. martense N. Robson (Colombia: Magdalena), H. maguirei N. Robson (Ecuador: Azuay, Loja, El Oro), H. gleasonii N. Robson (Colombia: Norte de Santandér), H. prietoi N. Robson (Ecuador: Azuay), H. cassiopiforme N. Robson (Peru: Amazonas), H. magdalenicum N. Robson (Colombia: Magdalena; Venezuela: Zulia, Mérida), H. valleanum N. Robson (Colombia: Valle de Cauca), H. sprucei N. Robson (Ecuador: widespread; Peru: Piura), H. recurvum N. Robson (Peru: Amazonas, Junin, Pasco), H. wurdackii N. Robson (Peru: Amazonas), H. costaricense N. Robson (Costa Rica; Panama; Colombia: Bolivar, Antioquia), H. bolivaricum N. Robson (Colombia: Bolivar), H. parallelum N. Robson (Colombia: Norte de Santandér/ Bull. Br. Mus. nat. Hist. (Bot.) 16 (1): 1-106 Issued 29 January 1987 2 NORMAN K. B. ROBSON César), H. marahuacanum N. Robson (three subspecies): subsp. marahuacanum (Venezuela: Amazo- nas), subsp. strictissimum N. Robson (Colombia: Boyaca; Venezuela: Mérida), subsp. chimantaicum N. Robson (Venezuela: Bolivar), H. horizontale N. Robson (Colombia: Norte de Santandér, Santandér), H. jaramilloi N. Robson (Costa Rica; Colombia: Magdalena, César/Norte de Santandér), H. llanganaticum N. Robson (Ecuador: Tungurahua, Napo), H. selaginella N. Robson (Colombia: Boyaca, Arauca, Cundinamarca), H. cymobrathys N. Robson (Colombia: Boyaca). The following changes of rank (stat. nov.) are also made: H. phellos subsp. platyphyllum (Gleason) N. Robson, H. loxense subsp. aequatoriale (R. Keller) N. Robson, H. caracasanum subsp. turumiquirense (Steyerm.) N. Robson. Two others are also new combinations (comb. et stat. nov.): H. strictum subsp. compactum (Triana & Planchon) N. Robson, H. lancioides subsp. congestiflorum (Triana & Planchon) N. Robson. 1. Introduction It has proved to be impracticable to complete the systematic part of this monograph of Hypericum L. in the order in which the sections were treated in part 1 (Robson, 1977). This account of sect. 29. Brathys, which is part 7 of an estimated nine parts*, has been completed next after part 3 (sects 1—6a) as a result of work on it for Flora Neotropica. The distributional area of the shrubby members of sect. Brathys (Belize and the Greater Antilles to Bolivia) falls wholly within the area of that Flora. The publication of part 7 out of order has necessitated the abandoning of the double enumeration started in part 3, where each species has a running generic number as well as a sectional one. In parts 4-8, therefore, the species will have only sectional numbers; but a continuous enumeration will be included in part 9, along with addenda, corrigenda, and a revised sectional key. The division of sect. Brathys sensu lato into two sections, 29 Brathys (trees, shrubs, and shrublets, centred north of Amazonia) and 30 Spachium (subshrubs and herbs, centred south of Amazonia) has proved to be incorrect. Sect. Brathys in fact includes a group of subshrubs and wiry herbs that contains the type of sect. Spachium (H. gentianoides (L.) Britton, Sterns & Poggenb.), thus extending the area of the section northward to eastern U.S.A. In addition, the herbaceous H. piriai Arechav. (including H. hilaireanum L.B.Sm. = H. tenuifolium St. Hil. non Pursh) represents a development from Sect. Brathys in which the flower has apparently become adapted for specialized insect pollination in a manner partially analogous to those of sects Elodes and Adenotrias. It has been treated as an extension of variation of Sect. Brathys, and its inclusion extends the area of this section south of Amazonia to south-eastern Brazil and Uruguay (Fig. 1). The remainder of sect. Brathys sensu lato divides into two sections. One of these is centred (i.e. has the species with the most primitive characters) in the Andes of Venezuela and adjacent Colombia, extending southwards along the Cordillera to Chile and northwards to Canada; and it also occurs in SE Asia and Australasia (H. japonicum Thunb. ex Murray) and Africa and Madagascar (H. /alandii group). Until now this section, which includes both shrubs and herbs, has not been recognized as distinct and therefore has no name. It does, however, include the type species of the genera Knifa Adanson (H. mutilum L.) and Tridia Korthals (H. japonicum Thunb. ex Murray), both of which were treated as series of Sarothra sect. Spachium by Kimura (1951). I have therefore chosen Knifa as the epithet for this new section: Hypericum section Knifa (Adanson) N. Robson, stat. nov. Knifa Adanson, Fam. Pl. 2: 44, 541 (1763). Tridia Korth. in Tijdschr. Natirl. Gesch. Physiol. 3: 17(1836). Sarothra sect. Spachium series Knifa (Adanson) Y. Kimura in Nakai & Honda, Nova FI. Japonica 10: 233 (1951) ‘Kniffa’. Sarothra sect. Spachium series Japonica Y. Kimura, tom. cit.: 233 (1951). Type: H. mutilum L. (Y. Kimura, 1951, lectotype). The other section, comprising shrubs, suffrutices, and herbs, is centred in south-eastern Brazil and extends in area to Uruguay, northern Argentina, and Peru and thence northward along the * For part 1 see Robson (1977); for part 2 see Robson (1981); for part 3 see Robson (1985). THE GENUS HYPERICUM L. 3 Cordillera to Venezuela. It also occurs in Australasia and SE Asia (H. gramineum Forster f.). The type of Kimura’s Sarothra sect. Trigynobrathys (H. myrianthum Cham. & Schlecht.) belongs to this section (although that epithet is not especially appropriate to the concept adopted here), as well as the type of sect. Brathys subsect. Connatum R. Keller (H. connatum Lam.): Hypericum section Trigynobrathys (Y. Kimura) N. Robson, comb. nov. Sarothra section Trigynobrathys Y. Kimura in Nakai & Honda, Nova Fl. Japonica 10: 233 (1951). Type: H. myrianthum Cham. & Schlecht. (Sarothra myriantha (Cham. & Schlecht.) Y. Kimura). The characters by which species in these three sections may be recognized can be summarized as follows: 1. Leaf lamina with short, sometimes thick basal insertion, pairs free or united by interfoliar ridge or very rarely more broadly united; inflorescence-branching usually pseudo-dichotomous or MONOCHAS A door oe oe, ote Pace cule iatened te eevee ye edtraa st wo tot eee sect. 29. Brathys Leaf lamina with long narrow basal insertion, pairs free or united by interfoliar ridge or by lamina proper, sometimes with bases decurrent or leaves perfoliate; inflorescence-branching pseudo-dichotomous or monochasial or dichasial\, 24.0. 2..4/:..39: fuk: seteeccesacaceuvanaececesuewsesee 2 2. Leaves narrowed at the base or wholly narrow, united by interfoliar ridge, with basal vein single, venation pinnate or 1-nerved, or, if (secondarily) amplexicaul and 3—5-nerved (H. majus and H. lalandii groups p.p.), then midrib branches absent of weak.... sect. 30. Knifa Leaves amplexicaul with 3—7 basal veins and midrib clearly branched, the leaf pairs usually GOCUITENLOK SODIGHIMICS perloliate <2. .ss0425.so0eascag entanssucsion ters sect. 31. Trigynobrathys Morphology (Fig. 2) The species in sect. 29 Brathys vary from small trees (up to 4 m in height) to prostrate shrublets or wiry herbs. They all lack dark glands completely and have no resin glands on the lower surface of the leaf (the latter are common in sect. 3 Ascyreia but not in sect. 1 Campylosporus), and there are no fringing glands on leaves, sepals, and petals. It appears that all have marginal punctiform leaf glands (although these are often very small), and the laminar glands in the leaves are also always punctiform. In the sepals and petals, however, the laminar glands vary from wholly linear to wholly punctiform. (a) Leaves The reduction of the laminar leaf glands from lines to dots is not here associated with the development of reticulate venation, as it isin sect. 1 Campylosporus (Robson, 1981: 80, Fig. 11). In the species of sect. Brathys with the most primitive characters (1 H. terrae-firmae) the venation is parallel and open, even though the laminar glands are punctiform. The primary venation remains open in nearly all the woody species, thus allowing reduction to 1-nerved leaves to occur in several parts of the section. Only in some of the relatively primitive species, e.g. 24 H. magniflorum, is there a clear tertiary reticulum. In sect. Brathys the ultimate fate of the leaves has proved to be of major taxonomic importance. In the most primitive state (found in 1 H. terrae-firmae and 2 H. styphelioides) the leaves are free and completely deciduous, leaving a rhombic leaf scar. In all other species they are wholly or partly persistent, the persistent part being eventually shed with the cortex of the stem internode; and the leaf bases are usually united by an interfoliar ridge. The ultimate fate of these leaves depends on their thickness and size. Where they are thick and long, they break off without fading, leaving a perceptible part of the base (25 H. gleasonii to 61 H. cymobrathys except 54 H. millefolium). Where they are thin and long, they tend to droop and wither on the stem (3a H. phellos subsp. phellos in part, 10 H. simonsii, 11 H. papillosum, and 24 H. magnifiorum). Where these thin leaves have (in evolution) become smaller or narrower and thicker, they either fall at the apparent base (3 H. phellos and 4 H. garciae) or persist and wither (Spp. 5-9, 12-23), tending to become secondarily appressed if the base is narrow (e.g. 8 H. stenopetalum) and depressed if it is broad (e.g. 16 H. cuatrecasii). Finally, in 54 H. millefolium NORMAN K. B. ROBSON ‘guT] poydnis9jzUl ue Aq po}edIpUI Ie WeISeIP OY} 0} JURAD[OI UOTRLIVA SuIMOYs exe} JO s}IWIT *g Weg Ves ‘~g—-z9 “ddg jo sjlejap JoO.J ‘sUONIEs poyefo1 pue satoads Cg oY} Jo sdrysuonejoy :skywsg 67 “OST “BLA wnujpue Byourbejes 09 8S / < seplolouR} YS c MmRC HONEA / ¢ wa sepioosni SS ojjjwesel 6¥ wno}jeuebueyy 9 c wnueseoeied yuewye LS ' cS eeuopsed wnojuejepse@ 0S jwerep ; fife wnyoyouRl ey z € x \ s a 3 va wn Aydouskd ~ €¢ / WNHYOsoT WwW VS sepioojowid le [16 |0€| L sAyyesqowAo [08+]L9 meqenys LE ws € \ ' ‘ 4 ‘ wnjzo144s g v @]eB}U0Z10Y oo 9V vyy wnusoenyesew UnUsenss wnyd13sb1}83 “dss wnueoenyesew € € wnuwyssi39138 “dss wnoyejuswiyo “des + wnhueroenyesew WNIPBAIOG 0 ad 1eonids qVvV wnueoenyeiew a a wnulediunt L Vv wnjeJ1jsoid ve Als esSusedliBysOo 8e Wyoep sepioAiq 6€ sdiysuoijejoy skyjeig 62 98S LE 1O}e11d 8 rf esuesow ‘dss esuesejo ‘dss uaa 2 oy ee OS a seployeyudsys “dss seproyeydiys G of [19] eewsIj-88s110} | ES juinysnoue, (it) Lo} : sueoseoiew \ VZ~ \ ors sowjeud ; wnsoyiubew ~ ¢! ) i cccd yuoseel6 suowls- p " vA INBOISJOAIP, , aA Ole og aot pe wnsoyded ‘ s , ' wnueoxew L l ,sueyed, 97 we ay (iit) re) Ce \ Ge Sele RY \ wnsoujeo 6 M / 8 \,wnuewes, , \ . DE vs wnjeyedoue3s \ ‘ sepioipodoohy iy / i ‘¢ h 4 Zl 4 9S sepioAnyy 6 G ewsoj;cdoisseo wnueenboJo ! wnipuecep 4 " wnyAydayeid “dss sojyoud evjoi08 @) 3 / 1 ae sonoud q ¢ y i t : ! seu hob 2¢ L ! t ene na Soe ‘ c lL i ~~ _wrnonzeoyséu euejnoioe " 2 \ ¢ wnueysooe InM oe Se wnasnoed "Wseoes}eNd 91 ewsosuiqes Be esueyinb esuexoj WNyOsose] va él 7 teed 81 ee z L rf esueyew 9 esuenzel Pl C n6emyey 1eyaid THE GENUS HYPERICUM L. 2s the leaves are very small and remain attached to the stem, falling with it either when it disarticulates at the nodes or with the cortex. In those taxa in which the leaves fall at the apparent base, the true base has become united with the stem, surviving only as a protuberance (e.g. in 4 H. garciae). Such leaves can be distinguished from the wholly deciduous leaves of H. terrae-firmae and H. styphelioides by the absence of a true leaf scar. The order of these leaf characters in evolution would appear to have been: wholly deciduous — wholly persistent — partly deciduous — wholly persistent again (Spp. 5—9 and 62-80), see Fig. 2. The leaf base mostly remains parallel-sided or only slightly narrowed before widening to form the interfoliar ridge; but in one group (31 H. pimeleoides in part and Spp. 49-53, 55, 59 in part) it narrows perceptibly to form a pseudopetiole. Where the lamina has narrowed to the width of the pseudopetiole, however, members of this group cease to appear petiolate and have to be described as (secondarily) sessile (Spp. 54, 56-59 in part). (b) Inflorescences The inflorescence in woody members of sect. Brathys is nearly always 1-flowered (the primitive state in the genus). Only in a few of the more primitive species (e.g. 3 H. phellos, 26 H. mexicanum, and 61 H. cymobrathys) is there any development of cymose branching; and even then it is not usually constant and never elaborate. Apart from this relatively weak tendency to produce monochasia, the inflorescence in woody species is always strictly a synflorescence with variation resulting from different types and aggregations of 1-flowered branches. The basic form is the pseudo-dichotomy, i.e. the development of equal foliate branches from the axils of the leaf pair immediately below the flower. This is the only type of acropetal inflorescence development in the section; but the upper lateral branches are also often floriferous, and the relative frequency of these two modes of branching can be of taxonomic value. Thus in 10 H. simonsii, for example, the flowers are solely on lateral branches, whereas in 21 H. martense the majority of the pulviniform plant consists of a pseudo-dichotomous synflorescence, with the branches of the latest pseudo-dichotomies bearing lateral flowering branches (Fig. 8). The development of a suffruticose or herbaceous habit (6 H. piriai and Spp. 62-80) is accompanied by the elaboration of monochasial cymose inflorescences. (c) Flowers and fruits In section Brathys the perianth is normally pentamerous, with the sepals appressed to the petals. These are longer than the sepals and usually spreading (‘flowers stellate’), but in small flowers they may remain ascending (‘flowers obconic’). The sepals vary in width and hence in the number of nerves, which may be of taxonomic significance, as may also be the aspect (plane to incurved-cucullate). Variation in petal shape and size may help to separate related species; but the degree of development of the apiculus, though striking, is not discontinuous enough to be of taxonomic value. The number and size of the stamens are correlated in general with flower size as a whole, but the woody species with the fewest stamens (35 H. decandrum, which can have as few as 5) does not have the smallest flowers. That distinction belongs to 42 H. prostratum, in which they can be as small as 4 mm in diameter. Among the herbs H. gentianoides can have as few as 5 stamens and also has the smallest flowers. The styles vary in number from 5 (H. terrae-firmae, H. styphelioides) to 3, 4 as usual being an apparently unstable number, not characteristic of any one species. The majority of species have 3 styles, but in three advanced woody species this trend has been reversed so that an increased number (4—5) of styles is associated with very small flowers (41 H. juniperinum = H. brathys), 42 H. prostratum, and 60 H. selaginella). The form of the styles (outcurving, incurving, etc.) may be taxonomically useful. The stigma is primitively narrow but has frequently become broader, usually in a capitate form; in one group (Spp. 51-53) the style apex is clavate. The ovary and capsule remain enclosed in the perianth except when the flower is open, and so it is not always easy to observe them in herbarium specimens. Nevertheless, their shape varies and can be utilized taxonomically. NORMAN K. B. ROBSON ‘oul poydni9jul MOLeU B Aq poyedIPUT 91k WRITeIP 9Y} 0} JURAD[OI UONLLILA SUIMOYs BXe} JO s}IWTT “soloeds pa]A}s-¢—p Jo soouarins90 s1ydiowode paxejOst 2Y} 9J0N “SUOIDES po}e[a1 PUL UONDeS 9Y} UIYIIM sioJVIeYS UTeIII9 Jo sw] :skyso4g 67 “OS 7 “BL 8S eyisses x -- see se GO Gy. £ ¢ _7eye}01edopnesd” s pe a RS GS / LoS oy / IS . a / \ & go ~ a OF as es Rep Ee ee — ve ae y \ ae iG ‘ aes cv a Sy Ae 9V vyy ev ae opp avy / OY ates ge 6€ si9joeleyyO skyjeig 62 }998S ce ve LE 8c oma -\~ ' ‘ you / kyued Wa —— 4 \ ‘ H t \ “ i fas \ Wersy i nO} \ (1A) 4 ’ ’ Leis ; N2Gc4 67é ’ I A bd¢ ; I ‘ : Leese 1 \ 7 ue Ras cl \ i ae ys ' ‘ i 1 \ t, ' ‘ ras , \ i cr \ Ts ri ' i 2¢ : ‘ as i H vl ‘ Li ! ne a ‘ a A & L et \ c l \ , y Aare ’ \ of a = ee re \ ra ol MR 0S / a - rd cc aS x ‘ ‘\ ! de él \ f N H OZ ke VL L Ses. \ ic ¢ LG 8l \ ; lew a LR ERSTE PE Tech ed y Pury Gg ese eae, ape SN ws SEER Be SUIOA idea} jeseq . ‘ 1 é Pa ~ Ww“ r ree wwe wee ee ewe - 7 ‘ t 4 THE GENUS HYPERICUM L. yi Cytology and hybrids Very little cytological work has been done on the major part of Sect. Brathys, the only recorded chromosome count for a woody species being n = 12 for 7 H. irazuense (Part 2: table 7). As this is the primitive number for the whole genus (Part 2: 150), there is no available information about possible chromosome variation within the woody members of the section. The numbers n = 6, 12, 24 have been recorded for herbaceous species (see part 8). Hybrids in sect. Brathys are apparently very rare. The only specimen seen that might be a hybrid was collected in Costa Rica (J. & C. Taylor 11757 (NY)) and looks like H. irazuense Xx costaricense. Distribution and evolution (Fig. 3) In parts 2 and 3 it was shown that sect. 29 Brathys is directly related to the basic, African sect. 1 Campylosporus, the taxon in it that is most closely related to H. terrae-firmae being H. revolutum subsp. keniense from the East African mountains (Figs 1-3). Sect. 30 Spachium, as has been shown above, must be divided into three parts, the name-giving part: being included in sect. 29 Brathys. The other two parts, sects 30 Knifa and 31 Trigynobrathys, are more closely related to each other than either is to sect. Brathys. All three sections, 29-31, would seem to have stemmed from an early westward development from the original Hypericum stock in the eastern part of west Gondwanaland, i.e. in what was to become Africa. This western group appears to have divided into a northern (sect. Brathys) and a southern subgroup, the latter also having divided into a northern group (sect. Knifa) and a southern group (sect. Trigynobrathys) as a result of (climatic?) changes in the Amazon region. It clearly migrated into what is now South America while contact between Africa and eastern South America was still possible, 1.e. until the Lower Cretaceous (Turonian—100m yrs BP) (Raven & Axelrod, 1974; Owen, 1976; Howarth, 1981); but whether it had split into two before or after reaching that region is uncertain. From the evolutionary divergence pattern of Hypericum as a whole (Robson, 1981: Fig. 73; Robson, 1985; Fig. 1), it seems likely that the first split occurred in Africa. For a temperate genus such as Hypericum that is confined in tropical latitudes to high altitudes, the ranges of the Cordillera act as highways of dispersion to which the species are restricted and along which they migrate and differentiate. It is therefore often possible to trace presumed evolutionary lines (clades or, more accurately, morphoclines) in sequence along one or more of the Andean ranges. If the variation is continuous, as it is for examples in 20 H. laricifolium, no infraspecific taxa can be recognized, even though the forms at the extreme ends of the range of this species (‘H. /aricoides’ in Venezuela and ‘H. acerosum’ in Peru) look very different from the primitive form from northern Ecuador. If the variation becomes discontinuous, on the other hand, one finds a geographical replacement series of taxa each more advanced (apomorphic) than the one before. In such a clade it is frequently difficult or impossible to discover any apomorphic character in the less advanced taxon that would enable one to conclude that it had continued to evolve independently subsequent to its isolation from the more advanced one. In such clades, therefore, where it is difficult to believe in the Hennigian maxim that one taxon is automatically altered when it gives rise to another (i.e. that at each speciation the cladogram must dichotomize), I have depicted the relationship (Figs 1-3) by a line, not a dichotomy. (a) Wide disjunctions The majority of the primitive species in sect. Brathys occur in, or are confined to the mountains near, the Colombian-Venezuelan border area or to adjacent areas in Boyaca (Colombia) and Mérida, Tachira or Trujillo (Venezuela); but the most primitive (plesiomorphic) characters are found in the species from Belize (1 H. terrae-firmae) and Cuba (2 H. styphelioides), species that are so closely related that they have sometimes been united under the latter name, e.g. by Standley & Williams (1961). The question therefore arises as to whether (i) the original members of sect. Brathys migrated and differentiated southward from Belize/Cuba to Col- ombia/Venezuela before isolation of the two groups, or (ii) the differentiation took place NORMAN K. B. ROBSON “oul] poydnisajur ue Aq poyeoIpul Ie WeISeIP dy} 0} JURAIIOI UOPLIVA SUIMOYS BXk} JO SUIT “1x94 94} UI Passnosip Sape]d dBIPU! SoINSY ULWOL [[RUG *[e1JUSO PU YINOS “}soM ‘JSBO ‘YJIOU 9JLOIPUT $19}}9] BSBI-1OMOT *( (1OUTUL) pur] ssoroR -- - - ‘(4ofeur) pur] ssoroe —— ‘vas ssoioe ==) suonountsip 1ourw pue rofew SuLMoys sorseds Apoom [9 94} Jo uoNQINSIG :skyw1g 67 °109S ¢ “BI BlenzeusA A euBAnD 5 e891y BISOD YO | l Aen6nun Mab F] bi laba foto) ao 9-9U09 6) 9 | 1€6) — sissies a reese a eiiog og 37 / Oe] eH OH band no x S-M3d 6 C MA-uUg (e) I l \| bz ese VL Cc / 2U09 (4) @) I] 3a 1] Moone a ee 9+MND x8) Ae xx) ae 2) q 4 : [0819 One Pe tue 8 23 VSS ee, Sia ae Beet = NU ee pe ee eS ae ee 9S 3-909 Pass 6¥ (xix) [ € oo le MA-2U09 2uU-eUu0D MA-2UQ9D LV eSA 27-8 83 rad x Ve D3d-MA él 8-09 83 eu-2U09 LE 9-uad OS a Fi uad-83 6€ 9€ EZ vod |Z ieee Sl n-esya UO!}IN!IySIq skyjeig 6zZ }0eS THE GENUS HYPERICUM L. 9 northwards from South America to the Greater Antilles before isolation, or (iii) a uniform population was split and the two groups evolved in isolation (Fig. 3a/b). What is clear is that this distribution gives strong support to the theory that the first land link between South America and North American was by way of the Greater Antilles (Rosen, 1976, 1985; Robson, 1981: 213; Melville, 1981: 247). The remaining evolution of the section has been wholly within South America, except for: (i) four species that have reached Costa Rica/Panama after the establishment of the second land link (Isthmus of Panama), (7 H. irazuense, 38 H. costaricense, 49 H. jaramilloi, and 50 H. cardonae), (ii) three shrubby species in Hispaniola (52 H. ekmanii, 53 H. pycnophyllum, and 54 H. millefolium), and (iii) the derivatives of 61 H. cymobrathys, which occur throughout Central America and the Greater Antilles and extend their area of distribution into eastern U.S.A. The three Hispaniolan shrubs terminate a clade that originates with two of the Costa Rican species, which also occur in the basic area of the Colombian- Venezuelan border (49 H. jaramilloi, 50 H. cardonae) (Fig. 3, clade xxviii). Occupying an intermediate position is 51 H. caracasanum of northern Venezuela, of which the more advanced subspecies (51b subsp. turmiquirense) is found at the north-eastern extreme of the Andean range and is also morphologically close to 52 H. ekmanii. These data suggest that the ancestor of the Hispaniolan species reached that island either directly by long-distance dispersal or via the Lesser Antilles, which, according to the above mentioned theory, would by then have arisen as a result of volcanic action; but overland access before the separation of the Greater Antilles from South America cannot be ruled out. A three-fold origin of Hispaniola from parts previously associated respectively with Yucatan, eastern Cuba, and the Bahamas, such as has been proposed by Rosen (1985), would make the overland route less likely (see also p. 91). The immediate derivatives of 61 H. cymobrathys are three species of the Colombian and Venezuelan Andes (see Part 8). Of these the Colombian H. chamaemyrtus is the sister-species of the H. gnidioides — H. gentianoides group, which radiates from the Honduras Republic: south to Panama (H. gnidioides), north to north-eastern U.S.A. (H. gentianoides), and east to the Greater Antilles (H. diosmoides group). On the other hand, there is a wide geographical disjunction between the Venezuelan H. pseudocaracasanum and its sister-group in the south- eastern U.S.A. (H. denticulatum et aff.). (b) Species with leaves persistent or deciduous near the base (Spp. 3-23) In this group (Fig. 3 clade (c)) the leaves are rarely broadest above the middle, and so the leaf-venation is parallel to pinnate or 1-nerved, not flabellate. All species have leaves that are either wholly persistent or deciduous at the apparent base. They form two subgroups: clade (1) with leaves that are free (Spp. 3-9); clade (ii) with leaves that are united by an interfoliar ridge or rarely more intimately (Spp. 10-23). Subgroup (i) radiates from 3 H. phellos of north-eastern Colombia and adjacent Venezuela. The variation in this species can be described; but it has not been possible to express this variation in a formal classification, apart from the recognition of two derivative subspecies. Subspecies H. phellos has therefore had to be differentiated plesiomorphically (i.e. by the possession of primitive characters), viz. by having plane or recurved (not incurved or revolute) leaves that are deciduous at the base of the free part (see p. 22). Four separate derivative clades (iv, vi, vii, viii) end with taxa in which the leaves are secondarily wholly persistent, and in three of these (all but vii), the leaves are plane or incurved. The taxa in these three are all distributed in or near the Colombia-Venezuela border area near H. phellos except 7 H. irazuense (Costa Rica and adjacent Panama), which is confined to the nearest high ground in Central America to the Colombian Cordillera and is apparently derived from H. phellos subsp. oroqueanum (César/Norte de Santandér border). In clade (vii) the tendency for the margin of coriaceous leaves to become reflexed (3a iii H. phellos subsp. phellos ‘patens’) is exaggerated, so that in 4 H. garciae it is revolute, hiding most of the lower lamina. In this species, which has a disjunct distribution in the central Colombia/ western Venezuela area, the leaves tend to become narrow and sometimes longer, a tendency that is more evident in 5 H. acostanum, a local species of southern Ecuador that is smaller and 10 NORMAN K. B. ROBSON less woody than H. garciae. Although isolated from the nearest H. garciae population by a wide disjunction (central Colombia to southern Ecuador), H. acostanum is separated from the species that terminates clade (vii) by an even wider one. 6 H. piriai is a woody herb of south-eastern Brazil and Uruguay in which the flowers have apparently become adapted to specialized insect-pollination. Thus the stamen fascicles have the filaments united for more than half way (asin, say, H. elodes), and two pairs of fascicles are more or less completely united. The petals are oblique to erect, sometimes making the flower pseudo-tubular, again as in H. elodes. How the gap between southern Ecuador and south-eastern Brazil was traversed is not clear; quite possibly it was by ancient long-distance dispersal. I have not yet discovered any similar disjunctions, but the morphocline southward along the Andes to Ecuador suggests that a vicariant interpretation is unlikely to be correct. An early offshoot of clade (ii) comprises two isolated species (clade ix) in which the relatively large herbaceous leaves fade and droop while on the stem. 10 H. simonsii is confined to the northern part of the Sierra de Santa Marta (Magdalena), whilst 11 H. papillosum has a restricted distribution in Boyaca. In the remainder of the clade (ii) the leaves are smaller and either broad-based and deflexed (clade xi) or narrow-based and spreading only (clade xii). At the extremes of clade (xi) the leaves become too narrow to droop and so remain spreading (15 H. myricariifolium, 18 H. loxense in part). The species at the base of these clades (13 H. thuyoides) has a vicariad that is scarcely more than subspecifically distinct (12 H. lycopodioides), in which the leaves have become coriaceous. H. thuyoides itself, from central Colombia (Cundinamarca), is variable, with the broader-leaved (typical) form near clade (xi) and the narrower-leaved form (‘genistoides’) at the base of clade (xii). Clade (xii), in turn, divides into two distinct subclades: in one (xiv), the areas of 22 H. hartwegii and 23 H. maguire? represent a major disjunction to southern Ecuador, whereas in the other (xiii), H. thuyoides is near the origin of two trends initially in opposite directions, respectively northwards to Boyaca (19 H. sabiniforme) and southward to northern Ecuador (20 H. laricifolium). As has already been mentioned, H. laricifolium shows continuous variation southwards to central Peru and northwards to western Venezuela; but the morphologically reduced population further north, in the Sierra Nevada de Santa Marta (Magdalena) is sufficiently distinct to warrant specific recognition (21 H. martense). (c) Species with leaves with sheathing leaf-bases or marginal secretions (Spp. 24-26, 61) In contrast to the early offshoot (f) in Boyaca in which the oblong to elliptic leaves of the basic species have parallel venation (61 H. cymobrathys), nearly all early members of clade (d) (Fig. 3) have leaves broader above the middle with flabellate venation. A partial exception is 24 H. magniflorum, in which the Boyaca population has parallel-veined leaves, but the more northern one (straddling the Colombia- Venezuela border) has flabellate venation. H. magniflorum is also exceptional in that the leaves are relatively thin and droop when fading, instead of breaking off above the base like those of all the other species. It forms the base of one sub-branch of Branch (e) (clade xv), in which the broad, flabellate-veined leaves have sticky marginal leaf secretions, capitate stigmas, and usually a cymose inflorescence. 25 H. gleasonii occurs in the border area of Colombia/Venezuela, whereas 26 H. mexicanum has its primitive form in Boyaca and Cundina- marca and differentiated back northwards into the area of H. gleasonii. (d) Species with incurved-cucullate leaves but no marginal secretions (Spp. 27-30) Apart from those species with marcescent or secreting leaves in clade (xv) there remains clade (e), which splits into two subclades, (xvi) with leaves cucullate at the apex and (xvii) in which they have a plane apex (apart from a few of the more specialized species where the apex is somewhat cucullate). The basic species of clade (xvi), 27 H. stuebelii, is a northern Peruvian relict, the area of which is thus separated from that of its nearest relatives in north-central Colombia by a considerable disjunction. All the other members of this group are confined to the south Ecuador/north Peru region except 30 H. decandrum, which has extended its area into north Ecuador. H. stuebelii has leaves broader than those of H. magniflorum, with flabellate venation; but the leaves of the other species (Spp. 28-30) are smaller and narrower with only a midrib or, at most, one pair of THE GENUS HYPERICUM L. 11 vein branches. Like H. stuebelii, 28 H. prietoi and 29 H. cassiopiforme appear to be relicts, but 30 H. decandrum is more widespread. (e) Species with sessile leaves deciduous above the base but not incurved-cucullate (Spp. 31p.p.—48). The primitive form of 31 H. pimeleoides, which is found in the Cerro de Oroque (César/Norte de Santandér), has large, broad, sessile, plane leaves; but within the range of this species are forms where the leaf base has narrowed to a pseudopetiole. The remaining major woody clades in sect. Brathys are respectively characterized by these two leaf forms, the species of clade (xviii) (Spp. 32-48) having sessile leaves, whereas in all but some extreme species of clade (xix) they are pseudopetiolate. In clade (xviii) the basic species, 32 H. magdalenicum, is almost confined to the isolated northern Sierra de Santa Marta (Magdalena) and the adjacent Sierra de Perija on the border with Venezuela. Apart from 43 H. parallelum, a close relative with smaller, densely imbricate leaves isolated on the Cerro de Oroque, this variable species has given rise to two clades, the members of which differ by a complex of characters. Clade (xx), which derives from the more primitive, broader-leaved form of the species, includes species with leaves that twist as they spread. This character is evident only where (a) the leaves are long enough to twist at all and (b) the internodes are not too short to preclude individual twisting. In two extreme species, 35 H. aciculare and 42 H. prostratum, the leaves are often too short to twist, while in 36 H. recurvum and 37 H. wurdackii the internodes are too short to allow it. The two subsidiary clades of clade (xx) each has a relict basic species, both of which occur at a considerable distance to the south of the H. magdalenicum area. In clade (xxii), 33 H. valleanum (Colombia: Valle de Cauca) forms a stepping-stone to the species of clade (xxiv), which are all from Ecuador and Peru. Of these, 34 H. sprucei occurs from northern Ecuador to northern Peru, with the distribution of the derivative 35 H. aciculare overlapping its area in south Ecuador and just extending into Peru. The species with recurving leaves (36 H. recurvum, 37 H. wurdackii) are in north and central Peru, H. wurdackii being a bizarre relict with densely imbricate leaves forming four ‘wings’. Clade (xxv) proceeds along the Cordillera from H. valleanum in the opposite direction to that of clade (xxiv). A small population of 38 H. costaricense is found on the Bolivar/Antioquia border as well as the main one in Costa Rica and adjacent Panama, there being no suitably high ground between them. 39 H. bryoides, clearly a reduced version of H. costaricense, is confined to the South American Cordillera (Norte de Santandér to Boyaca). In the relict basic species of clade (xxiii), 40 H. bolivaricum (Colombia: Bolivar), twisting leaves are combined with a condensed synflorescence and capitate stigmas, all characters that are found in the variable 41 H. juniperinum (= H. brathys). In the latter species, which is distributed from western Venezuela (Trujillo) to southern Colombia (Putumayo), there is a trend towards an increase in the number of styles and placentae from 3 to 4 and then to 5. The population from Cauca shows another trend, among 3—4-styled forms, from tall erect shrubs to smaller shrubs with decumbent stems. This trend is continued back north-eastward in 42 H. prostratum, a dwarf shrub or shrublet from the northern half of the Colombian Cordillera Oriental with stems decumbent to prostrate. The remaining derivatives of 32 H. magdalenicum are all related to more advanced forms of that species (32’ in Fig. 3). Apart from 43 H. parallelum they form two clades: (xxvi) of which the members have polished, acicular leaves, rounded-incurved in section and twisting when large enough, and (xxvii) in which the species have leaves that are dull to metallic and flat to incurved but not usually rounded in section and become recurved not twisted. Clade (xxvi) comprises three geographically separate taxa that I have treated as subspecies of 44 H. marahuacanum. The subspecies with the largest flowers (44a subsp. marahuacanum) is confined to the Cerro Marahuaca in southern Venezuela (Amazonas), an isolated massif from which no other Hypericum species are known; and one of the derivative subspecies, 44c subsp. chiman- taicum, is equally isolated further east, on the Chimanta Massif and Cima de Roraima (Bolivar) near the Guyana border. The remaining derivative subspecies, 44b subsp. strictissimum, occurs 12 NORMAN K. B. ROBSON in two quite separate areas (Venezuela: Mérida and Colombia: Boyaca) and differs from subsp. marahuacanum in size and condensation of parts. In comparison with H. marahuacanum, the species of the H. strictum group (Spp. 45—48) are south-western in distribution, being confined to the area from Cundinamarca and Meta (Colombia) to western Venezuela. The relatively broad-leaved 45 H. lancifolium, from the Colombia-Venezuela border area, has a dwarf, high-altitude derivative, 46 H. horizontale. on the Colombian side of the frontier and a southern derivative complex (47 H. tetrastichum, 48 H. strictum) with H. tetrastichum extending from Boyaca northward into Venezuela and H. strictum occurring farther south. (f) Species with pseudopetiolate leaves deciduous above the base but not incurved-cucullate (Spp. 31 p.p., 49-60) The above heading is not strictly accurate in that a few of the advanced members of this group (clade xix) have developed a somewhat cucullate apex. But their leaves are linear and would not be confused with those of species in clade (xvi). The species with pseudopetiolate leaves and their derivatives in clade (xix) (Spp. 49-60) would appear to have evolved relatively early in the history of sect. Brathys, as they have spread to almost the furthest limits of the shrubby part of the section (Hispaniola and Costa Rica to Bolivia). The basic species of this clade, 49 H. jaramilloi, has a disjunct distribution in northern Colombia (Cerro de Oroque, Sierra de Santa Marta) and Costa Rica; and an early species in the eastern subclade (xxviii), 50 H. cardonae, has a similar but wider distribution, in Costa Rica and the Colombia/Venezuela border region south to Boyaca. The rest of the species of this clade show a geographical replacement pattern eastward: 51 H. caracasanum in the mountains of northern Venezuela, with 51b subsp. turumiquirense at their extremity in Sucre, and the remainder (Spp. 52-54) in Hispaniola (see p. 9). The second (westward) derivative clade of 49 H. jaramillo, clade (xxix), is based on 55 H. ruscoides, from south Colombia and north Ecuador, which gave rise southward to clade (xxx), a series of sessile-leaved species (56 H. llanganaticum in central Ecuador, 57 H. struthiolifolium in Peru, and 58 H. andinum in south Peru and Bolivia) with acute leaves and (at least initially) long styles with relatively narrow stigmas. In another direction, evolution resulted in two species with obtuse, cucullate leaves and short styles with relatively broad stigmas. Of these, 59 H. lancioides comprises a southern subspecies with a pseudo-dichotomous synflorescence (59a subsp. /an- cioides) and a northern one in which the flowering branches are lateral and congested (59b subsp. congestiflorum). A final southward development from the latter has resulted in the dwarf 60 H. selaginella, with prostrate or ascending stems and 3-5 styles and placentae. 2. Systematic treatment Sect. 29. BRATHYS (Mutis ex L.F.) Choisy Prodr. Monogr. Fam. Hypéric.: 38, 58 (1821).* Small trees, shrubs, shrublets or wiry herbs up to 6 m tall, evergreen, glabrous or very rarely with simple hairs, without dark glands; branching below inflorescence lateral. Stems 4(6)-lined and + compressed (ancipitous) when young, sooner or later becoming terete, eglandular; cortex exfoliating in strips or scales or irregularly; bark fissured, smooth. Leaves opposite, decussate, sessile to shortly petiolate, free or + deeply united to form interfoliar ridge or very rarely bases of lamina proper united, base sometimes sheathing, deciduous at base or (more usually) above base or persistent and (in woody species) either fading or shed with cortex; lamina entire, with venation parallel or flabellate to pinnate or 1-veined, open or closed, with tertiary venation + laxly reticulate; laminar glands punctiform; marginal gland dots dense, often very small; ventral * Including Sarothra L. (1753), Martia sensu Sprengel (1826) pro parte, Hypericum sect. Brathys subsect. Multista- mineum R. Keller (1893). Hypericum sect. Brathys subsect. Spachium R. Keller (1893), Sanidophyllum Small (1924), Sarothra sect. Spachium (R. Keller) Y. Kimura (1951), Sarothra sect. Spachium series Eusarothra Y. Kimura (1951), see Part 1 (Robson, 1977: 338); also Hypericum sect. Sarothra (L.) Reichardt in Martius, Fl. Brasil. 12(1): 186 (1878). THE GENUS HYPERICUM L. 13 resin glands absent. Inflorescence 1-flowered, with branching pseudo-dichotomous (usually from uppermost node) or 2—15-flowered and then with branching (within inflorescence) dichasial/monochasial from uppermost node, often with subsidiary branches from lower nodes; bracts and bracteoles foliar or reduced or sometimes transitional in form to sepals. Flowers stellate or occasionally obconic, homostylous. Sepals 5, free, persistent, erect in fruit, with margin entire; veins 3—11(—19); laminar glands linear to punctiform; marginal, submarginal and inframarginal glands absent. Petals 5, persistent, with apiculus subterminal to obsolete, acute to acuminate or obtuse; margin entire; marginal glands absent; laminar glands linear to punc- tiform. Stamen fascicles basically 5, very rarely distinguishable individually, usually forming a continuous narrow ring of 5—250 stamens (i.e. 1-50 per fascicle) or very rarely fascicles + united 2 + 2 + 1, persistent; filaments very shortly united or very rarely united above midway; anthers yellow to orange, gland amber; pollen type VIII. Ovary with 3-S parietal placentae, ©-ovulate; styles 3-5, free with bases contiguous; stigmas small to broadly capitate or clavate. Capsule 3—5-valved, coriaceous to chartaceous, with valves obscurely narrowly longitudinally vittate, occasionally with 1-3 flattish vesicles. Seeds narrowly cylindric, shallowly carinate or not, without apical expansion; testa scalariform-reticulate to ribbed-scalariform. BASIC CHROMOSOME NUMBER (x): 12; ploidy 1, 2, 4. Hasirat: open grassland or among stones or sometimes in scrub or moist montane woodland, often along streams, in dry to locally wet areas of the Cordilleran paramo and subpéramo vegetation; 1600—4875 m. Also in open Pinus or Pinus—Quercus woodland or Pinus savannah or on sterile white sand (Spp. 1 and 2, in Belize and Cuba, 0-800 m); or in open Pinus forest or in cloud forest among limestone rocks (Spp. 52-54 in Hispaniola); 1300-3175 m. See part 8 for herbaceous species. DisTRIBUTION: Belize, Cuba, Hispaniola; Costa Rica, Panama; Guyana (Roraima), Venezuela (Roraima, Chimantaé, Marahuaca) and Andes of Venezuela, Colombia, Ecuador, Peru, and Bolivia: also (subshrubs and herbaceous species) Porto Rico, Honduras, Guatemala, Mexico, and south-eastern and eastern U.S.A. 80 species (+ 8 subspecies). Key to sect. 29. Brathys 1 Leaves deciduous at base, leaving rhombic scar on stem; stem not articulate or with INTETIONUARTIOBES oonicc re cednce poeta tes ety Conn pubenatetels drat seebe wv antedee sewtidetepmbunnceNs 2 Leaves persistent (falling with cortex) or deciduous above base; stem articulate and/or with INCETCOMAE TIQGES ower os adeseacan ei ee wa adultes acco s ec uues ra csckb cpuv'uwede nue boos hanes naneadE os a 2(1) Leaves narrowly elliptic, acute, 16-30 mm long, chartaceous to subcoriaceous; sepals PDD WAIN ONG oh con vcs. © baie cee Sodus cuca eRieeh ep dean RRM eRCe get 1. terrae-firmae(p. 18) Leaves narrowly obovate or, if oblong to elliptic, then acuminate or shorter (10-16 mm long) or thickly coriaceous; sepals 5—15 mm long (2. styphelioides)......................:000605 3 3(2) Leaves narrowly elliptic to narrowly oblong, usually thinly coriaceous, spreading, often OCCUNVIIE Gre ave lane tet ou canctyeldraknce cote aoaanaeaes 2a. styphelioides subsp. clarense(p. 21) Leaves oblanceolate to narrowly obovate, + thickly coriaceous, + densely imbricate, SUL Ne ae vac ilar Pan eg sues sete snes He ek a TN Puy sur da vautws inverters cornea ths hewteu kage ated 4 4(3) | Leaves with 7-11 basal veins, glands not prominent 2b. styphelioides subsp. styphelioides(p. 21) Leaves with 5—7 basal veins, glands prominent......... 2c. styphelioides subsp. moaense(p. 22) 5(1) Leaves persistent or deciduous immediately above base of free part; pseudo-dichotomous TANCES OLE AUSENU OL SAIS ricer ennai enresccawe esheets eam cusesemnpenterstesanrzo steed 6 Leaves deciduous above base, i.e. leaving base protruding from stem; pseudo- chehotomous branches O1ten PregOMINANl 6.24565 365 ser. ou,se oases cess aiwnceeedn esses sn punssoscevene 38 6(5) Leaf bases either (i) sheathing or (ii) amplexicaul with stem fragile and articulated, the Hmermodes and leaves falling 25 ONG 6 ...o< lociccssccnts ae saasducsGGarcacvesserietec ioe noudeeesiseennanee 7 Leaf bases free to perfoliate or shallowly cupped; internodes and leaves not as above ........ 8 14 7(6) 8(6) 9(8) 10(9) 11(10) 12(11) 13(10) 14(13) 15(14) 16(15) 17(16) 18(15) 19(13) 20(19) 21(20) 22(8) 23(22) NORMAN K. B. ROBSON Leaf bases sheathing; leaves 10-18 mm long, withering; flowers 25—40 mm in diam. 24. magniflorum(p. Leaf bases amplexicaul; leaves up to 2 mm long, falling with internodes; flowers 5—7 mm in GUDARR Da siete cas otcae soos horace eectea conc Stay ai Sele ei-ea tan center war eepeaaostaars 54. millefolium(p. Leaves free, the base sometimes indurated or fused with the stem .................ccececeeeeeeee es Leaves united by interfoliar ridge or lamina prope? ........0...:i::0sc0esssesseeconennntnarssoroesees Leaves becoming deflexed, marcescent; stem internodes without or almost without corky CMEFQENCES 5.5 hone cwssldge eh sseapedealantaeees 3a. phellos subsp. phellos (i) ‘marcescens’ (p. Leaves becoming deflexed to appressed but not marcescent; stem internodes with or WItHOULCOLKY CMNOrBENICES co 2 ys voc2 bcc uy cp avon apy recone a4 sien on open pte guneane face “bees naseeet scene Leaves revolute, lower surface obscured, bases forming + woody projections; stem PLETTIOGES SINGER. acres eG dae pene saske 0 dacivae cc adeeb Genes Ceslerneurs as tesewasesesacioneveseserwcryss Leaves recurved to incurved, lower surface always visible; stem internodes usually with CORKY CMICTOCTICER SG 55.5 non easel esstecerene Wen ducen t eege we vaste ae Ses Paw tense are sete es etak comune: Leaves deciduous, 4-10 mm long, elliptic-oblong to linear; sepals ovate to lanceolate; SHU EO OEY tal nary etek Nera AE asides rad Aoanrap nat ae heehee he ounnaas nian aeeate ane 4. garciae(p. Leaves persistent, (8)10—40 mm long, linear; sepals narrowly lanceolate to linear or oblong; dwarf shrub to 0-8 m tall or suffrutex Or Berd 2.556. 265035cc0ssessccesnrenssecesersusso nce Flowers solitary, inflorescence-branching pseudo-dichotomous; flowers stellate; stamens not fasciculate;.dwart Shrub: &..,.:..4¢ 14. 5.5135 as ed jacaesagsssateessades sscseshwmes 5. acostanum(p. Flowers 1-25, inflorescence-branching monochasial; flowers obconic to pseudotubular stamens 3(5)-fasciculate; suffrutex or perennial herb ...................:seeeeeeeees 6. piriai(p. LOAVES With :o bdSal VEINS. SUDCOFIACEOUS sass ca25 235 ory tse ou Buy ndansak onensscaseneneateatatsenagertnanes Leaves with 1 basal vein (venation wholly pinnate), chartaceOuS................scceeseeeereeeeeees Leaves recurved before falling ................... 3a. phellos subsp. phellos (ii) ‘angustum’ (p. Leaves spreading to appressed before fallin. jive vcrce ce ciese ts: avec Waves rises don sentese soooxeene ts PG aVies CEGIGMOUS cet caetias cas sas ths saa Aba BE cau esniedaivaldcaeeleeecatenees nneseae cares eeme seas PE GAVGS POISISTO BG sce co nattars Adi oks dace sutsnicatms candugealeaupdyerecleben tirchesemaesnub asa ueaoenn opts Leaves spreading, not or scarcely becoming appressed, elliptic to oblanceolate 3a. phellos subsp. phellos (iii) ‘patens’ (p. Leaves becoming or remaining appressed, oblanceolate to narrowly oblong................++++ Leaves spreading, then appressed, before falling 3a. phellos subsp. phellos (iv) ‘phellos’ (p. Leaves remaining appressed, tetrastichous ............... 3b. phellos subsp. oroqueanum(p. Leaves with 3 unbranched basal veins; flowers 25-30 mm in diam.; styles 9-10 mm long, BOW AS ONG AS OVALY ces coun se vaTIAG Coates Catena 35. aciculare(p. 66) Leaves imbricate at first but not markedly tetrastichous, recurving, 5-8 mm long; flowers 152 min in diam.; petals $=12 min long 7.32, .0250. cc cceun-esecevs onneo 39. recurvum(p. 68) Leaves densely and persistently imbricate, ‘winged’-tetrastichous, scarcely spreading, 9-14 mm long; flowers 20—25 mm in diam.; petals 15-18 mm long....... 40. wurdackii(p. 69) Flowers 15—20 mm in diam.; leaves 5—15 mm long, acute; stems woody, not rooting 38. costaricense(p. 70) 18 70(63) 71(70) 72(62) 73(72) 74(73) 75(72) 76(75) 77(76) 78(77) NORMAN K. B. ROBSON Flowers 6—8 mm in diam.; leaves 3-5 mm long, subacute to obtuse; stems wiry, basally POORIELG os coarse oh ceo sa eee ea eee cat oy eras eta teres Fneeonniaee Fare ans 39. bryoides(p. 71) Leaves very narrowly elliptic, 15 x 3mm or larger; sepals 9-11 mm long, coriaceous; styles CoS IMLIOUS Sees seach ete ew ovce dn ens ee oy aan Yo caine ope nce Reiccsiog tees 40. bolivaricum(p. 72) Leaves linear, up to 14 x 1-5 mm; sepals 2—7 mm long, chartaceous to membranous; styles Eo hs 1110 (6) 1s Jan es ae Aneta af SPR Seances Sarre M AAP REE PU im ry re Me Envi dion tame man Sr L v1 Stems erect to decumbent; leaves 6-14 mm long; styles 1-5-3 mm long, longer than ovary 41. juniperinum(p. 72) Stems prostrate; leaves 2-5 mm long; styles 0-6—0-8 mm long, shorter than ovary 42. prostratum(p. 75) Leaves lustrous, margin markedly incurved, lamina twisting and sometimes recurving (44. Pit Tg Li) te 111117 /) NASSP OSSSREREED SEE TT Grr tte REE Re CHT re ee ee roe E Pre etd dee BU sn ces v8) Leaves dull or with metallic sheen, margin plane to incurved, lamina recurving only.......... 75 Stigmas small or slightly enlarged; ovary and capsule acute; inflorescence branching Pscudo-dichotomolus OF TACCMMOTA can pat con dves sopnaete enone cece scwcunet keek Tix neeasaaineeee 74 Stigmas enlarged to capitate; ovary and capsule + rostrate, inflorescence branching pseudo-dichotomous to lateral-congested 44c. marahuacanum subsp. chimantaicum(p. 80) Styles 4-5 mm long; sepals 7-9 mm long; inflorescence branching pseudo-dichotomous 44a. marahuacanum subsp. marahuacanum(p. 79) Styles 2-3(—4) mm long; sepals 4-6 mm long; inflorescence branching lateral, + racemi- | (ay 911 War rare OM SY AP oe Taree eee er ae! 44b. marahuacanum subsp. strictissimum(p. 79) Leaves plane or slightly incurved, oblong to oblong-linear or narrowly oblanceolate, the tips NOtTEG WHEN YOUNG saree. re... race. 255. et eaaee sca Fa suanh Gen ¥ veer sores 45. lancifolium(p. 80) Leaves markedly incurved to canaliculate, linear to linear-acicular, the tips usually red Wile YOUN Bais sont eee epee te sates ea ented er ere teads ve eRaene ss deca eek tec netes Rogrolmiweah rr ceeih 76 Plant a prostrate shrublet with branches matted, rooting; leaves often secund, not POET ASTICIOUS SS oe tirs eee ies eed sh ope ewed Came iesee ach oe estes 46. horizontale(p. 81) Plant an erect shrub with branches strict, not rooting; leaves never secund, usually + CEEEASTICHO US OC Rees ata eis de sages ees de Teen ee anaes poet tins venvaticts Muviseds 77 Leaves with metallic sheen, markedly tetrastichous, erecttosuberect 47. tetrastichum(p. 82) Leaves dull, not tetrastichous (except sometimes in young parts or plants); usually MULCUE VANS (ORs SeriChaIn As cat yy cote ee ees a sweeeone a artes sche Rn eee aes cates Codpaeea eee 78 Leaf and sepal apices acicular; sepals 1-3—2 mm wide, not ribbed; leaves 8-13 mm long; branches strict, usually all from one stem .................. 48a. strictum subsp. strictum(p. 84) Leaf and sepal apices subacute; sepals 1-1-2 mm wide, clearly ribbed; leaves S—7(8) mm long; branches ascending, usually branched from the base 48b. strictum subsp. compactum(p. 85) 1. Hypericum terrae-firmae Sprague & Riley 321 (K!, holotype; GH!, isotype). Adams in Rhodora 64: 236 (1962), non A. Richard (1845). Icon: Fig. 4. in Kew Bull. 1924: 12 (1924). Type: Belize, without precise locality, 1912? (fr), Peck H. styphelioides sensu Standley & Williams in Fieldiana Bot. 24 (7): 51 (1961); P. Shrub or small tree, 1-2 m tall, erect, with branches strict, pseudo-dichotomous or lateral. Stems orange-brown, 4-lined when young, soon terete, without corky wrinkles, cortex exfoliating in strips; internodes 4—6 mm long. Leaves sessile, free from the base, spreading to subimbricate and tetrastichous, deciduous at the base without fading; lamina 16-30 x 4-6 mm, narrowly oblong to narrowly elliptic, plane, not cucullate or carinate, concolorous, not or slightly glaucous, chartaceous to thinly coriaceous; apex acute, base narrowly cuneate to subangustate, not sheathing, * The leaves and petals in the marginal text figures are all respectively twice and one-and-a-half times natural size. In the inflorescence diagrams O = open flower, @ = flower bud, x = vegetative bud, 7 = continuing shoot. Note. In part 3 (Robson, 1985) the magnifications in the legends to figs 9-24 should all have been reduced by half. THE GENUS HYPERICUM L. by ae SSN / N Lf ZL, > : yi . N \ rN Aix Ll N N = 19 Fig. 4 H. terrae-firmae: (a) habit; (b) stem with leaf bases; (c) leaf scar; (d) leaf; (e) sepal; (f) petal; (g) stamens (partly cut away) and ovary; (h) stamen; (i) anthers; (j) capsule (a, b = 2; c X 1; d,j x 2; e-g x 3;h x 8;i x 15). (a—d, g, j) Hunt 66; (e, f, h, i) Sutton et al. 193. pairs free; basal veins 7, subparallel-sided, with short lateral branches, tertiary reticulation not visible; laminar glands dense, not prominent. Inflorescence 1- flowered, with pseudo-dichotomous branches from 1—2 nodes below; pedicel absent or very short; upper leaves transitional. Flowers 40—60 mm in diam., stellate. Sepals 14-22 x 3-7 mm, narrowly oblong, acute; veins 15, dichotomising and reticulate distally, with midrib prominent; glands linear. Petals bright yellow, 20-30 x 7-12 mm, c. 1:3 X sepals, oblanceolate; apiculus acute; glands linear, mostly uninter- rupted. Stamens c. 250, longest 9-11 mm long, c. 0-3—-0-4 x petals. Ovary (2-5—)3 20 NORMAN K. B. ROBSON —4-5 x 2-2-5 mm, + narrowly ovoid; styles 5, 5—7 mm long, 1-4-3 x ovary, erect, outcurved below apex; stigmas small. Capsule 8-10 x 5-7 mm, broadly ovoid, shorter than sepals. Seeds c. 1 mm long, not or shallowly carinate; testa finely scalariform-reticulate. In open pine or oak-pine forest or pine savannah on granite, often near streams; lowland to 550 m. Belize (Cayo, Belize). Map 1. BELIZE. Belize: Colonel English Pine Ridge, Belize-Cayo road, 20.1.1958 (fl), Gentle 9515 (DUKE, F, NY, S, US); Burrellboom, c. 9 m, 22.11.1970 (fr), McDaniel 12853 (F). Cayo: Mountain Pine Ridge, San Miguel, 400 m, 2.vi.1979 (fl & fr), Whitefoord 1935 (BM); Mountain Pine Ridge, Augustine, vii—viii. 1936 (fl), Lundell 6617 (F, K, MICH, NY, US). H. terrae-firmae, as previous authors have pointed out, is closely related to the Cuban H. styphelioides, especially to forms of subsp. clarense; but the thinner, longer, acute leaves, which spread more or less widely from the base, and the large flowers with longer, narrowly oblong sepals, allow the Belize plant to be accorded specific rank. Its nearest ancestral taxon is in East Africa (H. revolutum subsp. keniense (Schweinfurth) N. Robson), a subspecies of Sect. 1 Campylosporus. 2. Hypericum styphelioides A. Richard in Sagra, Hist. Phys., Pol. Nat. Ile Cuba (Bot. — Pl. Vasc.): 237 (1845); Leon & Alain, Fl. Cuba 3: 316, t. 141 (1953); Lippoldin Wiss Z. Friedr.-Schiller Univ. Jena (Math.-Nat. R.) 19: 378, f.2. (1970). Type: Cuba, [Pinar del Rio, near San Diego de los Banos], Sagras.n. (P, holotype). Shrub 0-15—2 m tall, erect, with branches erect, strict, pseudo-dichotomous or lateral. Stems orange-brown, 4-lined when young, soon terete, without epidermal wrinkles, cortex exfoliating in strips or scales; internodes 2-6 mm long. Leaves sessile, spreading from above the base to closely imbricate and tetrastichous, deciduous at the base without fading; lamina 5—25 x 2—8 mm, narrowly obovate to narrowly elliptic or narrowly oblong, plane or incurved, sometimes + concave, not carinate, concolorous, lightly to densely glaucous, thinly to thickly coriaceous; apex shortly acuminate to acute, base cuneate to scarcely angustate, not or slightly sheathing, pairs free; basal veins 5—9(—11), subparallel to flabellate, branching distally to form a lax reticulum, tertiary reticulum not visible; laminar glands dense, + prominent. Inflorescence 1(2)-flowered, with pseudo-dichotomous branches from 1-2 nodes below, sometimes with additional branches from immediately lower node; pedicel absent or very short; upper leaves transitional. Flowers 15—S0(—70) mm in diam., stellate. Sepals 5-15 x 2-—S mm, oblong or lanceolate to obovate, acute or obtuse-apiculate to obtuse; veins 5—17, dichotomising, and reticulating distally, with midrib prominent beneath; glands linear, sometimes distally punctiform. Petals bright yellow, (8—)10—22 x 2-10 mm, c. 1-5 X sepals, oblanceolate; apiculus acute; glands linear, uninterrupted. Stamens c. 70—200, longest 5-9 mm long, c. 0-3—0-5 x petals. Ovary 2-4 x 1-5-3 mm, ovoid to rostrate-subglobose; styles (4)5, 3-5 mm long, 1-2—1-5 X ovary, erect, outcurved below apex; stigmas small. Capsule (3—)4-8 x 2-7 mm, ovoid-subglobose, shorter than sepals. Seeds c. 1 mm long, not or shallowly carinate; testa finely scalariform-reticulate. In pine forest or pine savannah or often on sterile white sands, especially round margins of lakes and pools; 0-800 m. Cuba (Pinar del Rio, Las Villas, Oriente, Isla de Pinos). Map 1. Three subspecies of H. styphelioides have been recognized. Subsp. clarense, the most similar to H. terrae-firmae, has a disjunct distribution (central Cuba, western Cuba, Isla de Pinos), whilst the others are respectively eastern (subsp. moaense) and western (subsp. styphelioides). THE GENUS HYPERICUM L. aN vA Map 11. H. terrae-firmae @.2. H. styphelioides: a. subsp. clarense A, b. subsp. styphelioides O, c. subsp. moaense @. 2a. Hypericum styphelioides subsp. clarense Lippold in Wiss. Z. Friedr.-Schiller Univ. Jena (Math.-Nat. R.) 19: 379, f. 2. (1970). Type: Cuba, Las Villas, Santa Spiritus, Rio Gayabo, 7.xii.1911 (fl), Bro. Clemente 2884 (LS, holotype; NY!, isotype). Hypericum incurvum Urban, Symb. Anitill. 9: 405 (1925); Le6n & Alain, Fl. Cuba 3: 317 (1953); Lippold in Wiss. Z. Friedr.-Schiller Univ. Jena (Math.-Nat. R.) 19: 380, f. 2. (1970). Type: Cuba, Isla de Pinos, Santa Barbara, Westport, 3.xii.1920 (fl), Ekman 12072 (S!, holotype). Shrub 0-15—0-6(-1) m tall. Leaves 10-18 x 2-5-5 mm, narrowly elliptic to narrowly oblong, thinly to thickly coriaceous, not or lightly glaucous, plane to concave or incurved, spreading or outcurving; apex acute; basal veins 5(7). Flowers 15-30 mm in diam. Sepals oblong to lanceolate, acute or obtuse, 5-veined; glands prominent. Petals 7-16 mm long. Styles 5, 3—5 mm long. Savannahs, fields, and coastal sands, sometimes on weathered dolomite; lowland to montane. In Las Villas, western Pinar del Rio and Isla de Pinos. CUBA. Pinar del Rio: N. of San Diego de los Banos, 13.iv.1900 (st), Palmer & Riley 535 (NY, US); mountains near E] Guama, 6.iii.1900 (fr), Palmer & Riley 160 (NY, US). Las Villas: Lomas de Banao, i.1920 (fl), Lund 123 (NY); Trinidad Mts, Buenos Aires, near La Sabana, 5.viii.1936, Smith, Hodgson & Gonzales 3381 (F, NY, S, US). Isla de Pinos: Siguanea, 26.ii1.1916 (fl), N. L. Britton, E. G. Britton & Wilson 14926 (F, NY, S, US); between Mine de Oro and Playa del Soldado, 12.1.1956, Killip 45389 (US). Subsp. clarense forms a cline from eastern Las Villas (where the flowers are larger and the leaves are broader and flat) via eastern Pinar del Rio to Isla de Pinos (where the flowers are smaller and the leaves narrow with an incurved margin and out- curving towards the apex). 2b. Hypericum styphelioides A. Richard subsp. styphelioides Shrub 0-3-2 m tall. Leaves 8-25 x 4-8 mm, oblanceolate to narrowly obovate, plane, + densely glaucous, + thickly coriaceous, spreading to densely imbricate; apex shortly acuminate; basal veins 7—9(—11); glands not usually prominent. Flowers 15—50(—70) mm in diam. Sepals elliptic to obovate, acute or obtuse-apiculate to obtuse, 9-17-veined. Petals 10-22 mm long. Styles 5, 3-5 mm long. 22 NORMAN K. B. ROBSON In pine forest or savannah sand or in wet places; lowland. In Isla de Pinos, Las Villas and Pinar del Rio. CUBA. Pinar del Rio: El Sabalo, Finca Sabanalamar, near sea level, 22. xiii. 1937 (fl), Killip 32263 (NY, US); Herradura, 1904 (fl), van Hermann 562 (BM, F, K, NY, P, US); near Laguna Alcatraz Grande, 16.xii.1911 (fl), Shafer 11016 (NY, US). Las Villas: Banao Mts., Sabana de la Gloria, 29.vii.1918 (fl), Léon & Roca 7928 (NY). Isla de Pinos: San Pedro and vicinity, 12.1i.—22.11.1916 (fl), N. L. Britton, E. G. Britton & Wilson 14148 (F, NY, S, US); Los Indios, 4.ii.1953 (fl), Killip 42603 (NY, LIS); Subsp. styphelioides forms two incompletely correlated clines eastward, showing reduction in size and increasing leaf-imbrication. The only specimen seen from Las Villas (Léon & Roca 7928) is somewhat intermediate between subsp. styphelioides and subsp. moaense. 2c. Hypericum styphelioides subsp. moaense Lippold in Wiss. Z. Friedr.-Schiller Univ. Jena (Math.-Nat. R.) 19: 379, f.2 (1970). Type: Cuba, Oriente, Plano el Sierra de Moa, Bisse 11676 (JE, holotype; HAJB, isotype). Shrub 0-6-1-5 m tall. Leaves 7—11(—16) X 2-6 mm, obovate (or rarely oblanceolate) to narrowly elliptic, thickly coriaceous; slightly glaucous, + cucullate or incurved, usually + densely imbricate; apex apiculate-obtuse to acute; basal veins 5—7; glands prominent. Flowers 15-20 mm in diam. Sepals lanceolate, acute, 7-veined. Petals 8-12 mm long. Styles 5(4), c. 3 mm long. In pine woods and open, dry or marshy habitats, on laterite or serpentine soils; 0-800 m. Confined to Oriente. CUBA. Oriente: Minas de Iberia, Taco Bay, in mountains, c. 800 m, 7—8.xii.1914 (fl), Ekman 3787 (F, K, MICH, NY, S); Moa, near airfield, vii.1941 (fl), Howard 6020 (BM, NY, P, S, US); Maravi (near Barancoa), 25.xii.1914 (st), Ekman 4020 (NY,.5; US), The H. phellos group (Spp. 3-9) Fig. 1, p. 4 The most primitive form of 3 H. phellos has relatively large, marcescent leaves as in 10 H. simonsii. In the other forms of the species, however, the leaves are either (i) smaller and more rigid, deciduous at the apparent base or persisting, or (ii) narrower and more revolute, deciduous at the apparent base only after drooping but not withering. The extreme variability of H. phellos in a relatively restricted area, together with the absence of clear morphological disjunctions, makes it difficult to recognize infraspecific taxa in this species and to differentiate it from its nearest relatives. The form with marcescent leaves (3a subsp. phellos (i) ‘marcescens’) differs from H. simonsii in having leaves free at the base with the apex acute to acuminate and the margin recurved, as well as rounded sepals; but it intergrades with both the form having thicker, narrow, drooping leaves with recurved margins and densely corky internodes (13a subsp. phellos (ii) ‘angustum’) and the form with smaller, chartaceous, spreading leaves, also with recurved margins, but with an acuminate apex and sometimes only slightly corky internodes (3a phellos subsp. phellos (v) ‘diversicaule’). From (ii) ‘angustum’ two distinct trends are apparent in subsp. phellos: to (iii) ‘patens’, in which the leaves are thicker with more recurved margins and usually fall when spreading or soon after becoming appressed, and to (iv) ‘phellos’ (including the type), in which the leaves are narrower and become markedly appressed but not so thick. Variant (ii) ‘patens’ gives rise to the H. garciae group (Spp. 4-6), with narrow revolute leaves, as well as 3b H. phellos subsp. oroqueanum, with narrow continually appressed leaves, and 7 H. irazuense (Costa Rica), whilst the (iv) ‘phellos’ trend continues to 8 H. stenopetalum, which has a distribution overlapping that of H. phellos. On the other main division of H. phellos subsp. phellos, (v) ‘diversicaule’ is related to 9 H. carinosum, which has smaller, THE GENUS HYPERICUM L. 1-nerved leaves and smaller flowers, whereas (vi) ‘tamanum’, which differs from ‘diversicaule’ only in its relatively broader and more obtuse leaves and sepals and more clearly differentiated lateral shoots, gives rise to a high-altitude condensed form (3c subsp. platyphyllum) in which the leaves become incurved and remain appressed until they are shed. 3. Hypericum phellos Gleason in Bull. Torrey bot. Club 56: 106 (1929). Type: Colombia, Santandér, Paramo de la Puentes, above La Baja, 3700-3800 m, 25.1.1927 (fl), Killip & Smith 18199 (NY}, holotype; F!, GH!, K!, US! isotypes). Shrub or small tree 0-3—4 m tall, erect, with branches strict, almost always lateral. Stems yellow-brown, 4-lined and + ancipitous when young, becoming terete, often with transverse epidermal or corky wrinkles or emergences immediately below the node or + throughout the internode, cortex exfoliating in strips; internodes 1-5-3 mm long. Leaves sessile, spreading from the base or imbricate and tetrastichous, deciduous near the base after fading or persistent; lamina 4-20 x (1-)2-6 mm, narrowly to broadly elliptic or oblong to obovate or oblanceolate, plane or recurved, not concave, not or scarcely carinate, concolorous, not glaucous, thickly to thinly coriaceous; apex shortly acuminate to rounded, base cuneate to angustate, not sheathing, free or pairs united to form a narrow interfoliar ridge; basal veins 3-7, diverging, all or only the midrib branching, tertiary reticulation obscure or not visible; laminar glands rather dense, impressed above, + prominent beneath. Inflorescence 1(2—5)-flowered, terminal and on short lateral shoots, with extension shoots usually from 2—3 nodes below; peduncle and pedicels 1-9 mm long; upper leaves foliose or bracteose. Flowers 15—30 mm in diam., stellate. Sepals 4—9 x 1-5-3 mm, + narrowly oblong to lanceolate or elliptic or rarely oblanceolate, acute or acuminate to obtuse or rarely rounded; veins 5—7, branched distally, midrib not or scarcely prominent; glands mostly linear or up to half punctiform. Petals bright yellow to orange-yellow, sometimes margined red, 8-15 x 4-8 mm, c. 2 X sepals, oblanceolate to narrowly obovate; apiculus acute; glands all linear or distally interrupted. Stamens c. 40-200, longest 5-8 mm long, c. 0-5—0-7 x petals. Ovary 2-3-5 X 1-5-2 mm, ovoid; styles 3, (5—)6—9 mm long, 2-3 X ovary, spreading, incurved distally; stigmas small or subcapitate. Capsules 4-5-5 x 3-5-5 mm, subglobose to globose, shorter than sepals. Seeds c. 1 mm long, ecarinate; testa finely scalariform. On scrubby slopes and in moist woods immediately below them; 2600-4200 m. Colombia (Santandér, Norte de Santandér, César), Venezuela (Tachira). Map 2. Hypericum phellos can be distinguished from its nearest relatives only by a combina- tion of characters. Nearly always the stem internodes have corky ridges or emerg- ences, especially below the node, and the leaves are deciduous. Where they are persistent, they are either marcescent with apex acute (‘marcescens’) or broad and remaining appressed, thus apparently preventing the development of the corky emergences (subsp. platyphyllum). The other recognizable subspecies (subsp. oro- queanum) has leaves that remain appressed but eventually fall. 3a. Hypericum phellos subsp. phellos H. tamanum Cuatrec. in Ciencia Mex. 4: 65 (1943). Type: Colombia, Norte de Santandér, Paramo de Tama, vicinity of La Cueva, 3100-3200 m, 27.x.1941 (fl), Cuatrecasas, Schultes & Smith 12647 (COL, holotype; BM!, F!, U!, US!, iso- types). H. thymifolium Cuatrec. in sched., non Kunth. H. tachirense Steyerm. in sched. Shrub or small tree 0-3-4 m tall, with shoots rounded to pyramidal or narrowly cylindric, not or + markedly differentiated into elongate main stems and short laterals; internodes with slight or pronounced corky emergences. Leaves free, 23 24 NORMAN K. B. ROBSON 80 ia Map 2_ 3. H. phellos (all subspecies) @; 10. H. simonsii A; 11. H. papillosum O. spreading then deciduous directly or after decurving or after deflexing and withering; lamina 5-5—16 x 1-4-7 mm, oblanceolate to elliptic or oblong, acute to acuminate, margin recurved to plane. Sepals 4-9 x 2-3 mm, narrowly oblong or elliptic to lanceolate or ovate-lanceolate. Petals 10-15 x 4-8 mm. Stamens c. 40-200. Styles 5—9 mm long. Colombia (Norte de Santandér, Santandér), Venezuela (Tachira); 2500-3900 m. COLOMBIA. Norte de Santandér: Paramo de Fontib6n, 2600-2750 m, 15 —16.x.1941 (fl), Cuatrecasas, Schultes & Smith 12298 (BM, COL, GH, NY, U, US); between Mutiscua and Pamplona, 3400 m, 23.ii.1927 (fl), Killip & Smith 19727 (GH, S, US); Hoya del Rio Chitag4, Quebrada de Presidente, 3100-3300 m, 28.xi.1941 (fl), Cuatrecasas 13490 (COL, NY). Paramo de Tama, vicinity of the Cueva, 3100-3200 m, 27.x.1941 (fl), Cuatrecasas, Schultes & Smith 12647 (BM, F, US). Santandér: Péramo de Santurban entre Bucaramanga & Berlin, W. side, c. 3200 m, 3.i1.1960 (fl & fr), Barclay & Juajibioy 10422 (NY); Berlin to Picacho, 3058 m, 14.viii.1977 (fl), Renteria, Mantilla, Nino & Ortiz 555(3) (COL, MO); vicinity of La Baja, 2700 m, 14—31.i.1927, Killip & Smith 18756 (GH, NY, US). VENEZUELA. Tachira: Péramo de Batallén, near La Grita, 3100 m, 8.i11.1979 (fr), Kieft 104 (BM, U, VEN); SE. of Paramo de Tama, Pata de Judio, near the Colombia-Venezuela frontier, 2900-3000 m, 20.i.1968 (fl), Steyermark & Dunster- ville 101230 (NY, VEN). The following is a summary of the variation of H. phellos subsp. phellos. The variation is continuous, thus preventing the recognition of these variants as taxa. The THE GENUS HYPERICUM L. names in single quotation marks are merely for reference and have no formal nomenclatural status. Variant 3a (i) ‘marcescens’: Stems not differentiated, internodal corky emergences not pronounced; leaves eventually deciduous, drooping and marcescent, narrowly elliptic to oblanceolate, acute to acuminate, margin recurved (Santandér: Vetas, La Baja). Variant 3a (ii) ‘angustum’: Stems not differentiated, internodal corky emergences evident; leaves eventually deciduous, drooping but not marcescent, oblanceolate to linear, acute, margin recurved (Norte de Santandér: Paramo de Fontib6én, Mutis- cua). Variant 3a (ili) ‘patens’: Stems not differentiated, internodal corky emergences evident; leaves deciduous when spreading, oblanceolate to elliptic, acute to obtuse, margin recurved to plane (Santandér: Paramo de Santurban, Norte de Santandér: Rio Chitaga4, Tachira: Paramo de Batall6n). Variant 3a (iv) ‘phellos’: Stems not differentiated, internodal corky emergences evident; leaves deciduous after becoming + appressed, narrowly elliptic to oblan- ceolate, acute, margin recurved to plane (Santandér: La Baja, Paramo de Romeral). Variant 3a (v) ‘diversicaule’: Stems often differentiated, internodal corky emerg- ences evident; leaves deciduous when spreading, broadly elliptic to obovate, acute to obtuse, margin recurved to plane (Norte de Santandér: Péramo de Fontibén, Mutiscua). Variant 3a (vi) ‘tamanum’: Stems differentiated, internodal corky emergences evident; leaves deciduous when spreading or ascending, narrowly elliptic to oblong, subacuminate to acute, margin recurved (Norte de Santandér and Tachira: Paramo de Tama, Santandér: Paramo de Almorzadero). The type specimen of H. phellos apparently belongs to variant 3a (iv) ‘phellos’. 3b. Hypericum phellos subsp. oroqueanum N. Robson, subsp. nov. a subsp. phellos foliis ad caulem continue adpressis differt. Type: Colombia, Cordil- lera Oriental, limites entre Los Departmentos Norte de Santandér y César, Jurisdic- ciones, Cerro de Oroque, 3000—3700-—3900 m, 22-—27.vii.1974, Garcia Barriga & Jaramillo 20629 (COL!, holotype). Small tree 4 m tall, with shoots rounded, not differentiated; internodes with pronounced corky emergences. Leaves free, closely imbricate-appressed, tetra- stichous, not or scarcely spreading, deciduous; lamina 5-5-8 x 1-6-2:2 mm, narrowly elliptic-oblong to oblanceolate, acute, margin plane. Sepals c. 7 x 2:5 mm, broadly elliptic. Petals 12-14 x 4-7 mm. Stamens c. 100. Styles 6-7 mm long. Colombia (Norte de Santandér/César); 3700 m? COLOMBIA. Known only from the type. 3c. Hypericum phellos subsp. platyphyllum (Gleason) N. Robson, stat. nov. Hypericum platyphyllum Gleason in Bull. Torrey bot. Club 56: 106 (1929) pro parte, excl. spec. Killip & Smith 15616. Type: Colombia, Norte de Santandér, Paramo de Romeral, 3800-4200 m, 30.i.1927 (fl), Killip & Smith 18626 (NY!, holotype; A!, COL, GH!, US!, isotypes). Shrub 0-5-2 m tall, with shoots flattened terminally to rounded, not or slightly differentiated; internodes smooth. Leaves united by interfoliar ridge, ascending to closely imbricate, persistent; lamina 4-7 x 2—4 mm, broadly elliptic to obovate, obtuse to rounded, margin recurved to incurved. Sepals 5-7 x 2-3 mm, narrowly oblong to elliptic. Petals 8-14 x 4mm. Stamens c. 80. Styles 5-5-7 mm long. 25 26 SF NORMAN K. B. ROBSON Colombia (Norte de Santandér, Santandér), confined to the Paramos de Santurban and de Romeral; 2700-4100 m. COLOMBIA. Norte de Santandér: Paramo de Santurban, extremo este, 3300 —3500 m, 27.vi.1940 (fl), Cuatrecasas & Garcia-Barriga 10297 (COL, F); Paramo de Romeral, hoya del rio Cucutilla, 3600 m, 9.iv.1973 (veg), Cuatrecasas & Jaramillo 28751 (BM, COL, US). Santandér: Paramo de Santurban, entre Berlin y Vetas, 3700-3800 m, 22.ix.1969 (fl), Cuatrecasas & Rodriguez 27897 (BM, COL, US); Paramo de Romeral, 3800-4100 m, 29-30.1.1927 (fl & fr), Killip & Smith 18580 (NY). Killip & Smith 187716 (Paramo de Romeral) is intermediate between subsp. platyphyllum and subsp. phellos. Gleason (1929) confused subsp. platyphyllum with 29 H. gleasonii (q.v.). 4. Hypericum garciae Pierce in Bull. Torrey bot. Club 70: 174 (1943). Type: Colombia, Santandér, Paramo del Almorzadero, Peralonso, 3200-3300 m, 20.vii.1940 (fl), Cuatrecasas & Garcia- Barriga 9935 (F!, holotype; NY!, US!, isotypes). Hypericum chamaemyrtus sensu Gleason in Bull. Torrey bot. Club 56: 102 (1929), non Triana & Planchon (1862). Hypericum ericifolium Steyerm. in Fieldiana Bot. 28: 393 (1952) [‘ericaefolium’]. Type: Venezuela, Tachira, Paramo de Tama, 2 km above Betania, 7 km above Villapaez, 2500 m, 14.vii.1944 (fl), Steyermark 57212 (F!, holotype; NY!, isotype). Shrub 0-3-1-2(-3) m tall, erect, with branches strict or ascending, lateral or occasionally pseudo-dichotomous. Stems orange-brown to blackish, 4-lined and ancipitous when young, the principal lines broader, eventually terete, without epidermal emergences, cortex exfoliating in strips; internodes 1-4 mm long. Leaves sessile, outcurving or ascending and subimbricate, tetrastichous, deciduous at suprabasal articulation; lamina 4-10 x 0-5—2-3 mm, elliptic-oblong or narrowly lanceolate to linear, broadly revolute and often concealing lower surface apart from midrib, not carinate, concolorous, lucent above, papillose beneath, not glaucous, coriaceous; apex acute, base cuneate, not sheathing, pairs united to form a narrow interfoliar ridge; basal vein unbranched; laminar glands dense, slightly prominent. Inflorescence 1(2)-flowered, terminal and on short lateral shoots, occasionally with pseudo-dichotomous branches; peduncle and pedicels 1-5-3 mm long, upwards incrassate; upper leaves foliose. Flowers 17-20 mm in diam., stellate. Sepals 4-9 x 1-7-3 mm, lanceolate to ovate, acute to subacuminate; veins 5—7, obscure, not visibly branching, midrib prominent; glands linear, distally punctiform. Petals bright to deep yellow, sometimes tinged red, 8—13 x 3-6-5 mm, 1-5-2 X sepals, narrowly obovate; apiculus acute; glands linear, interrupted distally. Stamens 40-70(?-100), longest 5—8 mm long, c. 0-6—0-65 x petals. Ovary 1-5—2-5 x 1-2 mm, ovoid; styles 3(4), 3-5-7 mm long, 2:2—2-5(—3) X ovary, free, spreading-incurved; stigmas small or slightly capitate. Capsule 4-5-5 x 3-3-5 mm, ovoid to subglobose, shorter than sepals. Seeds c. 1:3 mm long, ecarinate; testa finely scalariform. In paramo, usually on dry stony or sandy soil; 2400-3935 m. Colombia (Norte de Santandér to Meta), Venezuela (Tachira). Map 3. COLOMBIA. Boyaca: Paramo de La Rusia, NW. to N. of Duitama, Serrania Pena Negra, Hoya de la Laguna Agua Clara, 3935 m, 10.xii.1972 (fl), Cleef 6992 (COL, U); Vado Hondo, Siberia, between Pefia de Arnical and Alto de Mégotas, 3290 m, 4.iv.1973 (fr & fr), Cleef 9332 (BM, COL, U). Cundinamarca: Municipio de Tausa, paramo vecindad a la poblacion, 3200 m, 26.x.1961 (fl), Huertas & Camargo 5285 (COL). Meta: Péaramo de Sumapaz, Hoya El Nevado, Laguna El Sorbedero, 3550 m, 1.ii.1972 (fl), Cleef 1489A (COL). Norte de Santandér: Paramo de Hatico between Toledo and Pamplona, c. 2900 m, 12—13.i11.1927 (fl), Killip & Smith 20661 (K, NY, US). Santandér: Paramo del Almorzadero, 3600-3800 m, 28.xi.1941 (fl), Cuatrecasas 13496 (F, NY, P); Paramo Rico, near Vetas, 3750-3850 m, 18.i.1927 (fl), Killip & Smith 17659 (NY). THE GENUS HYPERICUM L. Map3 4. H. garciae @;5.H. acostanum @. VENEZUELA. Tachira: Paramo de Tama, El Paramito, 2550 m, viii.1939 (fl), Chardon 81 (VEN). H. garciae differs from the coriacous-revolute-leaved form of H. phellos ssp. phellos (‘patens’) principally by its narrower, closely revolute leaves, a character that distinguishes it from all other species in sect. Brathys, except H. acostanum and H. piriai. In addition, these species are unique in having the leaf-base incorporated in the stem and forming a swelling, thus indicating that the leaf is not deciduous at the base. The largest flowers and leaves in H. garciae occur in Boyaca, whence there is a reduction trend along the Cordillera Oriental north-eastward to Tachira. As the variation is continuous, H. ericifolium cannot be recognized as a distinct species. 5. Hypericum acostanum Steyerm. [in Acosta-Solis, Div. fitogeogr. y formac. geobot. Ecuador: 103 (1968), nomen] ex N. Robson, sp. nov. H. garciae Pierce affinis, sed foliis patulis persistentibus, pro ratione angustioribus, uninervis, margine revolutis, sepalis angustioribus, petalis longioribus, inter alia differt. Type: Ecuador, Loja, Cant6n Catachocha, Hacienda La Hamaca, Loma Larga, 2200-2400 m, 15.iv.1944 (fl), Acosta Solis 7839 (F!, holotype; NY!, photo- graph). Shrub 0-1-0-8 m tall, erect, with branches strict, pseudo-dichotomous and lateral. Stems orange-brown, 4-lined and ancipitous when young, the principal lines broader, eventually terete, without epidermal emergences, cortex exfoliating in strips; inter- nodes 2-5 mm long. Leaves sessile, outcurving or spreading but not imbricate, zy NORMAN K. B. ROBSON scarcely tetrastichous, persistent, not articulated but sometimes breaking off above base; lamina 8-22 x 0-5—2-5 mm, narrowly oblanceolate to linear, broadly to narrowly revolute, sometimes concealing lower surface apart from midrib, not carinate, paler or ferrugineous and smooth beneath, lucent above, not glaucous, coriaceous; apex acute, base parallel-sided, not sheathing, pairs united to form a narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense, impressed. Inflorescence 1-flowered, terminal, often with pseudo-dichotomous branches; pedicel 3-15 mm long, not upwards incrassate; upper leaves foliose. Flowers 20-25 mm in diam., stellate. Sepals 6-9 x 1-1-5 mm, narrowly triangular-lanceolate to oblong-linear, acute, sometimes revolute above; veins 5—7, obscure, not visibly branching, midrib slightly prominent; glands linear, distally punctiform. Petals bright (?) yellow, tinged red in bud, 10-14 x 5—7 mm, c. 2 X sepals, obovate; glands linear and striiform. Stamens c. 100, longest 4-5 mm long, c. 0-4 x petals. Ovary c. 2 x 1mm, narrowly ovoid; styles 3, 3-3-5 mm long, c. 1:5 X ovary, divergent; stigmas broadly capitate. Capsule 4-5 x 3-4 mm, ovoid-subglobose, shorter than sepals. Seeds c. 1 mm long, ecarinate; testa finely scalariform. On steep shrubby slopes of paramo; 2200-3300 m. Ecuador (Loja, Morona-Santiago). Map 3. ECUADOR. Loja: between San Pedro and Portovelo, slope towards Portovelo, c. 2300 m, 13.x.1955 (fl & fr), Asplund 18133 (S); Celica-Guachanama road, Km 8 at the Roldés memorial monument, 2700 m, 19.11.1985 (fl & fr), Harling & Anderson 2223 (BM, GB). H. acostanum is clearly related to H. garciae, for which the nearest record is from central Colombia (Meta). 6. Hypericum piriai Arechav. in An. Mus. nac. Montevideo 3: 108 (1898), Fl. Uruguay 1: 108 (1898); L. B. Sm. inJ. Wash. Acad. Sci. 48: 314 (1958); Rodriguez Jiménez in Reitz, Fl. Ill. Catarin., Hipericaceas: 5, fig. 1 B (1980). Type: Uruguay, Mardonado, Pan de Aztcar cerca del Castillo de Piridpolis, xi.189- (fl & fr), Arechavaleta (MVM, holotype; GH!, US!, photographs). H. tenuifolium St.-Hil. in St.-Hil., Jussieu & Cambess., Fl. Bras. Merid. 1: 337 (1828); Walpers, Repert. Spec. nov. 1: 390 (1842); Dietrich, Syn. PI.: 1236 (1847); Reichardt in Martius, Fl. Brasil. 12: 188 (1878), non Pursh (1814). Type: Brazil, Sao Paulo, prope urbes Sancti Pauli, Villa do Castro et Curityba, iii. 1820 (fl & fr), St.-Hilaire 1174 (P!, holotype; US!, photograph). H. hilaireanum L.B. Sm. in J. Wash. Acad. Sci. 48: 314 (1958); Angely, Fl. Anal. Paranda: 451 (1965). Type as for H. tenuifolium St.-Hil. Subshrub or perennial herb 0-07-0-35 m tall, erect, with taproot, branches few, strict or decumbent, lateral, from base or lower half of stem. Stems green, 4-lined above, eventually terete, gland-dotted, with + prominent glands along raised lines; cortex eventually exfoliating in strips; internodes 3-21 mm long. Leaves sessile, outcurving and deflexing, not tetrastichous, persistent, not articulated but occasionally breaking off above base; lamina 10-40 x 1—2 mm, linear, revolute, concealing all or most of lower surface apart from midrib, not carinate, paler beneath?, dull above, + glaucous, subcoriaceous; apex acute to acicular, base parallel, not sheathing, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense, slightly impressed or not. Inflorescence 1—18(—25)-flowered, terminal and sometimes axillary, monochasial after first branching; pedicels 3-6 mm long; upper leaves bracteose. Flowers 8-20 mm in diam., obconic to pseudotubular. Sepals 6-16 xX 1-5-2-5 mm, linear-lanceolate to lanceolate or narrowly oblong, acute to acumin- ate, margin slightly revolute or plane; veins 3-5, unbranched, midrib slightly prominent or not; glands linear towards base or wholly punctiform. Petals bright (?) yellow, not (?) tinged red in bud, 7-12 x 3-4-5 mm, 0-8-1-2 x sepals, obovate- oblong; glands linear, distally punctiform. Stamens 25—60, longest S—8(—10) mm, c. 0-75 x petals, 3(5)-fascicled, the double (antesepalous) fascicles sometimes incom- THE GENUS HYPERICUM L. pletely united, with filaments within each fascicle united shortly or usually to 0-7 of their length. Ovary 1-5—4 x 1-1-5 mm, narrowly ovoid to narrowly ovoid-ellipsoid; styles 3, 2-4 mm long, c. 1-1-3 X ovary, outcurving; stigmas subcapitate. Capsule 5-7 X 2-2-3 mm, ovoid-cylindric to ellipsoid, shorter than sepals. Seeds 0-8-1 mm long, ecarinate; testa finely scalariform. In dry, stony or damp grassland; lowland to 1000 m. Brazil (SAo Paulo, Parana, Santa Catarina, Rio Grande do Sul), Uruguay. Map 4. BRAZIL. Sao Paulo: Sao Paulo to Villa Emma, xii. 1933 (fl), Brade 12978 (MO). Parana: Municipio Guarapuava, estrada para Laranjeiras do Sul, 15.xi.1957 (fl), Hatschbach 4253 (US); Municipio Ponta Grossa, Parque Vila Velha, Furnas, i.x.1965 (fl & fr), Hatschbach 12854 (K). Santa Catarina: Municipio Abelardo Luz, 8-12 km N. of Abelardo Luz, 900-1000 m, 15.xi.1964, Smith & Klein 13305 (NY, US); Municipio Xanxere, 11 km N. of Abelardo Luz, 500-600 m, 25.xii.1956, Smith & Klein 9231 (US). Rio Grande do Sul: Vila Manresa, prope Porto Alegre, 21.x.1951, Rambo 51325 (US); Porto Alegre, Morro da Policias, 21.ii.1902 (fl & fr), Malme Regn. II 1408 (S). URUGUAY. Mardonado: see type. Lavalleja: Co. Penitente, 10.1.1909 (fl), Berro 5533 (US). Map 4 6. H. piriai @. Despite the considerable distance between southern Ecuador and south-eastern Brazil, there can be no doubt that H. piriai is a close relative of H. acostanum. The character trends between H. garciae and H. acostanum are extended to H. piriai (for example: tall tall shrub—dwarf shrub—subshrub to herb; leaves short, deciduous —long, persistent; sepals short, broad—long, narrow, revolute—long to short, narrow, revolute to plane). However, the flower shows new tendencies, viz. (i) towards development of a cymose inflorescence and (ii) to become adapted to specialized rather than open pollination. Unlike the other examples of this change in Hypericum (H. elodes and the species of sect. 25. Adenotrias), H. piriai shows intermediate evolutionary stages and the specialization is less complex. Thus (i) the flower apparently varies from obconic to pseudotubular and the stamen filaments within each fascicle vary correspondingly from almost free to c. 0-7 united; (ii) the double fascicles are often incompletely united; (iii) there are no fasciclodes to help open the flower by expansion and no ligules on the petals to guide the tongues of foraging insects. Ze 30 NORMAN K. B. ROBSON I agree with Rodriguez Jiménez (1980) that H. piriai cannot be divided into a northern species (H. hilaireanum) and a southern one (H. piriai sensu stricto). The more northern ones, suffruticose with fewer larger flowers and longer sepals, are the nearest morphologically to H. acostanum. 7. Hypericum irazuense Kuntze [Rev. Gen. Pl: 61 (1891), in synon.; Williams in Fieldiana Bot. 29: 355 (1961), nomen] ex N. Robson, sp. nov. H. phellos subsp. oroqueano N. Robson affinis, sed foliis persistentibus nervis haud ramosis, differt; a H. stenopetalo Turcz. foliorum nervis basalibus haud ramosis floribus maioribus, stylis longioribus, differt. Type: Costa Rica, Cartago, Volcan Irazu, 3150 m, 25.vi.1874 (fl), Kuntze 2359 (NY!, holotype; K!, isotype). H. caracasanum var. scherzeri J. Zahlbr., in sched. H. silenoides sensu R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3(6): 214 (1893), in Bull. Herb. Boissier 6: 258 (1898), in op. cit. 11, 8: 177 (1908), in Engler & Prantl, op. cit., 2nd ed. 21: 181 (1925), non Juss. (1804). Icon: Fig. 5A. Shrub or small tree 0-4-5 m tall, erect, flat-topped, with branches strict, lateral and sometimes pseudo-dichotomous, Stems yellow-brown, 4-lined and ancipitous when young, becoming terete, with transverse corky emergences, cortex exfoliating in strips; internodes 1-5—4 mm long. Leaves sessile, imbricate and tetrastichous at first, spreading then becoming appressed, tardily deciduous near the base (usually after withering) or persistent; lamina 10-15 Xx 1-5-3-5 mm, narrowly elliptic to oblan- ceolate, plane to recurved, not carinate, concolorous, not glaucous, subcoriaceous to coriaceous; apex acute, base cuneate, not sheathing, pair free but with incomplete narrow interfoliar ridge; basal veins 3-5, parallel, inner pair (of 5) branching above, tertiary reticulation not visible; laminar glands rather dense, impressed above, not prominent beneath. Inflorescence 1-flowered, terminal and on short lateral shoots, with extension shoots from 2—3 nodes below; pedicel 2-5—4 mm long; upper leaves bracteose. Flowers 25-30 mm in diam., stellate. Sepals 7-5-9 x 1-5-2 mm, linear- lanceolate, acute; veins 5, unbranched, midrib not prominent; glands mostly linear, punctiform near apex. Petals bright yellow to orange-yellow, not (?) tinged red, 13-18 x 7-10 mm, c. 2 X sepals, oblanceolate to obovate; apiculus acute; glands linear, interrupted distally. Stamens c. 100, longest 6-9 mm long, 0-35-0-5 x petals. Ovary 1-5—2-5 x 1-1-5 mm, ovoid; styles 3, 9-10 long, 4-6 x ovary, free, spreading, distally incurved; stigmas small. Capsule 6-7 x 4-5 mm, broadly ovoid or ellipsoid to globose, shorter than sepals. Seeds 1-1-3 mm long, scarcely carinate; testa finely scalariform. 2n = 12. On open paramo slopes or among bamboo (Chusquea); 2700-3730 m. Costa Rica (San José, Cartago, Lim6n), on the cordillera and on Volcan Irazi and V. Turrialba, and in adjacent Panama (Chiriqui). Map 5 (p. 32). COSTA RICA. Cartago: Volcan Irazt, 3000-3300 m, 5.v.1928 (fl), Stork 2011 (F, MICH, NY); Finca Quemado to top of Volcan Turrialba, 2800-3200 m, 10.v.1971 (fl), Wilbur 14319 (DUKE); Cordillera de Talamanca, Cerro de la Muerte, Pan- American Hwy., 5 km above Millsville, 3400-3500 m, 22.vii.1949 (fl), Holm & Iltis 458 (F, K, NY, P). Limon: Chirrip6 National Park, between Casa de Administracié6n and peak, c. 3400 m, 13.11.1983 (fl & fr), Garwood et al. 1161 (BM). San José: Upper Rio Talari, lower Valle de los Conejos and trail to Valle de los Leones, 3250-3450 m, 21-23.viii.1971 (fl & fr), Burger & Gomez P. 82848 (BM, DUKE, F); La Asuncion, Cerro Sakira, 3400 m, 27.i.1968 (fl), Wilbur & Stone 10047 (DUKE, F, MO). PANAMA. Chiriqui: Cerro Fabrega and vicinity near Costa Rican frontier, 3150-3335 m, 7—8.iv.1976 (fl), Weston 10189 (MO). H. irazuense has apparently been derived from H. phellos subsp. oroqueanum in a similar way to the derivation of H. stenopetalum from H. phellos subsp. phellos, but it can be distinguished from H. stenopetalum both morphologically and geographi- cally. THE GENUS HYPERICUM L. 31 (} () N’ \/I\A Ny wa VY SY, Ni Wii SING win Sh N WZ Ws SS "yj NS — p> ss Wwe \\ ANY Ye \ i fl Fig.5 A. H. irazuense: (a) habit; (b) stem with leaves; (c) leaf; (d) sepal; (e) petal; (f) stamens (partly cut away) and ovary; (g) capsule. B. H. stenopetalum: (h) leaf; (i) sepal; (j) petal; (k) stamens (partly cut away) and ovary (a X 12;b x 1; d—g, i-k x 3;c,h x 4). A. Khan, Vickery & Tebbs 1398; B. Luteyn 6032. The record of H. irazuense from Guatemala is apparently erroneous. A Guatema- lan label was used for the Kew (K) specimen of Friedrichstal 1395, but the Chicago (F) specimen of this collection has a similar label with ‘Guatemala’ crossed out and replaced by ‘Cartago’. 8. Hypericum stenopetalum Turcz. in Bull. Soc. Nat. Moscou 31 (1): 385 (1858). Knuth in Reprium Spec. nov. Regni veg. Beih. 43: 484 (1927); Gleason in Bull. Torrey Bot. Club 56: 103 (1929). Type: 32 NORMAN K. B. ROBSON Map5 7. H. irazuense A; 8. H. stenopetalum @. Venezuela, Mérida, Sierra Nevada, 2940 m, 1846 (fl), Funck & Schlim 1139 (LE, holotype; BM!, P!, isotypes). H. decorticans Planchon & Linden in Weddell, Chloris Andina 2: 272 (1861); Triana & Planchon in Annis Sci. nat. (Bot.) IV, 18: 297 (1862), nomen. H. meridense Steyerm. in Fieldiana Bot. 28 (2): 394. (1952). Type: Venezuela, Mérida, between San José and Beguilla, Paramo de Pozo Negro, 2590-3220 m, 3.v.1944 (fl & fr), Steyermark 56272 (NY!, holotype; F!, isotype). Icon: Fig. 5B. Shrub or small tree 0-5—4 m tall, erect, branches strict or ascending, lateral or rarely pseudo-dichotomous. Stems yellow-brown, 4-lined and + ancipitous when young, becoming terete, without or with weak corky emergences, cortex exfoliating in strips; internodes 1-5-3 mm long. Leaves sessile, imbricate and tetrastichous at first, becoming appressed, eventually deciduous near base or persistent; lamina 4-15 x (1-)1-5—4-5 mm, narrowly or rather broadly elliptic to oblanceolate, plane or recurved or rarely concave to cucullate, not or scarcely carinate, concolorous, not glaucous, subcoriaceous to coriaceous; apex shortly acuminate to subacute, base cuneate to angustate, not sheathing, pairs free but with incomplete narrow interfol- iar ridge; basal veins 3—S, diverging, all or only midrib branching, tertiary reticula- tion obscure or not visible; lamina glands rather dense, impressed above, + prominent beneath. Jnflorescence 1-flowered, terminal and sometimes on + short lateral shoots, with extension shoots usually from 2nd or 3rd node below; pedicel 1-9 mm long; upper leaves foliose or bracteose. Flowers 15-27 mm in diam., stellate. Sepals 4-9 x 1-5-3 mm, narrowly oblong or elliptic to ovate-lanceolate or rarely oblanceolate, acute or shortly acuminate to obtuse; veins S—7, branched distally or not, midrib not or scarcely prominent; glands mostly linear or up to half punctiform. Petals bright yellow to orange-yellow, sometimes margined red, (6—)8-15 x 4-8 mm, c. 2 X sepals, oblanceolate to obovate; apiculus acute; glands all linear or distally interrupted. Stamens 65-120, longest 5—8 mm long, 0-5—0-7 x petals. Ovary 2-3 X 1-2 mm, + broadly ovoid; styles 3, (S—)6—9 mm long, 2-4 X ovary, free, spreading, distally incurved; stigmas small or subcapitate. Capsule 4-6 x 3-5-5 mm, broadly ovoid to globose, shorter than sepals. Seeds c. 1 mm long, ecarinate; testa finely scalariform. On open scrubby or stony paramo slopes and among shrubs or bamboos in the tree/shrub zone immediately below them; 2580-4200 m. THE GENUS HYPERICUM L. Colombia (Santandér), Venezuela (Tachira, Mérida). Map 5. COLOMBIA. Santandér: [Péramo de Almorzadero] Peralonso, Timocato, 3600 m, 29.ix.1969 (fl), Cuatrecasas & Rodriguez 27871 (BM, COL, US); Paramo de Santurban, c. 3000 m, 27.viii.1948 (fl), Barkley & Araque M. 188169 (COL, US). VENEZUELA. Mérida: Paéramo de Los Leones (La Lagunita, La Canada Grande) al W. de Murucuba, 3400-3500 m, 31.v.1930 (fl), Gehringer 159 (F, NY, VEN); Péramo de San José, monte El Cupio, 3000 m, 1.i1.1973 (fl & fr), Cuatrecasas, Ruiz-Teran & L6épez-Figueras 28440 (BM, US). Tachira: Paramo de Portachuelo, 2860 m, 23.x.1978 (fl), Luteyn 6032 (BM, MO, NY, VEN). Linden 446 (BM, K, W) is labelled ‘Caracas’. In the absence of other records from that area this probably merely indicates Venezuela. 9. Hypericum carinosum R. Keller in Bull. Herb. Boissier, I, 8: 182 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925); Knuth in Reprium Spec. nov. Regni veg. Beih. 43: 483 (1927); Gleason in Bull. Torrey bot. Club 56: 102—103 (1929). Type: Venezuela, Mérida, Paramo de Portachuelo, 2600-2700 m, 1846 (fl), Funck & Schlim 1459 (W!, holotype; BM!, P!, isotypes). H. nova species sensu Turcz. in Bull. Soc. Nat. Moscou 31 (2): 386 (1858); Knuth in Reprium Spec. nov. Regni veg. Beih. 43: 484 (1927). H. struthiolifolium sensu Trev., Hyper. Animad: 15 (1861) [‘struthiolaefolium’| pro parte, excl. typum. H. thymifolium sensu Triana & Planchon in Annis Sci. nat. (Bot.) IV, 18: 297 (1862) pro parte, quoad syn. Trevirani. H. stenoclados Cuatrec. in Ciencia Mex. 4: 64 (1943). Type: Colombia. Santandér, Paéramo de Almorzadero, extremo sur, Peralonso, 3200 m, 19.vii.1940 (fi), Cuatrecasas & Garcia Barriga 9919 (COL, holotype; F!, NY!, US! isotypes). Shrub 0-6—c.2-5 m tall, erect, with branches strict, lateral, short (flowering) and long (extension). Stems purplish-brown, 4-lined and ancipitous when young, eventually terete, without epidermal emergences, cortex exfoliating in strips; internodes 2—9 mm long. Leaves sessile, spreading from the base or ascending but not imbricate or tetrastichous, persistent; lamina 3-5—5 x 0-8-2 mm, narrowly elliptic or narrowly oblong to oblanceolate-spathulate, plane, rarely apically subconcave, markedly carinate, concolorous, not or slightly glaucous, chartaceous to subcoriaceous; apex acute to subacute, base cuneate to angustate, not sheathing, pairs almost free (interfoliar ridge rudimentary); basal vein 1, unbranched; laminar glands dense, not impressed above, prominent beneath. Jnfloresence 1—-2(—3)-flowered, terminal and on short lateral pinnate branches; peduncle and pedicels c. 2 mm long; upper leaves bracteose. Flowers 8-13 mm in diam., stellate. Sepals 2-4-5 x 0-5-1-8 mm, ovate-lanceolate to triangular-lanceolate, acute; veins 3(5), unbranched, midrib distally incrassate; glands linear, interrupted distally. Petals bright (?) yellow, 4-8 x 2-3 mm, c. 2 X sepals, oblong-lanceolate to obovate; apiculus acute; glands linear, interrupted distally. Stamens 40-80, longest 3-4 mm long, 0-5—0-6 x petals. Ovary 1-5-2 x 0-8-1 mm, ovoid; styles 3, 4-6(—7) mm long, c. 2-3 X ovary; stigmas small. Capsule c. 4 X 3-5 mm, subglobose, shorter than sepals. Seeds not seen. In thickets and woodland margins in the paramo; 2600-3200 (—4500) m. Colombia (Norte de Santandér, Santandér), Venezuela (Mérida). Map 6 (p. 37). COLOMBIA. Norte de Santandér: environs of Pamplona. 10.x.1943 (fl), de Garganta 626 (F). Santandér: vicinity of Las Vegas, 2600-3600 m, 21—23.xii.1926, Killip & Smith 15921 (GH, NY, US); vicinity of California, 2800 m, 11—27.i.1927, Killip & Smith 16904 (F, GH, NY, US). VENEZUELA. Mérida: Paramo de Los Leones, 3400-4500 m, 31.v.1930 (st), Gehringer 149 (VEN). H. carinosum is related to H. phellos subsp. phellos ‘diversicaule’, and its area of distribution is to the south-east of that of the latter. The two taxa, however, are quite 3D 34 NORMAN K. B. ROBSON distinct, H. carinosum being recognizable (inter alia) by the stem internodes without corky emergences, the smaller, persistent, 1-nerved leaves, and the smaller flowers. H. stenoclados has subacute, subcoriaceous leaves rather than the acute, char- taceous leaves typical of H. carinosum, but these differences do not merit taxonomic recognition. 10. Hypericum simonsii N. Robson, sp. nov. H. phellos affinis, sed caulis haud suberosis, foliis persistentibus deflexis oblongis apice rotundatis basi conjunctis venatione valde reticulata, sepalis subacutis vel rotundatis, stylis (S—)6—9 mm longis, differt. Type: Colombia, Magdalena, Sierra Nevada de Santa Marta, 1880 (fl), Simons s.n. (BM!, holotype). H. stenopetalum [var.] B majus Triana & Planchon in Annls Sci. nat. (Bot.) IV, 18: 297 (1862) [‘major’]. Types: Colombia, Magdalena, Santa Marta, source [‘prov.’] of Rio Hacha, 3700-3800 m, vii.1844 (fl & fr), Purdie s.n. (G, lectotype; GH!, K!, syntypes); ibid., 1851-1852 (fl), Schlim 838 (G, syntype). Icon: Fig. 6A. Shrub 1-2 m tall, erect, with branches ascending, mostly lateral. Stems reddish- brown, 4-lined when young, eventually terete, without epidermal wrinkles, cortex flaking irregularly; internodes 2-10 mm long. Leaves sessile, spreading from the base, not tetrastichous, becoming brown and deflexed, usually persistent until cortex is shed; lamina (7—)8—16(—20) x (2-2—)3-5-2(—7) mm, narrowly elliptic to narrowly oblong, plane, not concave or carinate, concolorous, pale green, chartaceous; apex rounded, base cuneate, not sheathing, pairs united to form a narrow interfoliar ridge; basal or near-basal veins 5—7, with branches and tertiary reticulation rather conspicuous; laminar glands dense, visible beneath only. Inflorescence 1 (2)- flowered, mostly terminal and lateral, occasionally pseudo-dichotomous; peduncle and pedicels 4-7 mm long; upper leaves not transitional. Flowers 20-35 mm in diam., stellate. Sepals 6-10-5 x 2-4-3-5 mm, lanceolate or narrowly elliptic to narrowly oblong or oblanceolate, subacute to rounded; veins 5—7, not or obscurely branched, midrib not prominent; glands distally punctiform. Petals bright yellow, 10-21 x 7-13 mm, c. 2 X sepals, oblanceolate; apiculus obtuse; glands distally interrupted. Stamens c. 100, ‘obscurely 5-fascicled’, longest c. 7 mm long, 0-4—0-5 x petals. Ovary 2:5—3-5 x 2-5 mm, ovoid; styles 3-4, 4-5-6 mm long, c. 2 X ovary, spreading, distally incurved; stigmas narrowly capitate. Capsule c. 8 X 5 mm, subglobose, shorter than sepals. Seeds not seen. In paramo, bushy prairies, and thickets; 3250-4100 m. Apparently confined to the Sierra Nevada de Santa Marta in Colombia (Magdalena). Map 2 (p. 24). COLOMBIA. Magdalena: near Rio Seville, 20.i1.1959 (fl), Barclay & Juajibioy 6557 (MO, NY); Sierra de Santa Marta, c. 48 km inland from Dibulla, c. 3850 m, vii.1932 (fl), Seifriz 455 (US); Sierra de Santa Marta, SE. slopes, Hoya del Rio Donachui, Meollaca, 3320-3260 m, 29.ix.1959, Cuatrecasas & Romero-Castaneda 24480 (COL, US); Sierra de Santa Marta, source of Rio Donachui, 4310 m, 22.v.1977, Starker White & Alverson 547A (NY); Sierra de Santa Marta, quebrada from Laguna Rio Frio, 3250 m, 28.vii.1972, Kirkbride & Forero 1751 (BM, COL). H. simonsii, which is apparently confined to the Sierra de Santa Marta massif, is related to H. phellos (Venezuela-Colombia border area). It is morphologically more specialized in that its leaves are united by an interfoliar ridge, but it is more primitive in that they have S—7 basal or near-basal veins. 11. Hypericum papillosum N. Robson, sp. nov. H. simonsii N. Robson affinis, sed habitu humiliore densiore, foliis sepalisque subtiliter papillosis, foliis angustioribus apice obtusis vel rotundatis basi breviter vaginatis, stigmatibus valde capitatis, differt. Type: Colombia, Boyaca, Sierra THE GENUS HYPERICUM L. 35 \ - fy. S pS a aaa px > Ne Fig.6 SYOIy WOH Q Wer RY 9 BN eee ro, VA, Si | » RASA WANG Ze Ma SW) a VV es QE WAKE y Si| ih My MY AN \ ass \ We Vv ‘3a SEES nN y WN BROS GA i GL SS NE Ni h i SENAY = Mg (8 Ug 7 ca) eS, row Se LO SYVAY YN IN ae Ss = ON, C3 RE (A) (RR A Wy ¥) A | a \ \ \! \J Lilt All; Mya] Wale d \ i M-T- LL EL Fig. 7 H. quitense: (a) habit; (b) stem with leaves; (c) leaf; (d) sepal; (e) petal; (f) stamens (partly cut away) and ovary; (g) capsule (a x 4; b, f, g x 4; c—-e x 6). All Camp 4133. mm long. Leaves sessile, spreading to deflexed, obscurely tetrastichous, turning brown, deciduous with the cortex; lamina 4—9-5 x 2-5—5 mm, sometimes increasing in size along shoot, oblong (lower) to broadly ovate, plane or slightly saccate or conduplicate, midrib prominent beneath, margin narrowly hyaline, otherwise pale green but not or only slightly glaucous, subcoriaceous; apex acute (lower) to obtuse, base broadly cuneate, pairs shortly united; basal or near-basal veins 3(5), with laterals and sometimes midrib branching, tertiary reticulum obscure or apparently absent; laminar glands mainly peripheral, dense, rather large, not prominent. Inflorescence 1-flowered, terminal and on very short lateral shoots, the whole shoot narrowly cylindrical, without pseudo-dichotomous branches; pedicel 3—7 mm long; upper leaves foliose. Flowers 8-12 mm in diam., stellate. Sepals 5 x 1-2 mm, elliptic to narrowly oblong, the outer sometimes broader, apiculate-obtuse to acute, sub- cucullate, margin narrowly hyaline; veins 3-5, not or scarcely branching, all promin- ent; glands all linear or distally punctiform. Petals bright yellow, 5-6-5 x 3-4 mm, 1-5-2 x sepals, obovate; apiculus subacute to obtuse; glands few, striiform and THE GENUS HYPERICUM L. punctiform. Stamens 30-45, longest 3—4 mm long, c. 0-65 x petals. Ovary 1-5-2 x 0-7-1 mm, narrowly ovoid; styles 3, 3-4 mm long, c. 2 X ovary, spreading; stigmas broadly capitate. Capsule 3-4-5 x 2-5—3-5, ellipsoid, equalling or slightly exceeding sepals. Seeds c. 0-7 mm long, ecarinate, testa finely scalariform. In damp meadows of the paramo; 2600-3800 m. Ecuador (Bolivar, Tungurahua, Cafar, Azuay). Map 7 (p. 39). ECUADOR. Tungurahua: around Lago Pisayambo, 3800 m, 26.xii.1964 (fl), Knight 268 (BM). Bolivar: Cordillera Occidental, Pucara de Telimbela, 2600-3000 m, 18.xi.1943 (fl), Acosta Solis 6825 (F). Canar: near Pimo, 3060-3120 m, 9.vii.1945 (fl), Camp E-4133 (BM, NY). Azuay: mountains above Sayausid, 3000-3200 m, 18. iii. 1974 (fl & fr), Harling & Anderson 12630 (GB). H. quitense is related to H. cuatrecasii but is much smaller in all parts. 18. Hypericum loxense Bentham Pl. Hartweg.: 126 (1843); R. Keller in Bull. Herb. Boissier II, 8: 177 (1908), in Engler & Prantl, Nat. Pfanzenfam. 2nd ed. 21: 181 (1925); Gleason in Bull. Torrey bot. Club 56: 102 (1929). Type: Ecuador, Loja, ‘in montibus Loxa’, viii.1842? (fl & fr), Hartweg 720 (K!, holotype; BM!, F!, P!, isotypes). Shrub or shrublet, 0-2—1(—1-5) m tall, erect or decumbent to prostrate, with branches ascending to strict, lateral, short (flowering) and long (extension), not pseudo- dichotomous. Stems reddish-brown, 4-lined and ancipitous when young, eventually terete, without wrinkles or emergences, cortex exfoliating in irregular flakes or strips; internodes 1-5—10 mm long. Leaves sessile, erect, imbricate and sometimes markedly tetrastichous when young, spreading and turning brown, deciduous with the cortex (or sometimes breaking off earlier); lamina 2-10 x 0-6-3 mm, not increasing in size along shoot, narrowly oblong or oblanceolate or linear to triangu- lar-lanceolate or triangular-ovate, incurved-conduplicate or subsaccate, midrib prominent beneath or not, margin narrowly hyaline, not or slightly glaucous, subcoriaceous to chartaceous; apex acute to obtuse, base narrowly cuneate or parallel-sided to broadly cuneate or truncate, pairs shortly united; basal veins 1(3-5), with midrib sometimes branching, tertiary reticulum rarely present; laminar glands dense to rather sparse, varying in size, prominent or not. Inflorescence either (i) wholly 1-flowered, terminal and on very short lateral shoots, the whole shoot narrowly cylindric, with lateral branches sometimes also bearing flowering branches but rarely with pseudo-dichotomous branches, or (ii) 1—c. 15-flowered in terminal and sometimes lateral dichasia or mixed dichasia and pseudo-dichotomies; peduncle and pedicels 2—9 mm long; upper leaves foliose, bracts gradually smaller. Flowers 6-10 mm in diam., stellate. Sepals 1-5—5-3 x 0-5—1-6 mm, elliptic or lanceolate to narrowly oblong, the outer sometimes broader, apiculate-obtuse to acute, some- times cucullate, margin narrowly hyaline; veins 3—S, not or scarcely branching, not or only the midrib slightly prominent; glands all linear or distally punctiform. Petals pale or bright yellow to orange, darker or reddish in bud, 3—6(—7-5) x 1-5-3 mm, 1-5-2 X sepals, obovate; apiculus obtuse or obsolete; glands few, striiform and punctiform. Stamens 20—40, longest 2-4 mm long, c. 0-65 X petals. Ovary 0-7—1-5 x 0-5-1 mm, narrowly ovoid; styles 3(4), 1-3-3 mm long, 1-3-2 x ovary, spreading; stigmas + broadly capitate. Capsule 2-5-4 x 1-5-2-2 mm, ellipsoid or ovoid- ellipsoid, equalling sepals. Seeds 0-7—0-8 mm long, ecarinate, testa finely scalar- iform. On dry slopes and in rocky places in the paramo; 2350-3800 m. Ecuador (Cotopaxi, Chimborazo, Canar, Azuay, Loja), Peru (Cajamarca, La Libertad). Map 8. H. loxense is closely related to H. quitense, which although overlapping it in distribution is apparently ecologically distinct. In addition, H. loxense has narrower and/or smaller leaves and usually smaller flowers. H. loxense comprises two populations of which subsp. aequatoriale is nearer to H. ne TEeeea es FALE LASY ii a 43 = sed) oO sed) 44 NORMAN K. B. ROBSON [ 80 Map 8 15. H. myricariifolium @; 18. H. loxense: a. subsp. aequatoriale ®, b. subsp. loxense O. quitense. The most primitive form of this subspecies is from Mt. Coraz6n (Cotopaxi), where it is geographically isolated from the rest (Chimborazo to Loja). It is always erect and has incurved-cucullate, relatively broad leaves. Subsp. loxense differs in having relatively narrow, plane leaves and a usually laxer and more spreading habit. Its most primitive form occurs in Loja, whence there is (i) a southern trend (with disjunction) to Peru (Cajamarca, La Libertad) in which the leaves become smaller and narrower, and (ii) a northern trend (Azuay, Cafar, Chimborazo) towards prostrate weak stems and somewhat smaller leaves. The occurrence of intermediates in Loja prevents the recognition of these taxa as species. 18a. Hypericum loxense subsp. aequatoriale (R. Keller) N. Robson, stat. nov. H. loxense var. aequatoriale R. Keller in Bot. Jahrb. 42: 129 (1908), in Bull. Herb. Boissier II, 8: 181 (1908). Type: Ecuador. Cafiar, between Azogues and Péaramo de Huairacaja, 2700-3000 m, ix—x.1905? (fl & fr), Lehmann 4795 (B+, holotype, F!, photograph; F!, K!, US!, isotypes). Leaves usually erect and + densely imbricate; lamina 2-5—5-2 x 1-2 mm, lanceolate or rarely oblong to broadly triangular-ovate, plane to subsaccate, chartaceous to subcoriaceous, apex acute to obtuse, base + broadly cuneate to truncate, midrib rarely prominent beneath. Inflorescence branches 1-—3-flowered. Sepals 2-3 x 0-5—1-1 mm, narrowly oblong to lanceolate, acute to subacute. THE GENUS HYPERICUM L. Ecuador (Cotopaxi, Chimborazo, Canar, Azuay, Loja). ECUADOR. Cotopaxi: Coraz6n, 3200 m, 22.vi.1876, André K. 436 (K). Chim- borazo: Péaramo de Cachenco, c. 3800 m, 27.vii.1959 (fl), Barclay & Juajibioy 8258 (NY). Cafiar: near Pimo, 3060-3120 m, 9.vii.1945 (fl), Camp E-4136 (BM, NY); Tipococha, 3400 m, 9.vii.1939 (fl & fr), Penland 994 (F). Azuay: Benos, SW. of Cuenca, c. 2700 m, 12.v.1974 (fl), Harling & Anderson 14561 (GB); San Miguel to Cuenca, 25 km, 3000 m, 15.ix.1969, B. & C. Maguire 61712 (BM, NY). Loja: between Loja and Cuenca, c. 20 km N. of Saraguro, 3080 m, 21.x.1980 (fl & fr), Croat 50854 (BM, MO). 18b. Hypericum loxense subsp. loxense Leaves spreading or more rarely erect and subimbricate; lamina 4-10 x 0-6-3 mm, narrowly oblong or narrowly elliptic or rarely oblanceolate to linear, plane to incurved, chartaceous, apex acute to rarely subacute, base narrowly cuneate to parallel-sided, midrib often prominent beneath. Inflorescence branches usually dichasial or mixed, more rarely 1-flowered. Sepals 2—4-5(5-3) x 0-5—1-2(1-6) mm, narrowly elliptic to oblong, acute. Ecuador (Chimborazo, Cafiar, southern Azuay, Loja), Peru (Cajamarca, La Liber- tad). ECUADOR. Azuay: Paramo Portete, 3600 m, 13.iii.1953 (fl), Prescott 793 (NY). Chimborazo: Urbina, towards Mt. Chimborazo, c. 3700 m, 27.vii.1939 (fl & fr), Asplund 7913 (S). Cafiar: near E] Tambo (c. 69 km S. of Sibombé), 2850-3000 m, 5.vii.1945 (fr), Camp E-4000 (BM, NY). Loja: Loja, S. of the town, c. 2350 m, 4.x.1955 (fl & fr), Asplund 17919 (S); between Cuenca and Loja, near Una, 1865 (fi & fr), Jameson 127 (K, W). PERU. Cajamarca: above Sunchubamba, 3680 m, 8.vi.1957 (fl), Ellenberg 1862 (U). La Libertad: Playapampa, c. 2700 m, 15—24.vi.1923 (fl), Macbride 4525 (NY). The disjunct populations in Chimborazo and northern Azuay differ in their decum- bent to prostrate habit, whereas the collection from Cajamarca differs in having oblanceolate subacute leaves and somewhat larger flowers than normal. 19. Hypericum sabiniforme Trev. Hyper. Animad.: 15 (1861); Triana & Planchon in Annis Sci. nat. (Bot.) IV, 18: 296 (1862) [‘sabinaeforme’|. Type: Colombia, Boyaca, Prov. de Tunja, Soata, 1300 m, iv.1843 (fl), Linden 1328 (BHU?, holotype; BM!, GH!, K!, P!, W!, isotypes). Hypericum thymifolium sensu Turcz. in Bull. Soc. Nat. Moscou 31 (1): 386 (1858), non Banks & Solander (1794) nec Kunth (1822). Hypericum hartwegii sensu Triana & Planchon, tom. cit.: 295. (1862) pro parte, excl. typum; R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 214 (1893) pro parte, excl. typum, in op. cit., 2nd ed., 21: 181 (1925) pro parte quoad specim. Colomb. Hypericum lindenii R. Keller in Bull. Herb. Boissier I, 8: 183 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925); Gleason in Bull. Torrey bot. Club 56: 102 (1929). Type: Colombia (as for H. sabiniforme), Linden 1328 (W!, holotype; BM!, GH!, K!, P!, isotypes). Hypericum trianae N. Robson in Cleef, Veg. Param. Colomb. Cord. Or.: 164, 305 (1981), nomen. Shrub 0-3—0-5 m tall, erect, bushy, with branches erect, + strict, lateral or rarely pseudo-dichotomous. Stems orange-brown, 4-lined when young, eventually terete, cortex exfoliating in strips; internodes 1-5-4 mm long. Leaves sessile or with pseudopetiole up to 1 mm long, outcurving to appressed, loosely to densely imbricate, sometimes markedly tetrastichous, deciduous at or sometimes slightly above the base; lamina 1-5-7 Xx 0-4-3 mm, ovate or elliptic to oblanceolate or narrowly oblong, + cucullate, incurved, with margin narrowly hyaline, midrib impressed beneath, concolorous, glaucous, coriaceous, markedly papillose or sub- papillose above, less so beneath; apex obtuse to rounded, base cuneate- 45 46 NORMAN K. B. ROBSON pseudopetiolate to angustate or parallel, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense, large, not or slightly promin- ent. Inflorescence 1-flowered, terminal and on lateral branches; pedicel 1-2 mm long; upper leaves transitional or foliose. Flowers 5—20 mm in diam., stellate. Sepals 2-5 x 0-5-3 mm, ovate-oblong to oblong, obtuse to rounded, plane or + cucullate, margin narrowly hyaline; veins 3—9, unbranched, with midrib prominent beneath or not; glands mostly linear to mostly punctiform. Petals bright (?) yellow, 4-11 x 2-6 mm, 2-2-5 xX sepals, oblong-obovate; apiculus acute; glands distally interupted to wholly punctiform. Stamens 60-125, longest 2-5—4-5 mm long, 0-4—0-65 x petals. Ovary 1-5-2 x 0-8-1 mm, ovoid to ellipsoid-subglobose; styles 2, 1-5—5 mm long, 1-2 X ovary, + spreading-outcurved; stigmas small. Capsule 3-5 X 2-5 mm, ovoid-cylindric to ellipsoid, exceeding sepals. Seeds not seen. In dry paramo; 2000-3870 m. Colombia (Boyaca). Map 9. COLOMBIA. Boyaca: Municipio de Susacén, paramo de Guantiva, 3140 m, 8.viii.1958 (fl), Jaramillo & Hernandez 965 (US); NW. of Belen, Quebrada Minas, Hoya Clla. Larga, 3870 m, 25.ii.1972 (fl), Cleef 1807 (COL; U); 1 km E. of Vado Hondo, Valle de Rio Cusiara, 2915 m, 31.iii. 1973 (fl). Cleef, Cuatrecasas & Jaramillo 9236B (COL); Paramo de Gina, 3200 m, 17.vii.1940 (fl), Cuatrecasas & Garcia Barriga. 9790 (F); Andes de Bogota, 2500-3000 m, (fl), Triana s.n. (BM). The type of H. sabiniforme, which superficially resembles a species of the African ericaceous genera Philippia or Blaeria, represents the extreme form of an eastward a9 | Map9 19. H. sabiniforme A; 22. H. hartwegii O; 23. H. maguirei @; 25. H. gleasonii @. THE GENUS HYPERICUM L. cline. At the western end of the cline is a form with much larger ovate to elliptic pseudopetiolate leaves, markedly papillose above, and much larger flowers. Varia- tion along the cline, however, appears to be continuous. The western form is near the broad-leaved form of H. thuyoides, but differs in having thicker, glaucous leaves with undulate to papillose epidermis. Wood 3682 (Cundinamarca, between Neusa and San Cayetano) is intermediate between 13 H. thuyoides ‘genistoides’ (see p. 38) and 19 H. sabiniforme. 20. Hypericum laricifolium Juss. in Annls Mus. Hist. nat. Paris 3: 160, t.16 f.1 (1804); Choisy, Prodr. monogr. fam. Hypéric.: 58 (1821), in DC., Prodr. syst. nat. regni veg. 1: 553 (1824); R. Keller in Bull. Herb. Boissier 6: 257 (1898), in op. cit. II, 8: 176 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925); Kunth in Reprium Spec. nov. Regni veg. Beih. 43: 483 (1927). Type: Peru, without precise locality or date (fl), J. de Jussieu* in Herb. A. L. de Jussieu 11800 (P — JUSS!, holotype; C!, F!, P!, isotypes). H. acerosum Kunth in Humboldt, Bonpland & Kunth, Nova Gen. et Sp. Pl. 5: 187, t. 457 (1822). Type: Ecuador, Azuay, ‘in Andium Assuay Quitensium’, 3240 m, vii. 1805 (fl), Humboldt & Bonpland (P-HUM!, holotype; P!, isotype). Brathys laricifolia (Juss.) Spach, Hist. nat. vég. Phan. 5: 366 (1836). B. acerosa (Kunth) Spach, Hist. nat. vég. Phan. 5: 449. 1836, in Annls Sci. nat. (Bot.) II, 5: 366 (1836). Hypericum laricifolium var. acerosum (Kunth) Weddell, Chloris Andina 2: 271 (1857). H. platypetalum Turcz. in Bull. Soc. Nat. Moscou. 31 (i): 388 (1858), non (Webb) Steudel (1840). Type: Ecuador, Pichincha, Quito, 1836? (fl), Jameson 60 (LE, holotype; BM!, K!, W!, isotypes). H. racemulosum Turcz. in Bull. Soc. Nat. Moscou. 36 (1): 580 (1863). Type: Peru. Amazonas, Chachapoyas, 1835 (fl), Mathews 1611 (LE, holotype; BM!, KW!, P!, isotypes). H. laricoides Gleason in Bull. Torrey bot. Club 56: 104. (1929). Type: Venezuela, Mérida, Paéramo de Timotes, Péramos de Chachapos, 3800—4200 m, 24.i.1928 (fi), Pittier 12713 (US!, holotype; F!, K!, MO!, isotypes). Icones: Cuatrecasas in Trab. Mus. Nac. Cienc. Nat. Madrid (Bot.), no. 27: t. 29 (1934); Humboldt, Bonpland & Kunth, tom. cit.: t. 457 (1822); Fig. 8A. Shrub or small tree (0-1—)0-3—3(—6) m tall, bushy or lax and spreading, with branches erect to spreading, lateral, short (flowering) and long, narrowly to broadly conical (extension). Stems 4-lined when young, eventually terete, blackish-brown, without epidermal emergences, cortex exfoliating in strips or irregular flakes; internodes 1-5-5 mm long. Leaves sessile, spreading or erect, sometimes imbricate, not markedly tetrastichous, persistent; lamina (2—)3—6(-8) x 0-3-2-2 mm, narrowly elliptic or narrowly elliptic-oblong to acerose, conduplicate to incurved, with midrib impressed beneath, margin narrowly hyaline and plane or subundulate, not glaucous, chartaceous to coriaceous; apex subacute to obtuse, cucullate, base narrowly cuneate to angustate, loosely clasping or not, pairs united to form narrow interfoliar ridge; basal vein 1, not or obscurely branched; laminar glands dense, scattered or in 1 row, impressed, usually visible beneath only. Inflorescence 1- flowered, on short lateral branches, rarely branching pseudo-dichotomously; pedicel 1-5-5 mm long; upper leaves foliose. Flowers 15—25(-—30) mm in diam., stellate. Sepals 3-7 x 0-7—2-2 mm, oblong or elliptic to narrowly ovate or lanceolate, subacute to acute, cucullate, margin hyaline; veins 7-9, not or scarcely branched, midrib distally incrassate; glands linear, distally punctiform. Petals bright golden yellow to orange-yellow, 7-14)-22) x 4-10 mm, 2-3 x sepals, very obliquely, obovate; apiculus obtuse; glands striiform and punctiform. Stamens 70-110, longest 4—7 mm long, c. 0-5 X petals. Ovary 2-3-5 x 0-8-1-5 mm, ellipsoid; styles 3, 4-9 mm long, 2—4-5 x ovary, spreading; stigmas scarcely to narrowly capitate. Capsule 4—7 x * The only specimen in Herb. de Jussieu that could be type material is labelled ‘Perou, 1780, Dombey.’ 48 NORMAN K. B. ROBSON YY Yirr Soy ; CO Yi WW VF f Fig.8 A. H. laricifolium: (a) habit; (b) stem with leaves (Ecuador); (c) stem with leaves (Venezuela); (d) leaf; (e) sepal; (f) petals; (g) stamens (partly cut away) and ovary; (h) capsule. B. H. martense: (i) habit; (j) stem with leaves; (k) leaf; (1) sepal; (m) petal; (n) capsule (a, i x ¥2; b,c, j X 2; d—h, k—-n x 4). A. Luteyn 6751 except (c) Luteyn 7583; B. Seifriz 421. 3-4 mm, ovoid to subglobose, equalling or exceeding sepals. Seeds 0-8—1 mm long, ecarinate; testa finely scalariform. In the subpéramo and pdramo, in open or more usually sheltered habitats, well- drained or not; 2200—4300 m. From western Venezuela (Lara, Trujillo, Mérida, Zulia) along the Cordilleras THE GENUS HYPERICUM L. Central and Oriental of Colombia and Ecuador to central Peru (Hudnuco, Ancash). Map 10. COLOMBIA. Antioquia: between Medellin and Retiro, around Las Palmas, c. 2800 m, 16.x.1947 (st), Gutiérrez, Barkley & Rosano 24(F). Boyaca: Paramo La Rusia, NNW. of Duitama, 3550 m, 11.ix.1969 (fl & fr), Cuatrecasas & Rodriguez 27730 (BM, COL, US). Caldas: Rio Otin, Nevado de Santa Isabel, Bagas de la Laguna Negra, 3800-3750 m, 24.xi.1946, Cuatrecasas 23160 (F, P, U). Cauca: Popayan, near Pitayo, 1841-1843 (fl), Hartweg 925 (BM, K, P, W); Paramo del Puracé, S. of the Volcan, San Francisco, 3450-3400 m, 23.vii.1943 (fl), Cuatrecasas 14571 (COL, F, P). César: Municipio de Valledupar, Corregimiento de Manaure, 2800 m, n.d. (fl), Rodriguez s.n. (COL). Chocé: Macigo de Tatama, valle El Encanto, 3420 m, 13.11.1983 (fl), Cleef & Salamanca s.n. (COL). Cundinamarca: Paramo de Guasca, vicinity of Bogota, 3150 m, 23.viii.1959 (fl), B. & C. Maguire 44029 (BM, NY); Paramo de Palacio, Lagunas de Buitrago, 3550 m, 27.iv.1973 (fl), Cleef 9577 B (COL). Magdalena: Sierra de Perija, plain between Cerro Venado and Cerro Avi6n, 3270-3350 m, 8.xi.1959 (fl), Cuatrecasas & Romero-Castaneda 25128 (COL). Narifo: Volcan de Cumbal, via Laguna Bolsa, 2740-2980 m, 11.ix.1944 (fl), Ewan 16139 (BM, S). Norte de Santandér: Ocania to Pamplona, Serra P1. , 2400-3300 m, 30.viii.1878 (fl), Kalbreyer 711 (K), Putumayo: Lagunda de la Cocha, Quebrada de Santa Lucia, 2850 m, 8.1.1941 (fl), Cuatrecasas 11824 (COL, F). Risaralda: Municipio de Pereira, vicinity of La Laguna de Ottn, 4000-4300 m, 3.ii.1980 (fl), Diaz, Valencia & Jaromillo 1738 (COL). Santandér: Paramo de Santurban, Laguna de Cunta, 3880 m, 21.1.1927 (fl), Killip & Smith 17962 (COL, F). Tolima: Paramo de Ruiz, 3200-3500 m, 16-17.xii.1917 (fl), Pennell 2999 (F, K, MO). Valle: Rio Map 10 20. H. laricifolium: typical form @, ‘laricoides’ O; 21. H. martense A. 49 50 NORMAN K. B. ROBSON Bugalagrande, Barragan, Paramo de Bavaya, Corrales, 3550-3400 m, 10.iv.1946 (fl), Cuatrecasas 20569 (BM, F, P, U, US). VENEZUELA. Lara: between Buenos Aires and Paramo de las Rosas, 2285 —3290 m, 11.11.1944 (fl), Steyermark 55501 (F). Mérida: Paramo de Mucuchies, x.1865 (fl), Moritz 1166 (BM, K, P, S); Sierra Nevada de Mérida, between La Canada del Espejo and La Canada de Los Anteojos, Torre de Befio, 4000-4130 m, 22.1.1973 (fl), Cuatrecasas, Ruiz-Terdn & Lépez-Figueras 28570 (BM, US); Paramo de Conejos, c. 8 km NE. of Mérida, 3800-4100 m, 27.xii.1967, Wessels Boer 2135 (NY, U, VEN). Tachira: Paramo de La Negra, 3000 m, 13.xi.1939 (fl), Burrus & Miiller 1019a (VEN). Trujillo: El Paramito, Jag6 to La Morita, hacia Tufiame, over 3000 m, vili.1958 (fl), Aristeguiéta & Medina 3488 (NY, VEN). Zulia: Zulia, 1983-1894 (fl), Mocquerys s.n. (NY, P). ECUADOR. Azuay: 54 km from Cuenca on Suscal road, 3350 m, 13.ix.1969 (fl), B. & C. Maguire 61696 (BM, NY). Bolivar: Cordillera Occidental, Gualicon Loma, 3240 m, 7.x.1943, Acosta Solis 6263 (F). Cafiar: Biblian to Canar, north of Biblian, c. 3400 m, 23.iv.1968 (fl), Harling, Storm & Strém 8655 (GB, NY). Carchi: Paramo El Angel, between El Angel and Tulcan, 3450-3500 m, 14.v.1973 (fl), Holm-Nielsen et al. 5298 (AAU, F, S). Chimborazo: Cerro Chiguaza, c. 3200 m, 24.ix.1968 (fl), Lugo 475 (GB, NY). Cotopaxi: Cotopaxi, SW. slope, 3700 m, 3.vii.1939 (fl), Asplund 7516 (K, P, S). El Oro: NE. Zaruma, Tioloma, Hac. Ambocas, 3100 m, 30.viii.1947, Espinosa 2034 (NY). Imbabura: Cotacache to Apuela, 21 km (Intec valley), 3300 m, 11.viii.1976 (fl), Dllgaard & Balslev 8687 (AAU, BM, MO, U). Loja: 10 km south of Loja, 2800 m, 1-3.viii.1959 (fl), Harling 6239 (S). Napo: Cerro Antisana, Los Corrales, near Papallacta, 3900 m, 21.vii.1960 (fl), Grubb et al. 218 (K, NY). Pichincha: Mount Guamani, 3600-3800 m, 15.ix.1939 (fl), Asplund 8757 (K, P, S). Tungurahua: between Ambato and Banos, 10.i.1981 (fl), D’Arcy 13975 (BM, MO). PERU. Amazonas: Chachapoyas, Cordillera Calla-Calla between Leimebamba and Balsas, 2900-3000 m, 21.x.1963 (fl & fr), Ferreyra & Acleto 15302 (MO, NY). Ancash: Cordillera Blanca, above Vicos, trail to Llacshacocha at Usacocha, 3350 m, 14. iii. 1964 (fl), Hutchinson & Wright 4392 (F, K, MO, P, S). Cajamarca: 28 km N. of Cajamarca towards Pedernal, 12 km W. to Hacienda Granja Porcon (SIPA), 3260 m, 6.vi.1963 (fl), D. & V. Ugent 5463 (GH, K, WIS). Huanuco: Mito, c. 2700 m, 8—22.ii.1922 (fl & fr), Macbride & Featherstone 1610 (F, S). La Libertad: Huama- cucho, Pullac to Parcay, 69 km, above Rio Chuagén, 3920 m, 7.viii.1964 (fl), Hutchinson, Wright & Straw 6188 (F, K, MO). Piura: Huancabamba, distr. Sondor, 2300 m, 21.vii.1975 (fl), Sagdastegui, Cabanillas & Dias 8214 (MO). San Martin: Huallaga, valley of Rio Apisoncho 30 km above Jucusbamba, 3300 m, 5.viii.1965 (fl), Hamilton & Holligan 1090 (K, S). H. laricifolium is closely related to H. thuyoides, and one collection from Cauca (Fernandez Pérez 7159) is intermediate in some characters. Nearly always, however, H. laricifolium can be distinguished by the narrow rounded leaves with impressed midrib and straight margin. H. laricifolium is very variable, but the variation appears to be continuous and not amenable to classification. From northern Ecuador and adjacent Colombia, where the leaves are relatively broad, there are clines north and south towards shorter, narrower, more densely imbricate leaves and smaller flowers with narrower, more acute sepals: (i) North-eastward along the Cordillera Central to Venezuela; the extreme form from Boyaca north-eastward looks very different, with young shoots densely clothed with appressed leaves (‘H. laricoides’). Both the typical form and ‘H. laricoides’ occur in Boyaca and some parts of Cundinamarca, the latter favouring damper habitats. In other parts of Cundinamarca (e.g. Paramo de Guasca) and in Caldas, Risaralda, and Tolima, however, there is a continuous series of intermediate forms that does not allow the recognition even of subspecies. (ii) Southward through Ecuador to central Peru (Hudnuco, Ancash) the leaves become gradually narrower and smaller, but not markedly imbricate, except for some populations from northern Peru. THE GENUS HYPERICUM L. 21. Hypericum martense N. Robson, sp. nov. H. laricifolio Juss. affinis, sed habitu multo humiliori caespitosa, ramificationi praecipue pseudo-dichotomo, foliis valde punctatis, sepalis valde nervatis, petalis plerumque minoribus, staminibus paucioribus, stylis brevioribus, capsula angus- tiora, differt. Type: Colombia, Magdalena, Sierra Nevada de Santa Marta, near source of Rio Yebosimeina, 3900 m, 25.v.1977 (fl), Starker-White & Alverson 586 (NY!, holotype; MO!, isotype!). Icon: Fig. 8B. Shrublet 0-08—0-16 m tall, forming clumps or cushions c. 15 mm in diam., with branches strict, mostly pseudo-dichotomous, or the outer decumbent but not rooting. Stems orange-brown, 4-lined and compressed when young, soon terete, cortex exfoliating irregularly; internodes 1-4 mm long. Leaves sessile, erect to closely imbricate, scarcely outcurving, not tetrastichous, persistent; lamina 4—7 x 0-6—0-8 mm, linear (upper) to acerose, incurved-canaliculate, cucullate, midrib not impressed, margin narrowly hyaline, concolorous, both surfaces dull, glaucous, subcoriaceous?; apex subacute to obtuse, base parallel, not clasping, pairs united to form very narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense, large, impressed beneath. Inflorescence 1-flowered, with pseudo-dichotomous bran- ches from node below; pedicel 3—4 mm long, slender; upper leaves not transitional. Flowers 8—10 mm in diam., obconic (?). Sepals 4—5 x 0-8—1-3 mm, narrowly oblong, acute, incurved above; veins 5, unbranched, all prominent; glands mostly linear but striiform to punctiform near hyaline margin. Petals 6-8 x 2 mmc. 1-5 X sepals, oblong-obovate; apiculus acute; glands few, punctiform. Stamens c. 25, longest 4—5 mm long, c. 0-65 X petals. Ovary c. 1 x 0-5 mm, ellipsoid; styles 3, 2—2-5 mm long, c. 2-2-5 X ovary, outcurving to erect; stigmas broadly capitate. Capsule 4-5 x 2-2-5 mm, narrowly ellipsoid, obtuse, exceeding sepals. Seeds 1-2 mm long, ecarinate; testa finely scalariform-reticulate. In rocky paramo, muddy places in high-rainfall areas; 3000-3900 m. Colombia (Magdalena); confined to the Sierra de Santa Marta. Map 10 (p. 49). COLOMBIA. Magdalena: Sierra de Santa Marta, SE. slopes, Hoya del Rio Donachui, Meollaca, c. 3400 m, 8.x.1959, Cuatrecasas & Romero Castaneda 24644 (US); above San Miguel, c. 3000 m, vii.1932, Seifriz 421 (US). H. martense is a high-altitude derivative of the ‘laricoides’ form of H. laricifolium. As a result of its dwarf, caespitose habit, however, the non-pseudo-dichotomous parts are much reduced, most of the clump being formed by pseudo-dichotomous (in- florescence) branching. The branches of a pseudo-dichotomy, however, usually bear lateral flowering branches. 22. Hypericum hartwegii Bentham Pl. Hartwegiana: 127 (1843) [‘hartwegi’]; Triana & Planchon in Anals Sci. nat. (Bot.) IV, 18: 295 (1862) pro parte, quoad typum; R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 214 (1893), in Bull. Herb. Boissier II, 8: 177 (1908), in Engler & Prantl, op. cit. 2nd ed. 21: 181 (1925) pro parte, excl. pl. Colomb. Type: Ecuador, Loja, mountains near Loja, Chiquiribamba, vii.1841 (fl), Hartweg 721 (K!, holotype; BM!, K!, P!, W!, isotypes). Shrub 0-3-0-9 m tall, erect, with branches strict to spreading, lateral, short (flower- ing) and long (extension) or rarely pseudo-dichotomous. Stems dark red-brown, 4-lined and ancipitous when young, eventually terete, without epidermal emerg- ences, cortex exfoliating in strips; internodes 5-7 mm long. Leaves sessile, erect to ascending, closely imbricate, tetrastichous, persistent; lamina 5-7 x 2-3 mm, elliptic to oblong or obovate, plane, apically subconcave, not carinate but with midrib + incrassate distally and subimpressed proximally beneath, margin narrowly hyaline, not glaucous, subcoriaceous; apex apiculate-obtuse to rounded, slightly cucullate, base cuneate to parallel, scarcely clasping, pairs united to form very narrow interfoliar ridge; basal vein 1, apparently unbranched; laminar glands quite 51 32 O HEE O Hh = NORMAN K. B. ROBSON dense, not impressed or prominent, visible on both sides. /nflorescence 1-flowered, terminal and on short lateral branches; pedicel 1-2 mm long; upper leaves foliose. Flowers 15—20 mm in diam., stellate. Sepals 4-6 x 1-8-2-3 mm, ovate to oblong- elliptic, acute to subacute; veins c. 5, unbranched (?), midrib distally incrassate; glands linear. Petals bright (?) yellow, 8-11 x 4-5 mm, 2-2-5 x sepals, obovate; apiculus acute; glands striiform and punctiform. Stamens 45-50, longest 6-7 mm long, c. 0-5 X petals. Ovary 2-2-5 x c.1-4 mm, ovoid-ellipsoid; styles 3, 5-6 mm long, c. 2-5 X ovary, Outcurving; stigmas small to subcapitate. Capsule and seeds not seen. Habitat and phenology not recorded. Ecuador (Loja). Known from only one collection. Map 9 (p. 46). ECUADOR. Loja: see type. H. hartwegii appears to be a very local relict species, being mostly closely related to H. lycopodioides (from east-central Colombia), which has spreading, incurved- conduplicate leaves with margin sometimes undulate, oblique-obovate petals, and more numerous stamens. See also 23 H. maguirei. 23. Hypericum maguirei N. Robson, sp. nov. H. hartwegii Bentham affinis, sed foliis minoribus, dense imbricatis vaide cucullatis costa media subtus impressa, floribus minoribus, differt. Type: Ecuador, Azuay, 5km past San Miguel on road to Cuenca, 3350m, 15.x.1969 (fl), B. & C. Maguire 61707 (BM!, holotype; NY!, isotype). Shrub 0-3-1 m tall, erect or spreading; branches strict or ascending, lateral, short (flowering) and long (extension), narrowly spiciform to pyramidal, sometimes secund. Stems dark red-brown, 4—6 lined when young, eventually terete, without epidermal emergences, cortex exfoliating in strips or irregular flakes; internodes 1-3 mm, long. Leaves sessile, suberect to erect, + densely imbricate, tetrastichous, persistent; lamina 2—4 x 0-4—-1-2 mm, + narrowly oblong to narrowly elliptic or linear, incurved, midrib impressed beneath, pale lucent green with margin narrowly hyaline and plane, not glaucous, coriaceous; apex obtuse to rounded, markedly cucullate, base narrowly cuneate to parallel-sided, not or scarcely clasping, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense or in one row beneath, not impressed or prominent, sparser above. /nflores- cence 1-flowered, terminal and on short lateral branches; pedicel 1-3 mm long; upper leaves foliose. Flowers (S—)8—15 mm in diam., stellate (to obconic ?). Sepals 2-5 xX 0-7-7-1 mm, acute, incurved-cucullate, margin hyaline; veins 3-5, unbran- ched, midrib prominent beneath; glands linear, distally punctiform. Petals golden to orange-yellow, 3-8 xX 1-5-4 mm, 1-5-2 xX sepals, obovate to oblong-obovate; apiculus subacute to obsolete; glands striiform and punctiform. Stamens 25-40, longest 2—6 mm long, 0-6—0-75 x petals. Ovary 1-2 x 0-5—1-2 mm, narrowly ovoid to ellipsoid; styles 3, 2-5 mm long, 1-5—2-5 X ovary, spreading-incurved; stigmas small. Capsule c. 3 X 2 mm, narrowly ovoid-ellipsoid, equalling sepals. Seeds not seen. On dry slopes with mostly low shrubs and ferns, and drier grassland; 2400-3350 m. Southern Ecuador (Azuay, Loja, El Oro). Map 9 (p. 46). ECUADOR. Azuay: Cuenca to Ona, Hacienda Pizhia, c. 2800 m, 23.ix.1955 (fl), Asplund 17766 (S). Azuay/Loja: Nudo del Cordillera Occidental y Cordillera Oriental entre Ofia y Rancho Ovejero, 2704—2800 m, 1-2.viii.1959 (fl), Barclay & Juajibioy 8452 (NY). El Oro: between Curtincapa and Guagra Uma, 13 km NE. of Curtincapa, 2850 m, 16.viii.1943 (fl), Steyermark 53912 (F). Loja: about Ofia, 2400 m, n.d. (fl), Jameson s.n. (K). Larger specimens of H. maguirei at first sight resemble a small version of H. lycopodioides, but differ from it in the sessile, markedly cucullate, imbricate leaves with midrib impressed beneath, as well as in size of parts. It is most closely related to H. hartwegii, which has larger, almost plane leaves and larger flowers. Although THE GENUS HYPERICUM L. these two populations are very close in distribution, they are morphologically so distinct that they should be treated as separate species. 24. Hypericum magniflorum Cuatrec. in Ciencia, Mex. 4: 64, f. 1. (1943); Steyermark in Fieldiana Bot. 28: 988 (1967). Type: Colombia, Norte de Santandér, Cordillera Oriental, Paramo de Tama, near La Cueva, 27.x.1941 (fl), Cuatrecasas, Schultes & Smith 12634 (COL, holotype; F!, GH!, P!, U!, isotypes). Icon: Fig. 9. Shrub 1-2-5 m tall, erect, with branches strict, mostly pseudo-dichotomous, some- times lateral. Stems yellow-brown, 4-lined when young, without epidermal emerg- ences, soon terete, cortex flaking irregularly; internodes 2-3 mm long. Leaves sessile, spreading abruptly from above base, + tetrastichous, becoming brown but ao Fig.9 H. magniflorum: (a) habit; (b) stem with leaves; (c) leaf; (d) sepal; (e) petals; (f) stamens (partly cut away) and ovary; (g) capsule (a, b x %; c—g x 2). (b) Cuatrecasas & Rodriguez 27789; others Steyermark 57378. 54 NORMAN K. B. ROBSON not deflexed, persistent until cortex is shed; lamina 10-18 x 4—8 mm, elliptic to oblanceolate-spathulate, plane, slightly cucullate, carinate, concolorous or paler beneath, not glaucous, chartaceous; apex acute to obtuse-apiculate, base angustate, sheathing, pairs united to form very narrow interfoliar ridge; basal veins 5—7, parallel or almost so, branching distally, with tertiary reticulation sometimes visible beneath; laminar glands dense above, sparse or absent beneath, slightly prominent. /nfloresc- ence 1-flowered, sometimes with several short flowering branches crowded together, usually with pseudo-dichotomous branches from Ist or 2nd node below; pedicel up to 5 mm long; upper leaves transitional. Flowers c. 35 mm in diam., stellate. Sepals 9-11 x 3-5-5 mm, broadly oblong to obovate-spathulate, acute to obtuse; veins 11-19, dichotomising distally, with midrib scarcely prominent beneath; glands punctiform in upper third. Petals rich yellow, 16-22 x 12-14 mm, c. 2 X sepals, obovate- oblanceolate; apiculus acute; glands punctiform distally. Stamens c. 80, longest 9-11 mm long, c. 0-5 x petals. Ovary 4-5 x 2-3 mm, broadly ellipsoid to subglobose; styles 3, S—9 mm long, 1-5-2 X ovary, erect, free, outcurved below apex; stigmas slightly enlarged to broadly capitate. Capsule c. 7 X 5mm, broadly ellipsoid, shorter than sepals. Seeds not seen. On rocky slopes of limestone outcrops; 3000—3800 m. Eastern Colombia (Boyaca, Santandér, Norte de Santandér) and adjacent Vene- zuela (Tachira). Map 11. COLOMBIA. Boyaca: Paramo de Chita, cabeceras de Rio Casanare, 3080 m, 19. vi. 1972 (fl), Cleef 4743 (BM, U). Norte de Santandér (see also type): Bucaraman- ga, Cumbre of Paramo Mortifo, Cucuta road, 3300 m, v.1948 (fl), Sandeman 6077 THE GENUS HYPERICUM L. (K). Santandér: Paramo de Santurban, c. 3000 m, 27.viii.1948 (fl & fr), Barkley & Araque 188155 (F, COL). VENEZUELA. Tachira: Paramo de Tama, near Colombo-Venezuelan bound- ary, 3045-3475 m, 15.vii.1944 (fl). Steyermark 57378 (F, NY); NNE. slopes of Pata de Judio, 2500-3100 m, 19.x.1978 (fi & fr), Luteyn 5945 (BM, VEN). Though clearly related to H. terrae-firmae, H. magniflorum is constantly 3-styled. The abrupt outward bend of the leaves is characteristic, as is the persistent leaf witha sheathing narrowed base. 25. Hypericum gleasonii N. Robson, sp. nov. H. magnifloro Cuatrec. affinis, sed foliis brevioribus, plusminusve imbricatis, inter- dum viscidis, supra basin deciduis, stylis brevioribus stigmatibus manifeste capitatis, differt. Type: Colombia, Santandér, Paramo de Almorzadero, c. 32-34 km S. of Chitaga, c. 3400 m, 12.v.1979 (fl), Luteyn 7653 (COL!, holotype; BM!, NY!, isotypes). H. platyphyllum Gleason in Bull. Torrey bot. Club 56: 106. (1929) pro parte, quoad Killip & Smith 15616 pro parte. Shrub 0-15-1-5 m tall, erect, with branches strict, mostly pseudo-dichotomous, sometimes lateral. Stems yellow-brown, 4-lined when young, eventually terete, without epidermal emergences, cortex flaking irregularly; internodes 2-3 mm long. Leaves sessile, outcurved-ascending to imbricate-erect, tetrastichous, deciduous above base before fading; lamina 7-10 x 3-6 mm, oblanceolate to obovate, incurved, not cucullate or carinate, concolorous, not glaucous, subcoriaceous to coriaceous; apex acute or apiculate to obtuse, base angustate, sheathing, pairs united (at least when young) to form narrow interfoliar ridge; basal veins 5—7, flabellate, branching distally (sometimes obscurely), reticulation obscure beneath; laminar glands dense, not or slightly prominent, visible beneath; marginal glands usually secreting viscous clear or white resin. Inflorescence 1-flowered usually with pseudo- dichotomous branches from 1st to 3rd node below and often with 1 or 2 lateral branches immediately below these, forming flower cluster when young; pedicel 3—6 mm long; upper leaves transitional. Flowers 20-40 mm in diam., stellate. Sepals 6-10 x 3-5 mm, elliptic to oblanceolate, acute to apiculate or obtuse; veins 9-15, sometimes dichotomising distally, with midrib scarcely prominent beneath; glands punctiform in upper third to two-thirds. Petals rich yellow, (always?) tinged red beneath, 12—20 x 10-11 mm, c. 2 X sepals, oblanceolate to obovate; apiculus obtuse to rounded; glands distally punctiform. Stamens 80-100, longest 6-10 mm long, c. 0:5 x petals. Ovary 3-4 x 2-3 mm, broadly ellipsoid to ovoid; styles 3, 4-6 mm long, 1-25-1-5 x ovary, free, suberect to spreading; stigmas broadly capitate. Capsule 5—6 x 5-6 mm, subglobose, shorter than sepals. Seeds not seen. In open paramo on fine talus or grassy slopes; 3200-3900 m. Colombia (Norte de Santandér, Santandér). Apparently confined to the Paramo del Almorzadero. Map 9 (p. 46). COLOMBIA. Norte de Santandér: Valley of rio Chitaga, 20 km SE. of Chitaga, 3475 m, 24.ix.1944 (fl & fr), St. John 207784 (COL, NY). Santandér: Péramo del Almorzadero, 3600-3800 m, 28.xi.1941 (fl), Cuatrecasas 13495 (COL, GH, NY, P); Péramo del Almorzadero, between Chitag4 and Cerrito, 3900 m, 31.xii.1959- 1.i1.1960 (fl), Barclay & Juajibioy 10389 (NY); Peralonso, Platera, 3400 m, 20.ix.1969 (fl), Cuatrecasas & Rodriguez 27828 (BM, COL, US); Peralonso, 3200 m, 19.vii.1940 (fr), Cuatrecasas & Garcia Barriga 9887 (COL). H. gleasonii is a relict species intermediate between H. magniflorum and H. mexicanum. It has larger flowers than H. mexicanum, and these are solitary, but it shares with that species the deciduous leaves with viscous secretions. In H. gleasonii the leaves are smaller than those of H. mexicanum (from Cundinamarca and Boyaca) 56 NORMAN K. B. ROBSON but with a comparable erect habit; the form of H. mexicanum that grows in the same area is a small-flowered shrublet with decumbent rooting stems. Killip & Smith 15616 (from the Paramo de Las Vegas), which Gleason (1929) placed in H. platyphyllum, is a puzzling collection. The BM and COL specimens belong to H. mexicanum, but the NY specimen has larger flowers and tends towards H. gleasonii. 26. Hypericum mexicanum L. Diss. Hyperic.: 5, f.2 (1776), Amoen. Acad. 8: 322, t.8, f.2 (1785); L. fil., Suppl. PI.: 345 (1781); Lam., Encyl. méth. (Bot.) 4: 169 (1797); Gleason in Bull. Torrey bot. Club 56: 103 (1929). Type: Colombia [Cundinamarca], Mutis in Herb. Linn. 943: 31 (LINN!, holotype; BM!, isotype). Hypericum ? mexicanum sensu Choisy, Prodr. monogr. fam. Hypéric.: 60 (1821). Hypericum mutisianum Kunth in Humboldt, Bonpland & Kunth, Nov. Gen. et Sp. PI. 5: 185 (1822), nom. illegit.; R. Kellerin Bull. Herb. Boissier I, 8: 178 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925); Knuth in Reprium Sp. nov. Regni veg., Beih. 43: 483 (1927). Type: Colombia, Cundinamarca, near Santa Fé de Bogota, 2457 m, 1805 (fl), Humboldt & Bonpland s.n. (P-HUM!, holotype; B-WILLD, isotype); Brathys mutisiana (Kunth) Spach, Hist. nat. vég. Phan. 5: 447 (1836); in Annls Sci. nat. (Bot.) II, 5: 366 (1836). Hypericum resinosum Bentham, Pl. Hartweg.: 165 (1845); R. Keller in Bull. Herb. Boissier I1, 8: 177 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925). Type: Colombia. Cundinamarca, 1841-1843 (fl), Hartweg 924 (K!, holo- type; BM!, P!, W!, isotypes). H. mexicanum var. intermedium Kuntze, Rev. Gen. pl. 1: 59 (1891). Type: Vene- zuela, Mérida, Silla de Caracas, Kuntze s.n. (NY, holotype). Shrub or shrublet 0-15—1-5(—2) m tall, erect or decumbent and sometimes rooting at the base, with branches strict, pseudo-dichotomous and sometimes lateral especially from short upper internodes. Stems yellow- to orange-brown, 4-lined when young, eventually terete, cortex flaking irregularly; internodes 2—8 mm long. Leaves sessile, outcurving-ascending to imbricate-erect, tetrastichous, deciduous above base usu- ally without withering; lamina 7-17 x 3-10 mm, elliptic or oblanceolate to obovate, plane to subconduplicate, not or scarcely cucullate, not carinate, concolorous, not glaucous, coriaceous; apex obtuse or rounded, base cuneate to angustate, + sheathing, pairs united to form narrow interfoliar ridge; basal veins 5—7, flabellate, branching and anastomosing distally, reticulation clearly to obscurely visible be- neath; laminar glands dense to sparse, usually visible and sometimes slightly prominent beneath; marginal glands secreting viscous clear or white resin. Inflores- cence 1—5(—13)-flowered, cymose, corymbose, often with 1-3 pairs of pseudo- dichotomous branches from 1-3 nodes close below; peduncle and pedicels 1-6 mm long; upper leaves transitional. Flowers 15—25 mm in diam., stellate. Sepals 5—9 x 2-4 mm, elliptic to oblanceolate-spathulate or obovate, acute to apiculate or obtuse; veins 5—9, sometimes dichotomising distally, with midrib scarcely prominent be- neath; glands punctiform distally. Petals golden yellow, (always?) tinged red be- neath, 9-12 x 4-9 mm, 1-3-2 X sepals, obovate-oblanceolate to obovate-oblong; apiculus obtuse to obsolete; glands punctiform distally. Stamens c. 50, longest 5—7 mm long, c. 0-5 X petals. Ovary 2-5-3 x 1-8-2 mm, broadly ellipsoid to ovoid; styles 3, 2-5-4 mm long, 1-1-2 x ovary, suberect; stigmas broadly capitate. Capsule 5—7 x 4—6 mm, broadly ellipsoid to subglobose, equalling or exceeding sepals. Seeds c. 0-7 mm long, ecarinate; testa finely scalariform. In open paramo on fine talus or grassy slopes or in scrub; (1700)2400—4200 m. Western Venezuela (Mérida, Tachira) and north-eastern to central Colombia (Norte de Santandér to Valle de Cauca). Map 12. COLOMBIA. Antioquia: entre Rio Negro y Sta Helena, 20650 m, vii.1944 (fi), Garcia Barriga. 11094 (COL). Boyaca: Sierra Nevada del Cocuy, Quebrada de San THE GENUS HYPERICUM L. Map 12 26. H. mexicanum ®; 28. H. prietoi O; 29. H. cassiopiforme @. Paulino proximo Alto Ritacuva, c. 3650, 3.v.1959 (fl), Barclay & Juajibioy 7525 (COL, MO, NY); Paramo de Gina, 3200 m, 17.vii.1940 (fl), Cuatrecasas & Garcia Barriga. 9783 (COL). Cundinamarca: Paramo de Guasca, 2840 m, 15.xii.1938 (fl), Balls 5700 (BM, COL, K); Sabana de Bogota, Municipio de Tenjo, Vereda de la Punta, 2800-2670 m, 29.i.1966 (fl & fr), Cuatrecasas & Jaramillo 26755 (BM, COL, US); Cerro de Monserrate, 3000 m, 10.iii.1966 (fl), Duque-Jaramillo 2877 (COL). Meta: Macizo de Sumapaz, alredores de la Laguna La Guitarra, 3370-3400 m, 5.vii.1981, Diaz P. 2485 (COL). Norte de Santandér: Paramo de Fontib6n, 2700 m, 27.vii.1940 (fl), Cuatrecasas & Garcia Barriga. 10088 (NY, US). Santandér: Paramo de Santurban, 3600 m, 27.viii.1948 (fl), Araque & Barkley 18s004 (COL, F). Valle: Rio Cali, Pichinde, 1700 m, vii. 1938 (fl), Duque-Jaramillo 4107 (COL). VENEZUELA. Mérida: Silla de Caracas, Kuntze s.n. (NY). Tachira: Paramo de Tama, 4000 m, 1937 (fl), Cardona 123 (US, VEN). H. mexicanum is a misnomer, as the type comes from Cundinamarca and the species as a whole is confined to Colombia and adjacent Venezuela. In attributing Mutis collections to Mexico, Linnaeus made the same mistake here as he did in the case of Vismia baccifera. H. mexicanum differs from H. gleasonii in having larger, more spreading leaves and smaller flowers, which are frequently in condensed cymes. The two species remain distinct where their distributions overlap in the Paramo del Almorzadero. Gleason (1929) confused the eastern form of H. mexicanum with H. platyphyllum, which has penninerved (not flabellate-nerved) leaves and smaller stigmas and is treated here as a subspecies of H. phellos. ay NORMAN K. B. ROBSON 27. Hypericum stuebelii Hieron. in Bot. Jahrb. 21: 321 (1895); R. Keller in Bull. Herb. Boissier I1, 8: 177 (1908), in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925) Type: Peru, Amazonas, Chachapoyas, between Pacasmayo and Moyobamba, near Mojon Cruz, 3300 m, iv.—vi.1875? (fl), Sttibel Peru 27 (B+; holotype; F!, GH!, NY!, photographs). Hypericum stuebelii Hieron. ex R. Keller in Bot. Jahrb. 42: 129 (1908). Type: as for H. stuebelii Hieron. Shrub 0-3-1 m tall, erect, with branches strict or ascending, pseudo-dichotomous. Stems reddish-brown, 4-lined when young, soon terete, cortex exfoliating in irregu- lar flakes; internode 3—4 mm long. Leaves sessile, closely imbricate and tetrastichous at first, gradually outcurving, deciduous above the base usually before fading; lamina 8-14 x 3-7 mm, broadly elliptic-ovate to elliptic-lanceolate, incurved-cucullate, midrib prominent distally beneath, margin scarcely distinct, concolorous, yellow- green, not glaucous, coriaceous; apex obtuse or apiculate to acute, base angustate, sheathing, pairs united to form rather deep (0-4-1 mm) interfoliar ridge; basal or near-basal veins 7—9, flabellate, not clearly branched, tertiary reticulum not visible; laminar glands dense, not prominent, visible above and sometimes beneath. Jn- florescence 1-flowered, usually with pseudo-dichotomous branches from node be- low; pedicel 4—6 mm long; upper leaves not transitional. Flowers 25—40 mm in diam., stellate. Sepals 10-12 x (2—)3—6 mm, elliptic-ovate to lanceolate, obtuse to acute, cucullate; veins 7-11, sometimes dichotomising distally, prominent beneath; glands linear, punctiform in distal third. Petals bright? yellow, 15S—20 x 5-10 mm, 1-5—1-7 x sepals, obovate; apiculus subacute to obtuse; glands linear, interrupted distally. Stamens 80-150, longest 7-10 mm long, c. 0-5 x petals. Ovary c. 4-5-5 X 2:5 mm, ellipsoid; styles 3-4, c. 7 mm long, c. 1-7 X ovary, free, suberect to outcurved; stigmas not or scarcely enlarged. Capsule c. 7-8 X 5 mm, broadly ellipsoid to subglobose, shorter than sepals. Seeds not seen. On open hillsides, (always ?) in damp or wet areas; 2400-3300 m. Northern Peru (Amazonas, San Martin), almost confined to a small region round Chachapoyas. Map 11 (p. 54). PERU. Amazonas: Chachapoyas, 1-5 km W. of Molinopampa, Jalca zone, 2400-2450 m, 18.vii.1962 (fl), Wurdack 1353 (F, GH, K, NY, P, S, US); La Jalca near Chachapoyas, 2700-3300 m, i.1930 (fl), Williams 7585 (F, NY); Chachapoyas, Pass of Piscohuafuna, 2700 m, viii.1938 (fl), Sandeman s.n. (K); Cordillera of Piscohuanuna between Chachapoyas and Moyobamba, xii.1868 (fl), Raimondi 2096 (NY). San Martin: Bagazan, 1835 (fl), Mathews 1609 (K). H. stuebelii appears to be a relict species with no ancestral relatives nearer than NE. Colombia (H. magniflorum, H. pimeleoides). Its leaves vary from broad, obtuse, and almost plane to narrow, acute, and incurved. 28. Hypericum prietoi N. Robson, sp. nov. H. stuebelii Hieron. affinis, sed ramis strictioribus, foliis minoribus densissimis manifeste quadrifariis, 1-nervis, parte distali obtuse carinata parte proximali subtus costa media impressa, floribus minoribus, differt. Type: Ecuador, Azuay, Eastern Cordillera, between Ona and the rio Yacuambi, 3000-3360 m, 10-19.ix.1945 (fr), Prieto P-302 (NY!, holotype). Shrub c. 0-3 m tall, erect; branches very strict, pseudo-dichotomous and lateral. Stems reddish-brown, 4—6-lined when young, eventually terete, cortex exfoliating in irregular flakes; internodes c. 2 mm long. Leaves sessile, very closely imbricate, markedly tetrastichous, laterally compressed distally, not becoming outcurved, deciduous above the base before fading; lamina 7-8 x 1-5-2 mm, oblanceolate, distally obtusely carinate and cucullate, midrib impressed proximally beneath, margin very narrowly hyaline, concolorous, lustrous yellow-green, not glaucous, coriaceous; apex acute, base parallel-sided, scarcely sheathing, pairs united to form shallow to rather deep (to 1-5 mm) interfoliar ridge; basal vein 1, unbranched; THE GENUS HYPERICUM L. laminar glands dense, not visible beneath. Inflorescence 1-flowered, with pseudo- dichotomous branches from node below; pedicel 2—2-5 mm long; upper leaves not transitional. Flowers stellate? Sepals 7-8 x 2 mm, lanceolate, acuminate, compress- ed-cucullate; veins 3—5, unbranched, prominent beneath; glands linear. Petals and stamens not seen. Ovary with styles 3, c. 4 mm long, suberect; stigma narrowly capitate. Capsule 5 X 2-5 mm, broadly ellipsoid, shorter than sepals. Seeds 1-2 mm long, ecarinate; testa finely scalariform. On the crest of the Cordillera; 3000-3360 m. Ecuador (Azuay). Map 12 (p. 57). ECUADOR. Azuay: Eastern Cordillera between Ona and the rio Yacuambi, 3000-3360 m, 10—19.ix.1945 (fr), Prieto P-302 (NY). H. prietoi, which has been collected once only, resembles H. stuebelii but has smaller, laterally compressed, and markedly tetrastichous leaves and stricter branches, some of which are lateral. 29. Hypericum cassiopiforme N. Robson, sp. nov. H. stuebelii Hieron. affinis, sed ramis basalibus lateralibusque strictioribus, foliis minoribus, oblanceolatis vel obovato-oblanceolatis, floribus minoribus, staminis paucioribus, stylis brevioribus, differt. Type: Peru, Amazonas, Chachapoyas, 1-5 km W. of Molinopampa, 2400-2450 m, 18.vii.1962 (fl), Wurdack 1355 (US!, holotype). Shrub c. 0-3 m tall, erect, with branches strict, basal and lateral. Stems reddish- brown, 4-lined when young, eventually terete, cortex exfoliating in irregular flakes; internodes 2—4 mm long. Leaves sessile, imbricate-tetrastichous but not usually laterally compressed distally, becoming outcurved, deciduous above the base before fading; lamina 7-8 x 2-5-3 mm, oblanceolate, incurved-cucullate but scarcely carinate, midrib slightly prominent to scarcely impressed beneath, margin narrowly, hyaline otherwise concolorous, lustrous yellowish-green, not glaucous, sub- coriaceous; apex subacute, base cuneate, broadening below, scarcely sheathing, pairs united to form narrow interfoliar ridge; basal vein 1 with 2 obscure near-basal branches; laminar glands dense, more visible above than beneath. Inflorescence 1-flowered; pedicel c. 3-5 mm long; upper leaves not transitional. Flowers c. 25 mm in diam., stellate. Sepals 8 x 1-8—2-3 mm, elliptic to oblong, acute, incurved-cucullate, reddish; veins 5, unbranched, prominent beneath; glands linear, distally punctiform. Petals bright yellow tinged red, c. 14 x 9mm, c. 1-7 X sepals, obovate-oblanceolate; apiculus acute; glands linear. Stamens c. 50, longest c. 10-11 mm long, c. 0-5 x petals. Ovary c. 4 X 1:7 mm, ellipsoid; styles 3, c. 5 mm long, c. 1-2 X ovary, sub- erect ?; stigmas narrow. Capsule and seeds not seen. In the Jalca zone; 2400-2450 m. Peru (Amazonas). Map 12 (p. 57). PERU. Amazonas: 1—5 km W. of Molinopampa, 2400-2450 m, 18.vii.1962 (fl), Wurdack 1355 (US). H. cassiopiforme, like H. prietoi, is based on a single collection and appears to be related directly to H. stuebelii. It is smaller, with smaller leaves and flowers and more slender stems, and is intermediate in form between H. stuebelii and H. decandrum. It differs from the latter by the broader leaves, the larger flowers with broader sepals and the longer styles with narrow stigmas. 30. Hypericum decandrum Turcz. in Bull. Soc. Nat. Moscou 31 (1): 389 (1858). Type: Ecuador [Pichincha (?)], Quito, [18657] (fl), Jameson 62 (LE, holotype; BM!, K!, W!, isotypes). Hypericum gnidioides var. polytrichoides R. Keller in Bull. Herb. Boissier 11, 8: 183 (1908). Type: Ecuador, Chimborazo, in pascuis montis Titaicin, 3900 m, xi.1859 (fr), Spruce 5599 (G!, holotype, F!, photograph; BM!, C!, GH!, K!, NY!, P!, W!, isotypes). \ Y ag NORMAN K. B. ROBSON Icon: Fig. 10. Shrub or shrublet (or perennial herb?) 0-1—0-6 m tall, erect or decumbent and rooting with branches strict, basal (from decumbent and rooting portion) and pseudo- dichotomous or sometimes lateral. Stems reddish- to yellowish-brown, 4-6-lined when young, soon 2-lined, eventually terete, cortex exfoliating in irregular flakes; internodes 1—5 mm long. Leaves sessile, erect and subimbricate to narrowly spread- ing or slightly outcurving, deciduous above the base before fading; lamina 4-15 x 0-7-3 mm, + narrowly oblanceolate to narrowly oblong, + incurved-cucullate, midrib distally prominent otherwise impressed or plane beneath, concolorous or with margin rather narrowly hyaline, lustrous pale green, not glaucous, sub- coriaceous; apex acute, base angustate, broadening below, not or scarcely sheathing, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched or with 1-2 pairs of obscure lateral branches; laminar glands dense, visible above only, or obscure. Inflorescence 1-flowered, with pseudo-dichotomous branches from node below; pedicel 2-5-7 mm long; upper leaves not transitional. Flowers 6-20 mm in diam., stellate. Sepals 3-5—9 x 1-3 mm, narrowly elliptic to narrowly oblong, acute, compressed-cucullate, green (not reddish); veins 5-7, unbranched, sometimes becoming prominent; glands linear, punctiform in distal third to half. Petals bright yellow tinged red (? always), 5—9 x 2-3 mm, 0-9-1-2 x sepals, obovate-oblong to oblanceolate-oblong; apiculus acute; glands absent. Stamens (5)10—45, longest 2-5-4 mm long, c. 0:5 X petals. Ovary 2-5-3 xX 1-1-5 mm, ellipsoid; styles 3(4), 1-5-3 mm long, c. 0-5-1 X ovary, divergent; stigma narrowly to broadly capitate. Capsule 3-5-5 X 2-3 mm, ellipsoid to cylindric-ellipsoid, shorter than sepals. Seeds 0-7—0-8 mm long, ecarinate; testa finely scalariform. In grassy or shrubby slopes of the paramo and subparamo; (2200 ?) 2700-3930 m. Ecuador (in scattered localities from Carchi to Loja), Peru (Amazonas). Map 11 (p. 54). ECUADOR. Azuay: Loja to Cuenca, between Km 113 and 124, 2700-3000 m, 26.ix.1969 (fl & fr), B. & C. Maguire 44320 (BM, NY). Caniar: Rivera NE. of Pindilig, 2900 m, 13.xii.1980 (fi & fr), Holm-Nielsen, Jaramillo & Coello 29222 (AAU). Carchi: Nudo de Boliche, Voladero, 3800 m, 12.vi.1939 (fl & fr), Penland 900 (F, GH, NY). Chimborazo: Mt. Titaictin, 3900 m, xi.1858 (fl), Spruce 5599 (BM, C, GH, K, NY, P, W). Loja: N. of Saraguro, 3045 m, 4.viii.1943 (fr), Steyermark 53743 (F, NY). Morona-Santiago: Paramo de Hatanga, Km 36 on road Sigsig- Gualaquiza, E. of pass, 3300 m, 14.xii.1980 (fl), Holm-Nielsen, Jaramillo & Coello 29386 (AAU). Napo: Valle de Chalupas near junction of Rio Chalupas and Rio Aguas Buenas, 3200 m, 20.i.1980 (fl), Holm-Nielsen 21014 (AAU). Pichincha?: Quito, 1865? (fl), Jameson 62 (BM, K, W). Tungurahua: Cordillera de Llanganates, near junction of rio Golpe and rio Sangarinas (Desaguadero), 26.xi.1939 (fl & fr), Asplund 9989 (S). PERU. Amazonas: Cerro San Marino, Amaybamba, 3100 m, iii.1919 (fr), Bues 567 (US). H. decandrum is most frequent towards the south of Ecuador and absent or rare in some central parts of that country. It is related to H. cassiopiforme but differs in the longer, often narrower, less closely imbricate leaves and the smaller flowers with shorter, capitate styles. It is quite variable, the most reduced, weaker-stemmed form from the extreme south of its range being considerably different from the sturdy northern plants. 31. Hypericum pimeleoides Planchon & Linden ex Triana & Planchon in Annls Sci. nat. (Bot.) IV, 18: 299 (1862); Gleason in Bull. Torrey bot. Club 56: 103 (1929). Type: Colombia, Norte de Santandér, Prov. de Ocafia, Cerro Pelado, 2270 m, ii. 1846-1852 (fl), Schlim 443 (sphalm. 243) (COL?, holotype; BM!, K!, KW!, P!, US!, isotypes). Hypericum caracasanum var. ocanense R. Keller in Bull. Herb. Boissier I, 8: 183 (1908). Type: as for H. pimeleoides (G!, holotype). THE GENUS HYPERICUM L. 61 Fig. 10 H. decandrum: (a) habit, large form; (b) habit, small form; (c) stem with leaves; (d) leaf; (e) sepal; (f) petal; (g) stamens (partly cut away) and ovary; (h) capsule (a, b x 4%; c—h x 5). (b) Espinosa 923; others B. & C. Maguire 61708. Shrub 0-4-1-5 m tall, erect, with branches strict, pseudo-dichotomous and occa- sionally lateral. Stems yellow- to orange-brown, 4-lined and compressed when young, the subfoliar ridge broad, soon terete, cortex exfoliating in strips; internodes 3-6 mm long. Leaves sessile or with broad petiole to 1-5(—2) mm long, densely imbricate and scarcely outcurving to laxly imbricate and narrowly spreading, + markedly tetrastichous, deciduous above petiole without fading; lamina 10-24 x 3-5—10-5 mm, broadly to narrowly elliptic, plane, not or scarcely cucullate, midrib not prominent beneath, margin not distinct or rarely indurated, concolorous, glaucous, coriaceous; apex acute or subapiculate to obtuse, base cuneate to angus- tate, sheathing or not, pairs united to form narrow interfoliar ridge; basal or near-basal veins (3)5—9, flabellate, branching and often visibly anastomosing distal- ly, tertiary reticulation obscure or apparently absent; laminar glands dense, often NORMAN K. B. ROBSON obscure beneath. /nflorescence 1-flowered, with pseudo-dichotomous branches from one or two nodes below; pedicel 4—12 mm long, incrassate upwards; upper leaves not transitional. Flowers 18-35 mm in diam., stellate to subcyathiform. Sepals (7)10—14 x (2)3-8 mm, elliptic-subcircular to elliptic-oblong or rarely oblong, acute to subacute; veins (5)7—11, dichotomising distally, with midrib not or slightly promin- ent; glands linear, distally punctiform. Petals bright yellow, 10-22 x 4-10 mm, c. 1-3-1-5 X sepals, obovate; apiculus acute; glands linear, distally punctiform. Sta- mens c. 40-90, longest 5—10 mm long, c. 0-5 x petals. Ovary (2-5)3—4 x 2-2-5 mm, ovoid-ellipsoid to subglobose; styles 3—4, (2)3—7 mm long, 1—1-75 x ovary, outcurv- ing to diverging, stout; stigmas scarcely to broadly capitate. Capsule 5—8 x 4-6 mm, cylindric-ellipsoid, equalling sepals. Seeds c. 1-2 mm long, ecarinate; testa finely scalariform. In open paramo; 3000-4300 m. Colombia (César/Norte de Santandér to Boyaca). In the Cordillera Oriental from Cerro de Oroque to Vado Hondo. Map 13 (excluding Cundinamarca records, which belong to 55 H. ruscoides). COLOMBIA. Arauca: Sierra Nevada del Cocuy, Quebrada El Playén, Agua Tendida, 3100 m, 10.vi.1973 (fl), Cleef 10174 (U). Boyaca: Sierra Nevada del Cocuy, near El Play6én, c. 3800 m, 10.ix.1957 (fl), Grubb, Curry & Fernandez-Perez 777 (COL, K, US); Paramo de Chita, cabeceras de Rio Casanare, Km 93 de la carretera a La Punta, 3080 m, 19.vii.1972 (fl), Cleef 4744 (BM, U); Pea de Arnical, N. de Vado Hondo, 3600 m, 6.iv.1973 (fl), Cleef 9418 (U). César/Norte de Santandér: 20 kmalS. Map 13 31. H. pimeleoides @; 32. H. magdalenicum A; 33. H. valleanum A; 34. H. sprucei @; 37. H. wurdackii O. THE GENUS HYPERICUM L. de Abrego, Las Jurisdicciones (Cerro de Oroque), 3700-3900 m, 19—21.v.1969 (fl & fr), Garcia-Barriga & Jaramillo 19744 (COL). Norte de Santandér: Paramo de Las Jurisdicciones, SW. of Ocania, 3000-3500 m, 8.vii.1984 (fl), Wood 4526 (K). H. pimelioides shows a continuous morphological trend southwards from a tall form in the Cerro de Oroque with stout branches, large, sessile, sheathing leaves, and large flowers with long styles and scarcely broadened stigmas, to a relatively low form in the Vado Hondo area with slender branches, small, petiolate, not sheathing leaves, and small flowers with short styles and broad stigmas. In the Cerro de Oroque population the large sessile or subsessile leaves have up to 9 free basal veins and a thin margin, whereas in the Vado Hondo population the smaller petiolate leaves have only one vein (the midrib); but 1—2 pairs of midrib branches originate near the base, so that the venation is reduced-flabellate. This is in contrast to the leaf venation of plants in the Sierra Nevada del Cocuy and elsewhere in Boyaca that belong to H. cardonae. These have rather similar leaves, but they are relatively narrower and less crowded, the margin is not or only slightly indurated, and the midrib branches, where visible, arise at least one third of the distance from the base. H. pimeleoides appears to be related to H. magniflorum and H. stuebelii because of its primitively broad-based leaves with flabellate venation and large flowers with 3-5 styles. It is the basic species to which the remainder of sect. Brathys is related, respectively through 32 H. magdalenicum and 49 H. jaramilloi. 32. Hypericum magdalenicum N. Robson, sp. nov. H. pimeleoidei Planchon & Linden ex Triana & Planchon affinis, sed foliis semper sessilibus angustioribus basi haud vel vix vaginatis venatione pinnata, sepalis plerum- que angustioribus, ovario angustiori, stigmatibus haud vel vix capitatis, differt. Type: Colombia, Magdalena, Sierra Nevada de Santa Marta, quebrada from Laguna Rio Frio, 3250 m, 28.vii.1972 (fl), Kirkbride & Forero 1748 (COL!, holotype; NY!, isotype). Shrub (0-2—-)0-7-1-5 m tall, erect, with branches strict, pseudo-dichotomous and sometimes with 2—3 laterals closely beneath flower. Stems orange-brown, 4-lined and compressed when young, the subfoliar ridges broad, soon terete, cortex exfoliating irregularly; internodes 2—4 mm long. Leaves sessile, densely imbricate, erect or outcurving, markedly tetrastichous, deciduous above base without fading; lamina 8-18 x 2-6-2 mm, broadly to rather narrowly elliptic or oblanceolate, plane to incurved-conduplicate, not or slightly cucullate, midrib plane or slightly impressed beneath, margin not distinct, epidermis sometimes undulate, concolorous, some- times glaucous, + thinly to thickly coriaceous; apex acute, base narrowly cuneate, scarcely sheathing, pairs united to form narrow interfoliar ridge; basal vein 1, with c. 3—4 pairs of ascending lateral branches, sometimes branching distally, not prominent or impressed beneath, tertiary reticulation not visible; laminar glands + dense, sometimes visible beneath. Inflorescence 1-flowered, with pseudo-dichotomous branches from node below; pedicel 4-5 mm long, not incrassate upwards; upper leaves not transitional. Flowers 20-30 mm in diam., stellate. Sepals (7—)9-11 x (2-)2-5-3 mm, elliptic, acute, veins c. 9-11, obscurely reticulating distally, with midrib not or slightly prominent; glands linear, punctiform near apex. Petals deep yellow, (10—)14—17 x (5-)9-11 mm, 1-5-2 x sepals, obovate; apiculus acute; glands linear, distally interrupted to punctiform. Stamens c. 100-120, longest c. 5-8 mm long, c. 0-5 X petals. Ovary 2-5-3 x 1-5-2 mm, ellipsoid to subglobose; styles 3, 6-7-5 mm long, 2-5—3-5 X ovary, spreading-incurving; stigmas narrow to narrowly capitate. Capsule (6—)8—9 X 4-6 mm, ellipsoid, shorter than sepals. Seeds c. 1-2 mm long, ecarinate; testa finely scalariform. In paramo forest, scrub or damp grassland; 2400-4875 m. Colombia (Magdalena), Venezuela (Zulia and Mérida). In the Sierra Nevada de Santa Marta and the Sierra de Perija, with one probably conspecific record from the Paramo del Molino. Map 13. COLOMBIA. Magdalena: Sierra Nevada de Santa Marta, alredores de cabeceras 64 NORMAN K. B. ROBSON de Rio Sevilla, 3320-3410 m, 22.1.1959 (fl & fr), Barclay & Juajibioy 6620 (COL, MO, NY); Sierra Nevada de Santa Marta, valley of Rio Yebosimeina, 3500—4000 m, 22.v.1977 (fl & fr), Starker White & Alverson 547 (COL, MO, NY); Sierra de Perija, E. of Manaure, quebrada de Floridablanca, 2700-2800 m, 9—12.xi.1959, Cuatre- casas & Romero Castaneda 25171 (COL). VENEZUELA. Mérida: Péramo del Molino, 2600 m, 19.i.1922 (fl), Jahn 940 (VEN). Zulia: Perija, Cerro Fetari, iv.1952 (st), Urbano 2 (VEN). H. magdalenicum differs from the sessile-leaved form of H. pimeleoides by (i) the leaves with pinnate venation but with a scarcely sheathing base and (ii) the narrower petals and sepals and usually shorter styles. H. magdalenicum, which has a pivotal position in relation to over half the remaining species of sect. Brathys (Spp. 38-53), is rather variable. The nearest form to H. pimeleoides (from Sierra de Santa Marta, Rio Sevilla) has relatively broad, plane leaves with conspicuous gland dots beneath. Most other forms from Sierra de Santa Marta, as well as those from Sierra de Perija, have narrower, incurved- conduplicate leaves in which the gland dots are inconspicuous or superficially invisible (except in some Sierra de Perija populations). Of these, the population from the Rio Donachui area has stiffly erect, relatively thick stems and markedly imbricate leaves (tending towards 43 H. parallelum and 40 H. bolivaricum), whereas the other populations have more divergent, relatively thin stems and less markedly imbricate, though distinctly tetrastichous leaves (tending towards 50 H. lancifolium and its relatives). 33. Hypericum valleanum N. Robson, sp. nov. H. sprucei N. Robson affinis, sed foliis latioribus, nec profunde incurvatis nec cucullatis, manifeste glanduloso-punctatis, pedicello 3—4 mm longo, stylis crassius- culis, differt. Type: Colombia, Valle de Cauca, Los Farallones, filo de la cordillera, matorrales de paramo en el cerro La Torre, c. 3750 m, 10.x.1944 (fl), Cuatrecasas 177861 (F!, holotype & isotype). Shrub c. 1 m tall, erect, with branches strict, pseudo-dichotomous and lateral. Stems orange-brown, compressed? when young, soon subterete to terete, cortex exfoliat- ing irregularly; internodes 1-2 mm long. Leaves sessile, subimbricate, narrowly spreading and twisting, subtetrastichous, deciduous above base without fading; lamina 11-17 xX 2-3 mm, narrowly elliptic, shallowly incurved, not cucullate, midrib impressed, margin very narrowly hyaline, concolorous, not glaucous, coriaceous; apex sharply acute, base angustate, not sheathing, pairs united to form very narrow interfoliar ridge; basal vein 1, unbranched; laminar glands dense, large, visible beneath. Inflorescence 1-flowered, with pseudo-dichotomous branches from node below; pedicel stout, 3—4 mm long; upper leaves not transitional. Flowers c. 20 mm in diam.?, stellate. Sepals 7-10 x 2-2-8 mm, narrowly elliptic to lanceolate, acute to acuminate, incurved above, margin narrowly hyaline, veins 5, the outer branched, midrib not or scarcely prominent; glands linear, distally punctiform. Petals bright (?) yellow, 10-12 X 5-6 mm, c. 1:2-1-4 X sepals, obovate; apiculus sharply acute; glands striiform to punctiform. Stamens c. 60 ?, longest 5-6 mm long, c. 0-5 x petals. Ovary c.2 X 1:5mm, ellipsoid; styles 3, c. 3mm long, c. 1-5 x ovary, stout, suberect; stigmas narrow. Capsule and seeds not seen. In paramo; 3750 m. Colombia (Valle de Cauca). Known only from the type collection. Map 13 (p. 62). COLOMBIA. Valle de Cauca: Cordillera Occidental, Los Farallones, cerro La Torre, c. 3750 m, 10.x.1944 (fl), Cuatrecasas 17861 (F). H. valleanum is intermediate in form and distribution between the more primitive form of H. magdalenicum (with broad punctate leaves) and H. sprucei, differing from the former by its shorter styles and narrower leaves, and from the latter by its broader, thicker, and densely punctate not cucullate leaves and its stout styles and generally stronger habit. It appears to be a relict population. THE GENUS HYPERICUM L. 34. Hypericum sprucei N. Robson, sp. nov. H. aciculari Kunth affinis, sed caulibus crassioribus, foliis latioribus et saepissime longioribus, coriaceis, sepalis petalisque longioribus, staminibus numerosioribus, differt. Type: Ecuador, Tungurahua, [Volcan] Tungurahua, 2400-3000 m, xi.1857 (fl), Spruce 5110 (BM!, holotype; c!, GH!, K!, P!, S!, isotypes). Hypericum struthiolifolium sensu R. Keller in Bull. Herb. Boissier II, 8: 180 (1908) [‘struthiolaefolium’|; Gleason in Bull. Torrey bot. Club 56: 102, 105 (1929), pro parte excl. typum. Icon: Fig. 11. Shrub 0-2-2 m tall, erect to ascending, with branches strict, lateral or more rarely pseudo-dichotomous. Stems orange-brown, 4-lined and ancipitous when young, eventually terete, cortex exfoliating irregularly; internodes 1-5-7 mm long. Leaves sessile, densely to loosely imbricate, erect, spreading and twisting, scarcely tetra- stichous, deciduous above base without fading; lamina 6-12 x 0-8-1-7 mm, very narrowly oblanceolate to linear, incurved-cucullate, midrib impressed beneath, margin very narrowly hyaline, concolorous, not glaucous, coriaceous; apex acute, base parallel, not sheathing, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched; laminar glands rather dense, few or sometimes none visible beneath. Inflorescence 1-flowered, rarely with pseudo-dichotomous branches from node below; pedicel 1-5-3 mm long, not incrassate upwards; upper leaves not transitional. Flowers 20-30 mm in diam. , stellate. Sepals 7-11 x 1-5—2-5 mm, lanceolate, acute to acuminate, incurved above, margin hyaline, veins 5, midrib not prominent; glands linear, distally punctiform. Petals bright to deep yellow, 10-15 x 5-7 mm, c. 1-4 x sepals, oblong-oblanceolate; apiculus sharply acute to apiculate; glands striiform to punctiform, sparse. Stamens 60—75, longest 5—9 mm long, 0-5-0-6 x petals. Ovary 2-3-5 x 1-5-2 mm, ellipsoid-subglobose; styles 3(4), (2:5)3—4-5 mm long, (1-3)1-5 —2 X ovary, outcurving, slender; stigmas narrowly or not capitate. Capsule c. 5 x 3 mm, broadly ellipsoid to subglobose, shorter than sepals. Seeds c. 0-8 mm long (immature?), ecarinate; testa finely scalariform. Open, dry or damp paramo; 2725-4000 m. Ecuador (Carchi to Loja, apparently absent from Cotopaxi and Bolivar), Peru (Piura). Map 13 (p. 62). ECUADOR. Azuay: c. 10 km NE. of Sevilla de Oro, 3000—3200 m, 13.ix.1976 (fl), Dllgaard & Balslev 9449 (AAU, BM, MO); ‘Oriente’ border, Eastern Cordil- lera between Ona and rio Yacuambi, 3000-3360 m, 10—19.ix.1945 (fl & fr), Prieto P-297 (BM, NY). Canar: between Biblian and Camiar, c. 3350 m, 21.ix.1955 (fl), Asplund 17665 (S). Carchi: Base of Volcan Chiles, km 34—36 on Tulcan — Maldonado road, 3900-4050 m, 19.v.1973 (fl), Holm-Nielsen et al. 5888 (AAU, F, GB, MO, S, U). Chimborazo: Atilio, c. 3700 m, 29.i.1968 (fl), Harling, Storm & Strém 6705 (GB). Imbabura: Ibarra to Mariano Acosta, at pass (Hacienda Yura Cruz), 3600 —3800 m, 10.viii.1976 (fl & fr), Dllgaard & Balslev 8664 (AAU, BM, MO, U). Loja: Loja — Zamora road, 1 km east of pass (Km 12), 2725-2750 m, 17.iv.1973 (fl), Holm-Nielsen et al. K. 3698 (S). Morona-Santiago: Gualaceo—General Plaza (Lim6n) road, 3000-3100 m, 20.ic.1967 (fl), Sparre 18778 (S). Napo: Cordillera de Los Llanganati, Chihuila Sacho o Ainchilibi, c. 3750 m, 25—29.viii.1959, Barclay & Juajibioy 8996 (MO, NY); Sucumbios ‘Playon de San Francisco’, Camino al Cerro el Mirador, 3600-3800 m, 29.xii.1980 (fl), Jaramillo & Coelle 3903 (AAU, QCA). Pichincha: 15-4 km west of Papallacta, 3980 m, 26.iii.1972, MacBryde & Dwyer 1185 (MO). Tungurahua: Llanganati Mts, Lake Aucacocha, 3700 m, vii.1969, Edwards 58(K); Paramo of Minza Chica, 3800 m, 4.iv.1939 (fl), Penland 307 (F, GH). Zamora —Chinchipe: Loja to Zamora, 2800 m, 18.ix.1961, Dodson & Thien 685 (BM). PERU. Piura: above Huancabamba, 3000 m, iv.1912 (fl), Weberbauer 6133 (F, GH, S). H. sprucei provides a morphological and geographical link between H. valleanum (Colombia) and three species in southern Ecuador and Peru (Spp. 35-37). From H. 65 : 66 NORMAN K. B. ROBSON Cc Fig.11 A. H. sprucei: (a) habit; (b) stem with leaves; (c) leaf; (d) sepal; (e) petal; (f) stamens (partly cut away) and ovary; (g) capsule. B. H. valleanum: (h) habit. C. H. aciculare: (i) habit (a, h, i X 2; b x 2; c-g X 3). A. Ollgaard & Balslev 8454; B. Cuatrecasas 17861; C. Hartweg 719. valleanum it differs in the slender styles and the shorter, narrower leaves. For differences between it and the southern species, see their accounts below. H. sprucei varies clinally from north to south, the plants from Carchi being most similar to H. valleanum. Towards the southern end of its range (Azuay to Piura), the leaves become slenderer and more flexuous, and the habit low and multi-stemmed. In Chimborazo the leaves and flowers of some plants are relatively small, thus indicating a trend towards 35 H. aciculare. 35. Hypericum aciculare Kunth in Humboldt, Bonpland & Kunth, Nova Gen. et Spec. Pl. 5: 190 (1822); Choisy in DC., Prodr. syst. nat. regni veg. 1: 553 (1824); Jameson, Pl. Aequator. 1: 108 (1865); R. Keller in Bull. Herb. Boissier I1, 8: 176 (1908) ? (see below), in Engler THE GENUS HYPERICUM L. & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 181 (1925). Type: Ecuador, Loja, locis temporis prope Loxam (Regni Quitensi), 1908 m, viii.1805, Humboldt & Bon- pland s.n. (P!, holotype; F!, US!, photographs) H. struthiolifolium Juss. in Annls Mus. Hist. nat. Paris 3: 160 (1804) [‘struthiolae- folium’| pro parte, quoad var. innom. (p. 161, t. 16f. 2b). H. struthiolifolium [var.] B sensu Choisy, Prodr. monogr. fam. Hypéric.: 59 (1821) [‘struthiolaefolium’| pro parte, quoad spec. Peruv. H. struthiolifolium [var.] B minutum Choisy in DC., Prodr. syst. nat. regn. veg. 1: 533 (1824) [‘struthiolaefolium’|. Type: Peru, Dombey s.n. (G, holotype; P!, isotype). Brathys acicularis (Kunth) Spach, Hist. nat. vég. Phan. 5: 366 (1836), in Annis Sci. nat. II, 5: 366 (1836). H. laricifolium sensu Bentham, Pl. Hartweg.: 126 (1843-1844). Shrub 0-3-2 m tall, erect and bushy to decumbent and slender, with branches strict to ascending, lateral and frequently pseudo-dichotomous. Stems orange-brown, 4-lined and + ancipitous when young, eventually terete, cortex exfoliating irregularly; internodes 1-7 mm long. Leaves sessile, subimbricate or not, suberect, outcurving and usually twisting, not or scarcely tetrastichous, deciduous above base without fading; lamina 3-5-7 x 0-4—0-6 mm, very narrowly elliptic to linear-acicular, incurved, slightly cucullate, midrib impressed beneath, margin very narrowly hyaline, concolorous, not glaucous, subcoriaceous to chartaceous; apex acute, base parallel-sided, not sheathing, scarcely broadening but pairs forming narrow inter- foliar ridge; basal vein 1, unbranched; laminar glands dense to very sparse, visible beneath. Inflorescence 1-flowered, terminal, with pseudo-dichotomous branches from node below and terminating short unbranched lateral shoots; pedicel 1-5—2-5 mm long, not incrassate upwards; upper leaves not transitional. Flowers 10-17 mm in diam., stellate. Sepals 4-7-5 x 0-7—1-5 mm, narrowly lanceolate, acute, incurved above, margin hyaline, veins (3)5, midrib not prominent but whole sepal often becoming ribbed; glands linear, distally punctiform. Petals yellow or apricot to orange, sometimes tinged red, 6-10 x 3-7 mm, c. 1-5 X sepals, oblong-obovate, apiculus sharply acute; glands striiform to punctiform. Stamens 30-55, longest 3-5 mm long, c. 0-5 X petals. Ovary 1-5—2-5 x 1-1-5 mm, ellipsoid-subglobose; styles 3, 2-5-3(4) mm long, 1-2-2 X ovary, outcurving-ascending; stigmas narrow. Capsule 3-5-4 x 2-5-3-5 mm, broadly ellipsoid, shorter than sepals. Seeds 1-2 mm long, ecarinate; testa finely scalariform. Dry, open scrub on slopes of paramo, dense moist forested slopes, and wet sphagnum; 1800-3700 m. Ecuador (Azuay, Loja), Peru (Piura, Amazonas). Map 14 (p. 68). ECUADOR. Azuay: Km 85 on Pan-American Highway north of Loja, 2850—2950 m, 3.v.1973 (fl), Holm-Nielsen et al. 4808 (AAU, F, GB, S, U); Rio Collay south of El Pan, Huagrarancha slopes, 2650-3290 m, 6.vii.1943 (fl), Steyermark 53372 (F); *30-50 km S. of Cuenca, Paéramo de Tinajillas; 3300-3450, 17.ii1.1945, Camp E-2229 (BM, NY);* Rancho Ovejero, c. 3000 m, 1-2.viii.1969 (fl). Barclay & Juajibioy 8484 (NY). Loja: between La Toma and Loja, 1800-2600 m, 4.ix.1923 (fl), Hitchcock 21431 (GH, NY, US);* 7 km from Loja on road to Zamora, 2350 m, 8.ix.1959 (fl), B. & C. Maguire 61667 (BM, NY);* W. slopes of Cerro Villonaco, 2200-2250 m, 10.iv.1974 (fl & fr), Harling & Anderson 13351 (GB, NY). PERU. Piura: Huancabamba to Cuello del Indio, Mitopampa, 2650 m, 22.vii.1975 (fi), Sagdstegui, Cabanillas & Dios 8253 (MO). Amazonas:* Chachapoyas, 3—6 Km W. of Molinopampa, 2200-2450 m, 19.vii.1962 (fl), Wurdack 1403 (F, K, NY, S, US). *Pajonales, Pozuro, 3000 m, vii.1863 (fl), Pearce 252 (BM, K). H. aciculare differs essentially from H. sprucei in habit and size of parts. It most resembles some populations of the latter from Chimborazo, from which it is separated by a disjunction of merely c. 100 km. Other, less similar populations of H. sprucei, however, are found in the same area as H. aciculare, which occurs in two ecologically somewhat distinct forms with widely overlapping areas of distribution. If * Wet habitat form, see p. 68. 67 68 NORMAN K. B. ROBSON Map 14 35. H. aciculare A; 36. H. recurvum ®. these forms prove to be genetically distinct, it may be possible to recognize them as subspecies, as they are almost always distinguishable from one another. Plants on open or forested paramo slopes are bushy with branching mostly pseudo- dichotomous, whereas those in marshes and seepages are slender with branching mostly lateral (records marked*). Keller (1908 supra) keys out H. aciculare with H. nitidum Lam. as having connate styles. If his material did indeed belong to this species, then he must have interpreted it wrongly. 36. Hypericum recurvum N. Robson, sp. nov. H. sprucei N. Robson affinis, sed foliis aureo-marginatis apice longe acuminatis, recurvascentibus haud torquescentibus, petalis angustioribus, differt; a H. aciculari Kunth isdem characteribus differt. Type: Peru, Amazonas, Chachapoyas, middle eastern Calla-Calla slopes, near Kms 416-419 of Leimebamba-Balsas road, 2900 —3100 m, 9.vii.1962 (fl & fr), Wurdack 1303 (K!, holotype; F!, GH!, NY!, P!, S!, US, isotypes). Shrub 0-15—0-7 m tall, many-stemmed, erect or decumbent and rooting, with branches strict, pseudo-dichotomous and sometimes lateral. Stems orange- to red- dish-brown, 6-lined when young, eventually terete, cortex exfoliating in irregular flakes; internodes 1—3 mm long. Leaves sessile, imbricate-suberect but not markedly tetrastichous at first, eventually becoming markedly outcurved but not twisted, deciduous above the base after turning brown; lamina S—8 x 0-8—1-2 mm, subulate, distally conduplicate and + cucullate, otherwise incurved with midrib slightly prominent proximally to slightly impressed beneath, margin narrowly glaucous, chartaceous; apex acicular, base broadened, markedly sheathing, pairs united to form narrow interfoliar ridge; basal vein 1, unbranched; laminar glands rather dense, THE GENUS HYPERICUM L. visible beneath when young. Jnflorescence 1-flowered, with pseudo-dichotomous branches from node below; pedicel 3-5—5 mm long, not incrassate upwards; upper leaves not transitional. Flowers 15-20 mm in diam., stellate. Sepals 7-9 x 1-1-5mm, narrowly-lanceolate-acerose, long-acuminate, cucullate, with golden-hyaline mar- gin; veins 3-5, unbranched, not prominent; glands linear, interrupted in distal 1/4. Petals golden yellow, 8-12 x 2-5-4 mm, 1-2-1-4 xX sepals, narrowly obovate- oblong; apiculus acicular; glands linear, distally interrupted. Stamens c. 60, longest 6-8 mm long, c. 0-7 x petals. Ovary 2-5-3 x 1-1-2 mm, narrowly ellipsoid; styles 3, 4—5 mm long, c. 1:7 X ovary, outcurved-ascending; stigmas narrow to scarcely capitate. Capsule 3-5 x 2-3 mm, ellipsoid, shorter than sepals. Seeds 0-7—0-8 mm long, ecarinate; testa finely scalariform. On open hillsides in exposed and (in Pasco) boggy areas; 2700-3100 (?—3300) m. Peru (Amazonas (Chachapoyas), Junin, Pasco), apparently almost restricted to the Cerro Calla-Calla. Map 14. PERU. Amazonas: Chachapoyas, Cordillera Calla-Calla between Leimebamba and Balsas, 2900-3000 m, 21.viii.1963 (fl), Ferreyra & Acleto 15284 (NY); between Leimebamba and Balsas, 2700 m, 1.vi.1963 (fl & fr), Lépez, Sagastegui & Callantes 4429 (NY); Cerro Calla-Calla, E. side, 19 km above Leimebamba on Balsas road, 3100 m, 4.vi.1964 (fl), Hutchinson & Wright 5505 (F, NY, UC, US); Jalca de Calla-Calla, 3100 m, 7.v.1970 (fl & fr), Sagdastegui 7451 (F, NY, VEN); Chachapoyas to Moyobamba, La Jalca, 2700-3300 m, 20.i.1930 (fl), Williams 7583 (F, NY); Chachapoyas, 1826 (fl & fr), Mathews 2130 (BM, K, P). Junin: Huancayo, Quebrada Occopilla, ii.1948 (fl), Soukoup 3636 (MO). Pasco: Oxapampa, Cordillera Yana- chaga, Cerro Pajonal 12 km SE. of Oxapampa, 2700-2800 m, 7.x.1982 (fr), Foster 8992 (BM, MO). H. recurvum is clearly related to the southern, low, many-stemmed form of H. sprucei, differing from it mainly in the often glandular, acuminate leaves, which becomes recurved, not twisted, and are suffused with a golden tint (at least when dried). The longer styles and narrower, acuminate sepals distinguish H. recurvum from H. andinum, which has an overlapping area of distribution. 37. Hypericum wurdackii N. Robson, sp. nov. H. recurvo N. Robson affinis, sed foliis longioribus, carinatis, densissime imbricatis, valde tetrastichis alas formantibus, floribus maioribus, differt; a H. aciculare Kunth isdem characteribus differt. Type: Peru, Amazonas, Chachapoyas, summit of Cerro Malcabal (Cerro Tumbe), 3-6 km SW. of Molinopampa, 2850-2900 m, 20.vii.1962 (fl & fr), Wurdack 1426 (K!, holotype; F!, GH!, MICH!, NY!, P!, S!, US, isotypes). Shrub 0-15—0-3 m tall, several-stemmed, erect, with branches strict, pseudo- dichotomous and sometimes lateral. Stems orange-brown, completely enclosed by leaves when young, eventually terete, cortex exfoliating irregularly; internodes 1 mm long. Leaves sessile, imbricate, markedly tetrastichous with the vertical rows winglike, not or scarcely outcurving, not twisted, deciduous above the base after turning brown; lamina 9-14 x 1-2 mm, linear-acerose, incurved-conduplicate, cucullate, carinate wholly or only at apex and base, margin narrowly golden-hyaline especially towards base, otherwise concolorous, yellow-green, not glaucous, char- taceous; apex sharply acute, base broadened, markedly sheathing, pairs united to form deep interfoliar ridge; basal vein 1, unbranched; laminar glands dense, invisible below in mature leaf. Inflorescence 1-flowered, with pseudo-dichotomous branches from node below; pedicel 2-3 mm long, not incrassate upwards; upper leaves not transitional. Flowers 20-25 mm in diam., stellate. Sepals 10-12 x 3-4 mm, lan- ceolate-acerose, sharply acute, not cucullate, with golden-hyaline margin; veins 7, unbranched, not or scarcely prominent; glands linear, punctiform in upper half. Petals golden (?) yellow, 15-18 xX 7-8 mm, c. 1-5 X sepals, narrowly obovate; apiculus acute; glands linear, distally interrupted. Stamens c. 100, longest 8-9 mm long, c. 0-5 X petals. Ovary c. 2:5 X 1-5 mm, ellipsoid; styles 3, 4-5 mm long, 1-7-2 X ovary, outcurved-ascending; stigmas narrow. Capsule c. 6 X 3-5 mm, ellipsoid, shorter than sepals. Seeds not seen. 30 m Tall closed- Tall — a — forest open-forest Trees 10-30 m Closed- Open-forest Open-forest Woodland Open- forest woodland Trees <10m Low closed- Low Low Low Low open- forest open-forest open-forest woodland woodland Shrubs >2 m Closed Open-scrub Open-scrub Tall Tall open- scrub shrubland shrubland Shrubs 0-25-—2 m Heathy Closed- Open- Open- Low Low open- - heathland heathland heathland shrubland shrubland Chenopodiaceous — — Low Low Low open- shrubland shrubland shrubland Shrubs <0:25m — —_— — Dwarfopen- Dwarf open- heathland heathland Hummock grasses — —_— — Hummock Open hummock grassland grassland a relatively high Foliage Projective Cover (FPC) e.g. (1) Closed-forest, (2) Tall open-forest, (3) Open-forest, and (4) Open-scrub. These structural formations are made up of overstorey and understorey vegetation, and depending on the amount of canopy cover, vary in FPC from 100% to 30%. They can be briefly described as follows: (a) Closed-forest (also referred to as rain-forest, it can also describe some mangrove forests). This community comprises many different tree genera and species which alter in composition with change in latitude. It is found in disjunct pockets from Cape York to Tasmania and grades from tropical mesophyll vine forests (with leaf size 45-180 cm’) to subtropical mesophyll and notophyll vine forest (with leaf size 20-25-45 cm?) to warm temperate notophyll vine forest to cool temperate microphyll moss forest (with leaf size 2-25—20-25 cm’). Closed-forest communities occur in humid areas where rainfall exceeds 1500 mm per annum; overstorey FPC in such a formation approaches 100%. These formations reach their best development on the basaltic soils of the Eastern Highlands. Several species of Ramalina grow in this habitat, two species are specific to the rain-forest canopy, and another three species are found in both closed-forest and open-forest communities. These closed-forests can contain trees which exceed 30m in height, but in monsoonal northern Australia and coastal Queensland, low closed-forests occur which are composed of trees reaching less than 10 m in height. During the wet season these forests have a FPC of 100%, but during the dry winters the FPC may be as low as 70% as some of the trees are semi- deciduous. In inland Queensland the low closed-forest is the only formation which supports any Ramalina species. (b) Tall open-forest (with sclerophyllous understorey with or without tree ferns) occurs in sub-humid sites of eastern, south-eastern and south-western Australia. Overstorey FPC in such a community varies from 70% to 50%. The dominant tree species in the overstorey is Eucalyptus, whereas the low tree/tall shrub stratum (less than 10 m high) contains 18 genera of angiosperms (Specht, 1981a). Several species of Ramalina have been found growing on twigs in the understorey, but rarely do species of Ramalina grow on eucalypts. (c) Open-forest (with either heathy sclerophyll or grassy understorey). This formation is the most widespread in Australia, occurring in the north, east, and southern parts of the continent; each area experiences quite different rainfall patterns ranging from summer rains to summer- winter uniform rainfall, to winter rains respectively. The genus Eucalyptus dominates these 114 G. NELL STEVENS communities, although species of Casuarina and Acacia are also plentiful. The overstorey FPC for open-forest communities varies between 70-30%. No Ramalina species grow in the tropical open-forests, but occur on Casuarina and Acacia in the temperate communities. (d) Mallee open-scrub (with savannah-chenopod understorey). This structural formation is common in the drier Mediterranean climate of southern Australia where species of Eucalypt dominate, but other genera, including Acacia are scattered between the mallee eucalypts. Overstorey FPC in these areas can be as low as 30%. The relatively low rainfall (250-800 mm p.a.) occurs in winter, followed by hot dry summers. Four species of Ramalina have been collected in this formation. Seral plant communities exist along the coastal fringe of Australia; these are coastal dune communities and saline coastal communities, the latter made up of salt-marsh, mangrove vegetation, and marine meadow (Specht, 1981a); only the coastal dune community and the mangroves act as phorophytes for Ramalina species. Such communities usually produce a series of structural formations, influenced by micro-habitat conditions; the overstorey FPC therefore differs in different parts of the community. (A) Coastal dune community: In the tropics, lowland rain-forest abuts the open coastline in areas where mangroves are not dominant; in the subtropics the sand dune vegetation comprises various shrub species of Banksia integrifolia, Callitris columellaris , and Casuarina equisitifolia as the dominant tree species. Along the southern coast-line, the foredunes are vegetated by heathy shrubs; Pimelea serpyllifolia, Leptospermum laevigatum, Leucopogon parviflora, Monotoca elliptica, Bursaria spinosa, and Acacia sophorae. Several Ramalina species occur in this community, the species varying with change in latitude and longitude. (B) The mangroves (saline coastal community): This vegetation occupies the muddy areas at the mouths of rivers and bays along the eastern, north-western, and northern coastlines of Australia, with pockets in South Australia. The structure of such communities varies from closed-forest to woodland to tall shrubland. The major concentration of mangrove vegetation (approx. 28 species) is confined to the tropics gradually dwindling in species numbers in the subtropics and eventually being reduced to one species, Avicennia marina, towards the temperate zone. Both FPC and the amount of rainfall vary considerably depending on the latitude of the mangrove community. Those mangrove species which act as phorophytes to the Ramalinae are Rhizophora stylosa, Rhizophora apiculata, Ceriops tagal, Bruguiera gymnorhiza, Avicennia marina, Aegiceras corniculatum, Lumnitzera racemosa, and Excoecaria agallocha. (4) Geology The saxicolous species of Ramalina are totally dependent on the presence of rock outcrops within their distribution range. Corticolous species, however, are only indirectly influenced through the presence or absence of particular plant communities which grow on the different soil types. The four main rock types involved in the areas where Ramalina species occur are (1) basaltic rocks (2) metamorphics and fine grained sedimentary rocks (3) sandstones and (4) granites. The major areas of basalt are Tertiary in age, and are confined to the Eastern Highlands. Most of these elevated areas are forested with either rain-forest or tall open-forest. Additional areas of basalt of Quaternary age are found in north-east and central-east Queensland, supporting low closed-forests, and western Victoria at lower altitudes on the coastal plain which supports extensive areas of open-woodland and grassland. In eastern Australia the metamorphic rocks and many fine-grained, compacted sedimentary rocks which are associated with them are of Palaeozoic age, and have been folded and elevated, forming mountain ranges and hilly country in the Eastern Highlands. They have been intruded by granite batholiths, which also make up large areas, having been revealed by prolonged erosion e.g. New England, Murrumbidgee, Kosciusko, and many other masses in the eastern states. These are areas of high elevation subject to cold winters. The sandstones which are most important are those sub-horizontal deposits of Mesozoic age RAMALINA IN AUSTRALIA 1i5 forming for example, much of the surface of the Sydney Basin, Moreton Basin, and the Carnarvon Range. The Grampians are composed of dipping sandstones of Palaeozoic age. In South Australia and Western Australia only a few species of Ramalina occur on trees growing on soils derived from old granites and metamorphic rocks (schists and quartzites). There are very few occurrences of Ramalinae in areas of calcareous rocks of the Tertiary basins and Pleistocene dune deposits. Historical understanding of Ramalina and Ramalinaceae (1) THE GENUS RAMALINA The genus Ramalina was first described by Acharius (1810) when he distinguished it from Parmelia by its cartilaginous thallus and apothecial characters. Acharius (1810) listed ten species (together with their forms) as belonging to the genus, namely Ramalina homalea, R. linearis, R. complanata, R. peruviana, R. polymorpha, R. fraxinea, R. fastigiata, R. scopulorum, R. farinacea, and R. pollinaria. Except for the first, all of these species have been retained in the genus to the present day. Many more taxa have been added to this original number in the last 175 years, some being removed later into separate genera because of differences found in the anatomy, the spore shape and size, or the colour of pycnidia. Montagne (1852) examined the cortex of R. homalea Ach. and found that the hyphae were arranged perpendicular to the surface without any mechanical support tissue in the cortex; he therefore erected a new genus Desmazieria Mont. to incorporate all Ramalina species possessing this anatomical feature, using R. homalea as the type. Massalongo (1854) erected the genus Cenozosia Massal. based on R. inanis Mont. as the type and characterized by its fistulose spongy thallus with a single layer of branching cells in the cortex. This monotypic genus has been retained as a distinct taxon up to the present. By using the colour of spermatia, as well as anatomical and thallus differences, Stizenberger (1862) divided the genus Ramalina into three parts (i) Cenozosia (Massal.) (type R. inanis Mont.) as a species with thallus hollow, spermatia black (ii) Desmazieria (Mont.) (type R. ceruchis Ach.) for those species with simple cortex, thallus interior cottony and spermatia black and (iii) Euramalina Stizenb. (type R. scopulorum Retz.) for species having a double cortex, thallus interior cottony, and spermatia pale or colourless. Nylander (1870) also used colour of spermogonia conceptacles for initial division within the genus, separating species with (1) totally black spermatia (2) partially black or (3) pale or colourless spermatia. Within group (1) he then used differences in anatomical structure of the cortex to separate R. inanis Mont. (= Cenozosia Massal.) from the rest. He placed five other species in another group which corresponds to Desmazieria Mont., namely R. ceruchis Ach., R. combeiodes Nyl., R. homalea Ach., R. flaccescens Nyl., and R. testudinaria Nyl. Ramalina melanothrix Laurer was separated from the rest because of its two-layered cortex (= Euramalina Stizenb.); its present position is in the genus Trichoramalina Rundel & Bowler. Group (2) contained only R. carpathica K6rber, a species which Bowler & Rundel (1977) placed in the genus Fistulariella Bowler & Rundel. Group (3) contained 56 species which today belong in three genera, Ramalina Ach., Niebla Rundel & Bowler, and Fistulariella Bowler & Rundel. Vainio (1890) also laid stress on anatomical structure of the cortex, accepting section Euramalina Stizenb. as having a chondroid cortex made up of longitudinal hyphae, and sections Desmazieria (Mont.) and Cenozosia (Massal.) as having a cortex made up of transverse hyphae. Vainio divided Euramalina Stizenb. into two stirps based on thallus structure (1) Fistularia Vainio species with inflated hollow branches and (2) Myelopoea Vainio species with a con- tinuous arachnoid or cottony medulla. He used branch shape to subdivide stirps Myelopoea into series Teretiusculae Vainio (containing species with terete or angular branches) and series Compressiusculae Vainio (containing species with flattened branches rarely two-edged). Hue (1901) followed Vainio in this division of Ramalina when he divided the genus into three groups on morphology: (A) Teretiusculae Vainio (B) Compressiusculae Vainio and (C) Fistulo- sae Hue. Steiner (1904) based his divisions of Ramalina on the arrangement of cortical tissue external to 116 G. NELL STEVENS the algal layer. His section Corticatae (outer layer consisting of more or less transversely arranged conglutinated hyphae) was equivalent to Desmazieria Mont., his section Bitectae (outer layer as above overlying a tissue of longitudinally arranged conglutinated hyphae) was similar to Euramalina Stizenb. He made a third division Ecorticatae (cortex composed of a layer of longitudinally oriented conglutinated hyphae which formed a closed ring). Ramalina arabum (Ach.) Meyen & Flotow is the type for Ecorticatae. Choisy (1954) commented that this thallus structure was similar to Alectoria Ach. Howe (1913-14) used spore characters to distinguish the sections he created within the genus Ramalina as he thought this represented a more natural classification. Section Ellipsosporae R. H. Howe contained species with ellipsoid or oblong spores (9-20 x 3:5—7-5 zm) and section Fusisporae R. H. Howe contained species with long fusiform spores (16-35 xX 3-7 wm) and section Bistortae R. H. Howe contained a species with sigmoid spores. Each of these sections he divided into series and these series correspond in the main with previous divisions, e.g. Howe’s series Desmazieria (Mont.) = section Corticatae Steiner = genus Niebla Bowler & Rundel. Series Myelopoeae = part of Myelopoea Vainio, series Fistulariae = stirps Fistularia Vainio = genus Fistulariella Bowler & Rundel. He made three other series, Tenuicorticate, Ciliate, and Fusisporae. This classification is not used. Du Rietz (1926) relegated the genus Desmazieria Mont. to subgenus Desmazieria (Mont.) (containing R. ceruchis, R. homalae, etc.) and Euramalina Stizenb. to subgenus Euramalina (Stizenb.), which he divided into section Tenuicorticatae R. H. Howe and section Bitectae Steiner. Section Tenuicorticatae contained species without a chondroid element and this section he further divided into subsection Solidae Du Rietz (species with a solid thallus, containing R. evernioides, etc.) and subsection Tubulosae Du Rietz (species with inflated fistulose thallus: R. inanis Mont. = Cenozosia Massal.). Du Rietz maintained section Bitectae Steiner (= Euramalina Stizenb.) but divided it into two subsections (a) subsection Myelopoea Vainio thallus solid, and (b) subsection Fistularia Vainio thallus inflated, fistulose. These extra subsections make the classification complicated although the basic criteria used for division are sound. Zahlbruckner (1907) combined the sections of both Stizenberger (1862) and Steiner (1904) and made three divisions in the genus Ramalina: (i) section Ecorticatae Steiner (ii) section Corticatae Steiner (which incorporated Cenozosia Massal. and Desmazieria Stizenb.) and (iii) section Euramalina Stizenb. (= Bitectae Steiner). Zahlbruckner subdivided Euramalina Stizenb. using the Vainio names Fistularia Vainio and Myelopoea Vainio and series Teretius- culae Vainio and Compressiusculae Vainio. Present day taxonomists are still divided in their acceptance of which species should be placed in the genus Ramalina sens. str. Bowler & Rundel (1977) established the new genus Fistulariella Bowler & Rundel for those species originally placed in Ramalina stirps Fistularia Vainio. They placed emphasis on the hollow, usually perforate thallus as traits warranting genus status, yet Australian material belonging to stirps Fistularia Vainio present a broad range of morphology varying from almost totally compressed to totally fistulose. It is therefore regarded here that the trait of inflation of the thallus should have no greater emphasis placed upon it than the roundness of branches. Rundel & Bowler (1978) proposed a new generic name Niebla to replace the genus name Desmazieria Mont. as they found the former name was a homonym for the earlier described genus Desmazeria Dumortier (Poaceae), and 13 taxa were transferred by them from Desma- zieria and Ramalina into Niebla Rundel & Bowler. Of the various divisions put forward up to the present day, the taxonomic system used by Vainio would most suitably classify the Australian material, e.g. the Australian Ramalinae would broadly divide into three groups: (1) fistulose thalli (2) terete and solid thalli and (3) compressed and solid thalli. (2) THE FAMILY RAMALINACEAE Agardh (1821: 93) erected the family Ramalinaceae [‘Ramalinez’], in which he placed the genus Ramalina. Since that time the genus has been alternatively included with Usnea and other RAMALINA IN AUSTRALIA Ly fruticose genera in the family Usneaceae, or isolated in the family Ramalinaceae. Table 2 shows the various classifications which record Ramalina in Ramalinaceae (with or without other members of Usneaceae) and the diversity of opinions as to the position of the genus. Apparently Zahlbruckner (1907) and Smith (1921) did not regard septate spores nor a cortex with mechanical tissue sufficiently different to separate Ramalina from other genera in the family Usneaceae; but Watson (1929) regarded spore septation as most important and placed Ramalina in the Ramalinaceae. With modern day improvement in techniques of microscopy and chemistry together with a better understanding of the fungal/algal composition, the natural relationships which exist between the genera are being more fully understood. Follmann & Huneck (1969) placed three genera in the family Ramalinaceae based on their chemotaxonomy and anatomy (viz. Ramalina, Ramalinopsis, and Desmazieria). Poelt (1974) also regarded these three genera as forming a natural entity and placed them in Ramalinaceae, but he added a fourth genus which was not closely related, e.g. Speerschneidera. Culberson & Culberson (1970) pointed out that chemical evidence supported segregation of Ramalinaceae as a separate family as no species in the Usneaceae proper produces orcinol meta-depsides, a category of substances richly represented in Ramalina. Henssen & Jahns (1973) have kept the genus Ramalina in the family Ramalinaceae, because of its two-celled spores and exobasidial conidiophores, and because of the specific lichen substances it contains. They stated that the structure of the apothecium of Ramalina resembles that of the Parmeliaceae in that there is a more or less complete algal layer beneath the hymenium, but the Ramalinaceae differ from the characteristic cupular exciple of the Parmeliaceae in that the algal layer is next to, not separated from the loosely interwoven subhymenium. Keuk (1979) divided the genera in Ramalinaceae into two groups based on anatomical and chemical differences, and colour of pycnidial walls (a) Ramalina, Ramalinopsis, and Tricho- ramalina and (b) Niebla (= Desmazieria Mont. and Cenozosia Massal.). The most modern division within the family Ramalinaceae has been based on cortical diversity. Bowler (1981) defined seven genera in the Ramalinaceae: Ramalina Ach., Fistu- lariella Bowler & Rundel, Niebla Rundel & Bowler, Ramalinopsis (Zahlbr.) Follm. & Huneck, Trichoramalina Rundel & Bowler, Dievernia M. Choisy, and Cenozosia Massal. Of these, the genus Ramalina contains the largest number of taxa. It can be seen from Table 2 that all recent authors who have specifically dealt with the question of relationships of genera within the two families Ramalinaceae and Usneaceae, have con- sidered them to be distinct morphologically, chemically, and taxonomically. Collectors of Australian Ramalinae Because the early explorers in Australian waters sailed in the temperate latitudes, it was only to be expected that the first Ramalina collected in Australia would be a temperate region species; this collection was made by Labillardiére, a botanist on the French ship ‘Recherche’ which sailed along the Great Australian Bight and visited Tasmania in 1791 (Ducker, 1979). This first specimen, held at H, (H-NYL 37226) was annotated ‘Ramalina leiodea Nyl. Labillardiere, C. van Dieman’. This was an erroneous identification which has now been determined as R. inflata J. D. Hook. & Taylor. In 1802, Robert Brown in his voyages with Matthew Flinders collected R. inflata from Port Jackson (Sydney, N.S.W.), but this specimen was determined as Lichen fastigiatus Ach. by Crombie (1880). When J. D. Hooker accompanied Sir James Ross in his Antarctic expedition in the ships ‘Erebus’ and ‘Terror’ in 1839, he collected the type of R. inflata J. D. Hook. & Taylor on the Lord Auckland Islands and from New Zealand he collected the type of R. geniculata J.D. Hook. & Taylor. They also visited Tasmania and collected material which was later published in Flora Tasmaniae (1859). In this work Churchill Babington cites R. tasmanica Ny]l. as a new species but 118 G. NELL STEVENS Table2 The classification of genera within the two families Ramalinaceae and Usneaceae as proposed by various lichenologists from 1814 to the present day. Author Date Class/Order Tribe/Family Genus Acharius, E. 1814 Class Homothalami Order Primus Scutellati Ramalina, Alectoria, Collema Order Secundus Peltati Usnea, Cornicularia Agardh, C. 1821 Ramalineae Ramalina Eschweiler, F. G. 1824 Usneaceae Usnea, Ramalina, Alectoria + Cornicularia, Dactylina, Desmaziera, Endocena, Evernia, Everniopsis, Letharia, Neuropogon, Oropogon, Siphula, Thamnolia. Montagne, J. F. C. 1852 Tr. Parmeliacea subtribe Usnea Ramalina, Usnea Nylander, W. 1857 _—_Lichenacei Ser. Ramalodei Tr. Ramalinei Ramalina, Alectoria, Evernia, Dufourea, Dactylina Tr. Usneei Usnea, Neuropogon, Chlorea Stizenberger, E. 1862 Class Tr. Usneaceae Homothalami Ramalineae Ramalina, Dactylina, Cetraria, Evernia, Dufourea Usneaceae Usnea, Alectoria, Oropogon, Argopsis Vainio, E. A. 1890 Order Tr. Parmelieae Ramalina, Usnea, Alectoria, Cyclocarpeae Evernia Hue, A. M. 1901 Series Radiatae Cyclocarpae Tr. Ramlineae Ramalina, Dufourea Tr. Usneae Usnea Tr. Alectoreae Alectoria Zahlbruckner, A. 1907. Order Usneaceae Ramalina, Usnea, Alectoria, Cyclocarpeae Oropogon, Dufourea, Evernia, Everniopsis, Endocena, Letharia, Thamnolia, Siphula, Dactylina. Smith, A. L. 1921 subseries Usneaceae Ramalina, Usnea, Alectoria, Cyclocarpineae Evernia, Letharia, Everniopsis, Oropogon, Thamnolia, Siphula, Dactylina, Dufourea, Endocena Watson, W. 1929 Order Ramalinaceae Ramalina Parmeliales Usneaceae Usnea, Alectoria, Letharia, Evernia, Everniopsis, Dactylina, Dufourea, Thamnolia, Siphula, Endocena Rasanen, V.J.P.B. 1943 Order Usneaceae Ramalina, Usnea, Alectoria, Cyclocarpeae Evernia, Letharia, Everniopsis, Dufourea, Dactylina, Oropogon, Ramalea Choisy, M. 1957 Ramalinaceae Ramalina Usneaceae Usnea, Desmazieria, Cenozosia RAMALINA IN AUSTRALIA 119 Table 2 — cont. Author Date Class/Order Tribe/Family Genus Follmann, G. & 1969 Ramalinaceae Ramalina, Ramalinopsis, Huneck, S. Desmazieria Henssen, A. & 1973 Order Ramalinaceae Ramalina Jahns, H. M. Lecanorales suborder Lecanorineae Poelt, J. 1974 Order Ramalinaceae Ramalina, Ramalinopsis, Lecanorales Desmaziera, Speerschneidera, suborder Usneaceae Usnea, Alectoria, Bryopogon, Lecaniorieae Cornicularia, Evernia, Everniopsis, Himantormia, Letharia, Sulcaria, Neuropogon, Oropogon Bowler, P. A. 1981 Ramalinaceae Ramalina, Fistulariella, Trichoramalina, Niebla, Dievernia, Cenozosia, Ramalinopsis examination of this material during the present study has shown this specimen to be yet another morphotype of the very variable R. inflata. The Reverend Dr. W. Woolls, a school master at Parramatta, N.S.W., in 1832, collected in that area, and one Ramalina specimen called R. gracilis Nyl. (MEL 9432) is actually R. exiguella Stirton. During the period 1842—46, J. P. Verreaux (a French ornithologist) who was sent to Australia by the Museum of Natural History in Paris, collected widely in Tasmania and mainland Australia. In his botanical collections were several Ramalina species which he had obtained during that period, either by collecting them himself or by purchase of material from others (Lamy, in litt.). Because of this practice of obtaining specimens from other collectors there is uncertainty about the locality given by Nylander for two ‘type’ specimens he records as collected by Verreaux from Swan River, Western Australia, these are R. australiensis Nyl. and R. confirmata Nyl. Both of these species grow in eastern Australia, and the Western Australian types are the only material collected from that area. Nylander himself appeared unsure of the locality of R. australiensis as he noted: ‘in Australia (prope Swan River, ni fallor) legit Verreaux 1846’. Verreaux collections held at PC indicate he travelled north as far as Moreton Bay, Queensland, and collected several Ramalina species from mangroves. The first woman recorded as a lichen collector was Amalie Dietrich, who collected lichens in Queensland for a wealthy Hamburg merchant named Godeffroy, for his private museum. Some of these collections are now housed at Minchen. They are collections of Ramalina species from the Rockhampton area. Friedrich Ludwig Leichhardt travelled through Queensland on his way to Port Essington (Darwin), 1844-45. He collected a Ramalina (now R. filicaulis N. Stevens) along the way but gave no exact location. As the occurrence of this species is rare today, more information about the locality would have been helpful in the present study. Ferdinand von Mueller, Victorian Government Botanist (1853-96), received lichen speci- mens from various collectors in Australia and these specimens were sent to overseas lichenol- ogists for determination, e.g. Anton Krempelhuber, Jean Miiller [Arg.], and James Stirton. Daniel Sullivan, a headmaster, collected lichens for von Mueller, one being the ‘type’ of R. glaucescens Krempelh., from Mt Ararat, western Victoria, and another was R. leiodea var. fastigiatula Mill. Arg. ‘type’ from the Grampians. Another Ramalina which Sullivan collected from Mt Ararat area (MEL 9471) now bears the name R. fimbriata Krog & Swinscow, a species common in East Africa. Charles French, a plant propagator at Melbourne Botanical Gardens 120 G. NELL STEVENS (Filson, 1976), collected the type of R. inflata var. fissa Mill. Arg. from King Island, Bass Strait. The type of R. lacerata Mill. Arg. was collected at Eucla, Western Australia, by J. Oliver. Collectors in New South Wales for the Victorian Botanical Department included T. White who collected the type of R. myrioclada Mill. Arg. from Twofold Bay, New South Wales. Charles Knight, a New Zealand surgeon, collected lichens in the Sydney area; he named one specimen R. subgeniculata C. Knight, which was subsequently changed to R. knightiana Zahlbr. Knight also named R. minuscula var. alba a specimen collected by James Keys at Mt Perry, Queensland. This name was published by Bailey (1886), but no description was given and later John Shirley (1888) described it. Messrs. Pentzke and Hartmann were two collectors in Queensland for the Victorian Botanical Department (Wilson, 1889). Pentzke collected the type material of R. farinacea var. nervulosa Mill. Arg. from Daintree River, north Queensland, and C. H. Hartman of Toowoomba collected the type material of R. geniculata var. compacta Mill. Arg. Collections made by F. M. Bailey, Government Botanist in Queensland, John Shirley, a school inspector, and Mrs M. Thozet, were sent to James Stirton, W. Leighton, Charles Knight, or Jean Miller [Arg.] for identification (Wilson, 1889). It was Stirton who named the two Ramalina types R. exiguella Stirton and R. perpusilla Stirton. Miller Arg. described a Rockhampton specimen collected by Mrs Thozet as R. farinacea var. squarrosa Mill. Arg. The type of R. farinacea var. dendroides Mill. Arg. was collected by E. Forde from the Hunter River, New South Wales. Both of these taxa have now been found to be morphotypes of R. peruviana Ach. The Reverend F. R. M. Wilson, from Kew, Victoria, collected prolifically (1877-97) in eastern Australia, from Tasmania to Brisbane, and more particularly along Port Phillip Bay and Gippsland, Victoria (Filson, 1976). He collected the type of R. calicaris var. australica Rasanen from Barwon Heads. Wilson himself described several taxa of Ramalina, e.g. R. unilateralis F. Wilson, R. brevis F. Wilson, and R. brevis var. brevissima F. Wilson. More recent collectors of Ramalinae in Australia are J. H. Willis, Assistant Government Botanist at the National Herbarium Melbourne (Filson, 1976), A. C. Beauglehole, who collected Ramalinae in South Australia and Victoria, J. S. Whinray, who made an extensive survey of the lichen flora of the Bass Strait Islands, and G. C. Bratt and J. A. Cashin, whose collections of Tasmanian lichens have proved invaluable in the present study. Other present day lichenologists who have collected Ramalina specimens which have been used in the present study are A. Archer (N.S.W.), J. A. Elix (A.C.T.), R. Filson (Vic.), G. Kantvilas (Tas.), R. W. Rogers (Qld.), N. Sammy (W.A.), C. Scarlett (Qld), and R. Seppelt (S.A.). Materials and methods The source of material for this research was threefold. (a) Personal collections (numbering over 10,000 specimens), the majority from Australia, with small collections from the Pacific Islands of Fiji, New Caledonia, Rarotonga, Tahiti, New Hebrides, New Zealand, and Hawaii. (b) Herbarium material from institutes, as acknowledged. (c) Herbarium and living material from private collections, as acknowledged. Most of the specimens examined during this study were tested by thin-layer chromatography using the techniques of Culberson (1972). Solvents B and C proved to be the most useful solvents to separate the acids found in the various species of Ramalina. The chemistry of type material was also checked. Examination of the surface features of the thallus (e.g. pseudocyphellae, soralia, and soredia shapes) and spore size and shape, was carried out by use of the Scanning Electron Microscope (SEM) using either a Cambridge or a Phillips 505. In order to observe the shape and thickness of the hyphae composing the outer cortex and the supportive tissue thin sections were examined by TEM, using an AEI Corinth 275 operated at 60 kV. Preparation of thallus material involved fixation of the tissue in glutaraldehyde, post fixation in osmium tetroxide, dehydration and then infiltration and embedding with Spurrs medium. The results were not always successful as the tissue tended to pull away from the medium along the outer cortex junction. RAMALINA IN AUSTRALIA 1A! All climatic data were derived from Meterological Bureau records (Climatic averages of Australia, 1975). Results Morphology (a) Soralia and soredia: Soralia are defined herein as those areas of thallus where clusters of soredia are found, and the term soredia is applied to the more or less spherical bodies composed of algae surrounded by short hyphae which act as vegetative diaspores. These vegetative reproduction bodies are produced by 11 of the Australian Ramalinae, and their constancy in occurrence enables this attribute to be used taxonomically. The position and form of the soralia are also useful criteria in the separation of each sorediate taxon. Round or ellipsoid soralia are produced by R. pacifica, R. caespitella, R. nervulosa var. nervulosa, R. nervulosa var. luciae, and R. nervulosa var. dumeticola (Plate 1, fig. 3); these occur marginally and laminally. Ramalina peruviana and R. tenella produce small, punctiform soralia (Plate 1, figs 1, 2) which are mainly apical in R. tenella and marginal or lateral in R. peruviana. The soralia found in the other four taxa are produced in several ways, e.g. (i) by disintegration of the lower cortex (R. fimbriata, Plate 1, fig. 5) (ii) by the partial separation, both marginally and apically, of the upper and lower cortices (R. canariensis) (iii) by lack of a continuous lower cortex (R. reducta, Plate 1, fig. 4) and (iv) by the splitting or disintegration of parts of the lower cortex at intervals along the branches (R. unilateralis, Plate 1, fig. 6). The soralia terminology used by Du Rietz (1924) and Beltman (1978) was not generally applicable to the Ramalinae. The composition of the soredia was examined under SEM; the algal cells were not discernable but the fragments of hyphae which make up the outer surface of the soredium were observed to differ in form in several of the taxa. The northern hemisphere species R. capitata is regarded as having soredia covered by an ‘epicortex-like’ layer (James, pers. comm.) (Plate 2, fig. 1), and such a structure was looked for in the Australian sorediate taxa. Four species were found to possess some semblance of an outer covering (Plate 2, figs 2,3, 4,5). Whether this layer should be compared to the ‘epicortex’ which occurs in some taxa in the Parmeliaceae as described by Hale (1973, 1981) is debatable; but it can be likened to the ‘Kittsubstanz’ (cementing substance) mentioned by Peveling (1970). Such a layer may act as a protection against wetting of the soredia. The Australian taxa which produce soredia with this outer covering are R. caespitella, R. reducta, R. tenella, and R. fimbriata, and even in these only a minority of the soredia show it clearly. The soredia of R. caespitella are small, spherical, and have a partially smooth surface (Plate 2, fig. 2). Soredia of R. reducta are irregular in shape and size but fusion of the external hyphal segments was apparent (Plate 2, fig. 3). Ramalina tenella soredia are small but well defined, and some have a distinct outer covering (Plate 2, fig. 4). Ramalina fimbriata soredia are Plate 1 (overleaf) SEM photographs showing soralia shapes in some of the Australian Ramalinae. Fig. 1 Apical, punctiform soralia — R. tenella. Scale 10 mm = 530 wm (x 19). Fig. 2 Marginal/lateral, punctiform soralia— R. peruviana. Scale 10 mm = 150 wm (x 65). Fig. 3 Round and ellipsoid soralia— R. nervulosa var. dumeticola. Scale 10 mm = 220 wm (xX 44). Fig. 4 Exposed lower surface with chondroid strands of cortical tissue across area of soralia— R. reducta. Scale 10 mm = 270 «um (X 36). Fig. 5 Patches of exposed medulla by disintegration of the under side — R. fimbriata. Scale 10 mm = 200 wm (xX 51). Fig. 6 Lower surface split apart and eroded at intervals — R. unilateralis. Scale 10 mm = 530 wm (X 19). Plate 2 (overleaf) SEM photographs showing soredia forms in some of the Australian Ramalinae. Scale indicated by black bar. Fig. 1 R. capitata —- showing smooth outer layer partially covering the soredium. Scale 10 mm = 22 wm (xX 440). Fig. 2 R. caespitella — showing portion of soredia covered by a smooth layer. Scale 10 mm = 22 wm (x 440). Fig. 3 R. reducta — showing soredia with smooth surface (x 364). Fig. 4 R. tenella — showing thin layer partially enclosing the soredium (xX 1236). Fig. 5 R. fimbriata — showing thin layer totally enclosing soredium (X 958). Fig. 6 R. unilateralis — showing soredium with a ‘woolly’ appearance; no outer layer present (X 1527). Fig. 7 R. nervulosa var. dumeticola — showing soredia with irregular surface (x 873). Fig. 8 R. canariensis — showing soredia with a ‘woolly’ appearance; no outer layer present (x 909). G. NELL STEVENS Plate 1 RAMALINA IN AUSTRALIA Plate 2 124 G. NELL STEVENS large and spherical, and the formation of an external layer could be discerned on many of the soredia at various stages of development; the most advanced layer formation is shown in Plate 2, fig. 5. All of these taxa are regarded as having granular soredia. The soredia produced by R. unilateralis (Plate 2, fig. 6) and by R. canariensis (Plate 2, fig. 8) are small in size and are composed of both thick and thin pieces of hyphae; there is no external fusion of the hyphae so the outer surface remains loosely interwoven giving a woolly appearance to the soredia (Plate 2, fig. 6). The soredia are farinose. The four taxa which belong in the R. farinacea complex (R. pacifica, R. nervulosa var. nervulosa, R. nervulosa var. luciae, and R. nervulosa var. dumeticola; Plate 2, fig. 7; Plate 3, figs 1, 2, 3) also produce soredia which have an irregular surface made up of separate hyphal pieces which are not fused (Plate 3, fig. 2), and the soredia are farinose. Ramalina peruviana produces soredia which have an irregular surface (Plate 3, fig. 4). Fibrils often arise from mounds of farinose soredia (Plate 1, fig. 2) and these features have the same structure as the soredia (Plate 3, fig. 7). (b) Apothecia: The production of apothecia as a sexual means of reproduction occurs in all but four of the Australian Ramalinae (i.e. seven of the 11 sorediate taxa also produce apothecia, but only rarely, and often the mature ascospores are few and difficult to find). The esorediate taxa generally produce numerous apothecia. The position of the apothecia on the branches can be of taxonomic importance, being terminal, subterminal, marginal, lateral, or laminal. Terminal refers to the apothecia at the end of a branch. Subterminal refers to apothecia positioned near the end of the branch, with the branch extending beyond the apothecium, forming a short attenuate branchlet or spur, which is usually bent at an angle in relation to the main branch. If the spur grows long it makes the branch appear geniculate. Marginal refers to apothecia on the edges of compressed branches, and laminal refers to apothecia occurring on the surface of the thallus. If branches are subterete to terete the apothecia occur laterally. When laminal apothecia occur specifically at the base of bifurcating branches they are referred to as positioned at the axil of the branch. Apothecial shape varies from concave to plane to convex; in some species the apothecia remain concave to maturity, in others the apothecia are concave to plane when immature, but are markedly convex at maturity. The thalline margin is usually distinct and entire, occasionally becoming incised at maturity; when the disc becomes convex the margin is often obscured. The size of the apothecia varies greatly from 0-2 mm to 5-0 mm with exceptions to 10 mm. Most apothecia are shortly stalked, but sessile apothecia are found in R. australiensis and R. filicaulis. The colour of the disc is usually yellow-green and pruinose; however, taxa in the R. celastri complex have been found with orange coloured discs. The apothecial tissue comprises a colourless hymenium formed of asci and paraphyses. The asci are clavate to subcylindrical and lie amongst sparsely branched paraphyses over a developed hypothecium. This hypothecial layer is usually brownish-yellow as is the epithecium [as used in a lichenological context — Henssen & Jahns (1973)]. (c) Spores: The number of spores per ascus is eight, usually distichous in arrangement (Plate 5, fig. 5), hyaline, 1-septate, varying considerably in size and shape, being broadly ellipsoidal to fusiform and either straight to slightly curved or reniform (Plates 4 & 5). Within a single ascus spores can vary from straight to curved, a feature also noted by Krog & Plate 3 SEM photographs showing soredia forms and surface features in some of the Australian Ramalinae. Scale indicated by black bar. Fig. 1 R. nervulosa var. luciae — showing soredia with irregular surface (x 750). Fig. 2 R. nervulosa var. nervulosa — showing fragments of hyphae making up the outer surface, without a surface layer (x 945). Fig. 3 R. pacifica — showing soredia with an irregular surface (x 525). Fig. 4 R. peruviana — showing soredium with an irregular surface (x 1300). Fig. 5 Punctiform pseudocyphellae — R. subfraxinea var. confirmata (x 75). Fig. 6 Enlargement of section of Fig. 5 (x 750). Fig. 7 Fibril produced from a mound of soredia — R. peruviana. Scale 10 mm = 30 wm (xX 341). Fig. 8 Tuberculate pseudocyphellae — R. subfraxinea var. subfraxinea. Scale 10 mm = 30 wm (X 341). = =| < -4 i n ~ < q < Z = 126 G. NELL STEVENS Swinscow (1975) and Landr6n (1972) (Plate 4, fig. 4); therefore the shape of spores was not used as a taxonomic character. The size of spores has also been found to be of limited taxonomic value as most mature spores of the Ramalina species lie within the size range (8—)10-16 x 3-5-6 um (Plate 4, figs 2,3, 4,5, 6). The relationship between the width and length of spores, as suggested by Krog & Swinscow (1974) was investigated in this study to see if this could be a taxonomic tool for separation of closely related taxa; however exceptions could always be found in the spore ratios which negated the usefulness of this characteristic (Plate 5, figs 2, 3, 4). Although the spores are 2-celled, some of the spores in the size range 14-16 x 4um appear 3 or 4-celled. These additional divisions are not produced by distoseptae so must be dense cytoplasmic strands (which take up the cotton blue stain) across the cell which produce a ‘mock septation’. The spores of each taxa examined under SEM showed surfaces which were remarkably smooth (Plates 4 & 5) except for a film which seemed to coat some of the spores or remained in patches attached to the spore surface (Plate 4, fig. 1; Plate 5, figs 1, 2, 4, 6). (d) Pycnidia: The occurrence of pycnidia amongst the Australian Ramalina species is not common. When present, they are always pale in colour, and produce pycnoconidia which are rod-shaped, usually 3-5 x 0-5 jm in size. It was observed that they mainly occurred on sterile thalli rather than on thalli bearing either apothecia or soredia. This suggests that they may provide an alternative means of propagation for the species. Their irregular occurrence and apparent uniformity prevented their taxonomic use. (e) Pseudocyphellae: Many lichenologists have discussed these morphological characters (Du Reitz, 1924; Culberson & Culberson, 1968; Duncan, 1970; Beltman, 1978; Bowler, 1981; and Hale, 1981). It is generally accepted that these features are cortical pores in the surface of the thallus which allow the exchange of gases. They are areas which have direct contact with the medullary hyphae. In the Ramalinae they vary in shape and size and several names have been applied to such surface features: striae, striations, tubercules, and papillae. Pseudocyphellae appear as shallow depressions (striae) in the cortex (Plate 3, figs 5, 6) or as white linear markings (striations) on the surface (Stevens, 1983a: figs 1B-D, 2A-B). In the Ramalina subfraxinea complex in Australia some of the taxa produce tubercles resembling raised warts, the tips of these structures containing small pseudocyphellae (Plate 3, fig. 8). The term ‘tuberculate pseudocyphellae’ was used by Krog & Swinscow (1975) to describe this form of pseudocyphellae. Similar surface features occur in several North American taxa, but are referred to as papillae, as in R. sinaloensis (Bowler & Rundel, 1972b) and R. complanata (Landr6n, 1972). Most sorediate Ramalina species produce pseudocyphellae of one kind or another; in the case of the R. farinacea taxa from the tropics, it was found that the elongate pseudocyphellae were precursors to soralia formation (Stevens, 1983a). Bowler (1981) reported that pseudocyphellae are mainly absent in the Fistulariella taxa and this was found to be so in the Australian fistulose material. As a diagnostic character for the separation of taxa, pseudocyphellae have been found to be of little value. Anatomy The anatomical structure of all Ramalina species found in Australia comprises (1) a cortex (which I regard as both the outer cortex plus the inner supportive tissue) and (2) a medulla (which includes the phycobiont). Plate4 SEM photographs showing spore shapes in the Australian Ramalinae. Fig. 1 R. subfraxinea var. leiodea— single spore. Scale 10 mm = 1-2 wm (X 8000). Fig. 2 R. australiensis spores. Scale 10 mm = 44m (x 2400). Fig. 3 R. pacifica spores. Scale 10 mm = 4 wm (X 2400). Fig. 4 R. subfraxinea var. confirmata spores. Scale 10 mm = 4 wm (X 2400). Fig. 5 R. subfraxinea var. norstictica spores. Scale 10 mm = 4 wm (Xx 2400). Fig. 6 R. subfraxinea var. subfraxinea spores. Scale 10 mm = 4 wm (Xx 2400). = s < oe = n 5 z g < g = : n Z a = rs = an i} a a Zz o RAMALINA IN AUSTRALIA 129 1) Cortex ae Australian taxa the outer cortical layer is usually 20 wm in thickness and is composed of thick walled, transverse or randomly oriented hyphae; it may be indistinct or even absent. The inner layer serves as supportive tissue and varies in thickness. It is composed of bundles of more or less periclinally arranged thick walled hyphae which make up a continuous or interrupted sheath around the internal medullary hyphae. The detail in the outer cortex is often difficult to detect under light microscopy when yellowish granules are present; a problem pointed out by Bowler (1981). To obtain as much detail as possible, sections were examined under TEM. Both cortical layers of hyphae are highly gelatinized and a uniform distribution of lumina is apparent; each lumen is surrounded by a thick matrix of fibrillar material arranged concentrically (Plate 6, fig. 1). Peveling (1974) commented that it is difficult to determine whether the different layers around hyphae are part of the proper cell wall or if they are extracellular additions to the wall. The change in orientation of the hyphae is apparent under TEM. In transverse section the randomly orientated outer layer hyphae appear elongate whereas the longitudinally oriented hyphae appear circular (Plate 6, fig. 2). Scattered between the hyphae are outlines of crystals, which are more dense in the outer region and may be usnic acid crystals (Plate 6, fig. 3). (a) Anatomical structure as a means of separating taxa. Krog & Osthagen (1980) divided the species of Ramalina found in the Canary Islands into four anatomical types based on the position of the supportive tissue: the farinacea type, the decipiens type, the bourgaeana type, and the duriaei type. The latter two types occur in species now placed in the genera Niebla and Dievernia and are therefore not relevant to Australian material. The majority of the Australian Ramalinae could be classed as belonging to the farinacea type (Figs 3A, B), as they possess a thin outer layer of tissue and a continuous inner layer of mechanical tissue of uneven thickness. Variations can occur in the arrangement of this inner supportive tissue if it becomes partly discontinuous, leaving sections of outer tissue unsupported (Fig. 3C). However, this is not comparable with the ‘decipiens type’ where some of the strands anastomose across the medulla. None of the Australian taxa could be separated on this trait. Bowler (1981) pointed out that the Ramalina taxa could be divided arbitrarily into two groups based on the degree of specialization in the cortical tissue. One group with a narrow, indistinct outer cortex intergrading with the underlying supportive tissue and the other group with a distinctly double-structured cortex, but he commented that intergradation could be seen between the two forms. The Australian taxa can be divided in this way generally, but thalli within one taxon can often show differences in thickness of the outer cortical layer. Those taxa with a sparse or indistinct outer cortex are R. caespitella, R. celastricomplex, R. pacifica, R. nervulosa, R. exiguella (Plate 6, fig. 4), R. peruviana, R. reducta, and R. unilateralis. Those taxa with a comparatively distinct double-structured cortex are R. australiensis, R. canariensis, R. filicaulis, R. fimbriata, R. glaucescens, R. inflata complex, R. litorea, R. subfraxinea complex (Plate 6, fig. 5), and R. tenella. (b) Variation in the anatomical structure within the thallus. It was noticed when serial sectioning thalli that differences in anatomical structure sometimes occurred depending on the portion of the branch cut. Two forms of variation were found. The first form of variation is (i) variation in the alignment of the hyphae in the supportive tissue in basal sections compared to the hyphal arrangement in the mid and apical sections. Basal sections of R. celastri subsp. ovalis show a typical thin outer cortex merging with the longitudinally oriented hyphae in the supportive tissue, as found in the anatomy of R. celastri subsp. celastri (Plate 6, fig. 7). Sections cut through the middle of the branch and towards the apex showed only a wide band of randomly oriented << Plate5 SEM photographs showing spore shapes in the Australian Ramalinae. Fig. 1 R. exiguella spores. Scale 10 mm = 4 wm (X 2400). Fig. 2 R. inflata subsp. inflata spores. Scale 10 mm = 4 wm (X 2400). Fig. 3 R. inflata subsp. perpusilla (inland taxon) spores. Scale 10 mm = 4 wm (x 2400). Fig. 4 R. inflata subsp. perpusilla (coastal taxon) spores. Scale 10 mm = 4 wm (X 2400). Fig. 5 R. celastri subsp. celastri spores within the ascus. Scale 10 mm = 8-3 wm (X 1163). Fig. 6 R. litorea spores. Scale 10 mm = 4 wm (x 2400). 130 G. NELL STEVENS 0.5 mm Fig.3 | Anatomical structures of Australian Ramalina thalli: A. Cross section showing uneven thickness of supportive tissue in typical ‘farinacea-type’ anatomy; B. Cross section of R. filicaulis terete branch showing typical ‘farinacea-type’ anatomy; C. Cross section of R. inflata subsp. perpusilla thallus showing variation from Fig. A — the supportive tissue being partly discontinuous; D. Algal clusters on the upper side only, as found in some thalli of the R. inflata group; E. Cross section of R. whinrayi showing thick outer cortex and irregular arrangement of supportive tissue which extends into the medullary region, with isolated rafts of supportive tissue also present. hyphae forming a network (Plate 6, fig. 6). This anomaly may be explained by the number of laminal apothecia that this taxon produces on both sides of the branch. The lack of supportive tissue and the random arrangement of hyphae could be due to the repeated formation of apothecial tissue from the middle of the branch to the apex. Brandt (1906) found that variations occurred within a single thallus, and illustrated this with serial sections along the branch of Ramalina curnowii. The second form of variation is (ii) variation in cortical structure between the upper side of the branch and the lower side. This was noticed by Brandt (1906) in the species R. populina. In Australian material such variation was found to occur in some specimens of R. inflata subsp. inflata and R. inflata subsp. perpusilla, where the inner cortex on the lower side is narrow, continuous and of even thickness, without an algal layer, but on the upper side it is partly discontinuous, of uneven thickness with algal clusters interspersed between the ridges (Fig. 3D). This formation appears to be a modification of the normal ‘farinacea type’ structure, to enable optimal functioning of the phycobiont under particular environmental conditions. It is not a constant feature of these taxa, so cannot be used taxonomically. Another taxon which has an atypical and complex anatomy (Fig. 3E) is R. whinrayi. The cortex consists of a thick outer cortex and a thick, discontinuous sheath of supportive tissue made up of bundles with ridges extending across the medulla or occurring as isolated rafts in the medulla; such intrusions produce islands of medullary tissue with algal clumps. This form of anatomy has been sighted in the northern hemisphere taxon R. capitata (= R. strepsilis). (2) Medulla The central portion of the thallus is occupied by medullary hyphae which are randomly oriented, either tightly compacted or loosely woven and arachnoid. In the fistulose taxa the hyphae may be RAMALINA IN AUSTRALIA 131 Plate6 TEM photographs showing anatomical features. External edge of outer cortex indicated by black arrow. Fig. 1 Lumen with surrounding fibrillar matrix. Scale 10 mm = 1-5 wm (X 6770). Fig. 2 Section from outer to inner cortex — R. exiguella. Scale 10 mm = 1 wm (X 10450). Fig. 3 Usnic acid? crystals in outer cortex — R. peruviana. Scale 10 mm = 1-8 wm (X 5682). Fig. 4 Section from outer to inner cortex showing narrow outer cortex. Scale 10 mm = 5 wm (xX 2125). Fig. 5 Section from ouer to inner cortex showing broad outer cortex. Scale 10 mm = 1:2 wm (xX 5000). Fig. 6 L.S. of mid-section of branch of R. celastri subsp. ovalis. Scale 1 mm = 10 wm (X 105). Fig. 7 L.S. of mid-section of branch of R. celastri subsp. celastri. Scale 1 mm = 19 wm (x 53). (Photographs for Figs 6 & 7 by courtesy of A. Henssen.) almost totally lacking, or occur in patches appressed to the inner cortex. Ramalina fissa produces very little medullary hyphae, whereas in some R. inflata subsp. inflata specimens the central cavity is criss-crossed by arachnoid hyphae to the same degree as R. glaucescens, a species regarded as having a solid, compressed thallus. 132 G. NELL STEVENS The sorediate species belonging to the R. farinacea group in Australia have a thick felty medulla compared to the R. farinacea taxa in the northern hemisphere which have a loosely compacted but continuous medullary layer. Other sorediate species, R. canariensis, R. fimbriata, R. reducta, and R. unilateralis, reveal large areas of exposed medulla which produce soralia. The most compacted medullary hyphae are found in R. whinrayi (Fig. 3E). Chemistry Usnic acid is found in the cortex of the Ramalinae in varying amounts but it is not used taxonomically as a diagnostic property. Atranorin occurs only rarely and in trace amounts so it is not used for taxonomic purposes either. It is the medullary substances produced by the Ramalinae — the depsides and the depsidones — which are important. (a) The orcinol depsides: These comprise both para- and meta-depsides. The para-depside divaricatic acid is particularly common and is found in ten Australian Ramalina species. It may occur as the sole medullary constituent or be accompanied by biogenetically closely related accessory acids such as stenosporic acid, nordivaricatic acid (both para-depsides) or by sekikaic acid (a meta-depside). Evernic acid, another para-depside occurs in one saxicolous species endemic to Australia, and lecanoric acid has been found as an accessory acid in this species. Evernic acid is rare in southern hemisphere Ramalinae, whereas in the northern hemisphere it is a relatively common con- stituent in this genus. The greatest number of lichen acids found in the Australian Ramalinae belong to the orcinol meta-depside group. Fourteen substances have been identified in this study. The most common metabolite is sekikaic acid which is found in nine species. In marked contrast it is found in only a few northern hemisphere Ramalina species. Sekikaic acid may occur as the sole medullary constituent or be accompanied by a number of minor acids viz. homosekikaic, ramalinolic, 4’-0-demethylsekikaic, 4’-O-methylnorhomosekikaic, and 4’-0-methylnorsekikaic acids. Cryptochlorophaeic acid is another orcinol meta-depside which has been observed alone or may occur accompanied by paludosic and 4’-0-methylcryptochlorophaeic acids. Boninic acid too, occurs as a major acid with the minor metabolites 2-0-methylsekikaic, 2,4’-di-0- methylnorsekikaic, 4’-0-methylpaludosic, and 4,4’-di-O-methylcryoptochlorophaeic acids (Chester & Elix, 1978). (b) The B-orcinol depsidones: Only four B-orcinol depsidones have been observed in Austra- lian Ramalinae. Salazinic acid is present in four species and may be accompanied by trace amounts of protocetraric acid, but protocetraric acid does not occur as a major component in any of the species. Norstictic acid occurs in one coastal species and psoromic acid is present in one inland species. The number of depsidones found in the Australian Ramalinae is less than the number found in Ramalaina species in Europe or North America, where fumarprotocetraric acid, hypopro- tocetraric acid, stictic acid, protocetraric acid, norstictic acid, psoromic acid, and salazinic acid all occur. Salazinic acid occurs in both sorediate and fertile Ramalina species which inhabit the Australian coastline (viz. R. pacifica, R. tenella, and R. fissa). In R. pacifica and R. tenella trace amounts of protocetraric acid are sometimes present accompanying the salazinic acid. The occurrence of these two acids in the same thallus is probably due to incomplete enzymic oxidation of the side chain methyl group in the conversion of protocetraric acid to salazinic acid (Elix, in litt.). Triterpenoids, derived from the mevalonic pathway occur in specimens of R. tenella, and terpenes have been found in several species but neither of these groups was utilized for taxonomic purposes. Chemical variation and species concept : Since the chemistry of a particular species is usually constant it has proved a useful property in RAMALINA IN AUSTRALIA 133 combination with morphological characters to define a lichen species. However, in some instances, morphologically identical taxa are found to contain different acids and the problem arises of how to treat them taxonomically. Elix (1982) pointed out that there are three common patterns of chemical variation (a) replacement type compounds (b) accessory type compounds and (c) chemosyndromic variation. (a) Replacement compounds: Replacement compounds occur in a number of Australian Ramalina taxa including R. whinrayi, R. glaucescens, the R. inflata complex, the R. farinacea complex, and the R. subfraxinea complex. The most common acids involved are divaricatic acid, sekikaic acid, and salazinic acid. The thalli of R. whinrayi produce either divaricatic acid + nordivaricatic acid or the replacement compound sekikaic acid with minor metabolites (4’-0-demethylsekikaic acid and trace amounts of 4’-0-methyl norhomosekikaic acid). On the Tasmanian coastline both of these races occur in the one population, their morphology is indistinguishable, and one thallus was found to contain both divaricatic and sekikaic acids. On the Bass Strait islands only the divaricatic acid race occurs on the islands in the Hogans Group (the most northerly islands), whereas on the Kents Group of islands and on Craggy Island only the sekikaic acid race was found, although most thalli contain trace amounts of divaricatic acid as well. Because of this dual occurrence of the acids in some thalli, both acid races are regarded as belonging to the one taxon, viz. R. whinrayi. An intensive study of the habitats in which each acid race occurs would be necessary to understand and interpret the segregation of the two acid races on the different islands. A survey similar to that carried out by Culberson (1969a) on taxa in the R. siliquosa complex may solve this problem if it could be found that the Tasmanian population occupied separate niches according to the acids present. In this case the chemical races could be regarded as taxonomically significant (Culberson, 1967, 1969b) and the explanation of the mixed acids would infer hybridization (Brodo, 1978). But the islands are small and inaccessible and such a survey was not possible during this study. The occurrence of replacement compounds was also observed in the R. inflata group, where lichen populations which are morphologically indistinguishable may contain either divaricatic acid or sekikaic acid as medullary substances. Here again such taxa have been regarded as chemical strains within the species; a decision strengthened by the presence of trace amounts of sekikaic acid in some thalli which contain divaricatic acid. A third replacement compound, salazinic acid, also occurs in the R. inflata complex. Biogenetically this depsidone is quite remote from the two depsides, yet thalli possessing this acid are remarkably similar to those with a depside chemistry. However, the depsidone taxon has been retained as a separate species (R. fissa), as it has a geographically different distribution pattern from the depside containing taxa. (b) Accessory compounds: Accessory type compounds are also present in the R. inflata complex, occurring sporadically in addition to the constant constituents, but having no correlation with any morphological or distributional variations (Elix, 1982). The presence of nordivaricatic acid in some divaricatic acid taxa and the presence of consalazinic or scabrosin derivatives in the salazinic taxon have been regarded as instances of accessory acids accompany- ing the major constituent and thus requiring no taxonomic recognition. Similarly the accessory acid connorstictic has been found in the norstictic acid taxon in the R. subfraxinea complex. Another Ramalina species complex comprising taxa with chemical variants of the replace- ment type and with accessory compounds is the tropical/subtropical R. farinacea complex. The acids involved are divaricatic acid (+ accessory compound stenosporic acid) and the sekikaic aggregate of acids and salazinic acid, but this group is further complicated by chemosyndromic variation occurring within taxa which produce the sekikaic aggregate of acids. (c) Chemosyndromic variation: A chemosyndrome is a group of biogenetically related metab- olites where one or two compounds are regularly the major components, and the minor 134 G. NELL STEVENS biosequentially related constituent of one taxon becomes the major constituent of the other (Culberson & Culberson, 1977; Elix, 1982). In the R. farinacea complex, the two tropical/subtropical taxa R. nervulosa var. luciae and R. nervulosa var. dumeticola contain the same acid components but these occur in different quantities in each taxon (Stevens, 19834). Correlation between secondary-product chemistry and ecogeography in the Ramalina subfraxinea complex The importance of chemistry in identifying taxa which have different amplitudes of ecological tolerance has been stressed by Culberson (1967, 1969b), Culberson & Culberson (1967), and Sheard (1978). The notable example of correlation between secondary-product chemistry and ecogeography is found in the R. siliquosa species complex of the northern hemisphere. This complex contains six sibling species each with a different depsidone chemistry (norstictic acid, salazinic acid, stictic acid, protocetraric acid, hypoprotocetraric acid, and acid deficient). Culberson (1969b) maintained that each chemical race was a different species which reflected the physiological differences between the taxa. However, the R. siliquosa chemistry comprises a replacement series of medullary depsidones which are able to be ranked by increasing numbers of oxidation steps in their biosynthesis (Culberson, Culberson & Johnson, 1977). The maritime R. subfraxinea complex in the southern hemisphere is similar to the maritime R. siliquosa complex in the northern hemisphere as it is made up of taxa which produce six different acids; four of the acids are depsides and two are depsidones (viz. sekikaic acid, divaricatic acid, cryptochlorophaeic acid, boninic acid, norstictic, and salazinic acids), but these acids are not closely related biosequentially (Elix, in litt.). Morphologically only the salazinic acid taxon is distinguishable; the other taxa show only slight morphological variations from one another, none of which is sufficiently unique and consistent that the chemical constituents of every individual can be infallibly predicted from appearance alone. Culberson, Culberson & Johnson (1977) found a correlation between the chemistry in the R. siliquosa taxa and different amplitudes of ecological tolerance of each taxon. Accordingly, a study of the thalli belonging to the R. subfraxinea complex was undertaken in order to find if a particular chemotype could be related to a particular type of habitat based on the amount of exposure to the sea (e.g. exposed = on trees or rocks facing the open sea; sheltered = on trees growing on the landward fringe of a mangrove community or on trees in clay pan habitats). In Table 3 the type of habitat (exposed or sheltered) at each location site is compared with the percentage of thalli containing each acid found at the site. The table shows that there is a difference in the percentage of certain acids at each site. The cryptochlorophaeic acid taxon occurs in areas experiencing the greatest amount of exposure to the open sea, the boninic acid taxon is generally found in exposed habitats (usually adjacent to quiet water), whereas the thalli containing sekikaic acid, divaricatic acid, or norstictic acid are mainly found in sheltered well protected areas away from salt spray. Salazinic acid containing thalli appear to tolerate a broad spectrum of habitats from exposed to sheltered. From Table 3 it can be seen that there is a definite tendency for different chemotypes to occupy some microhabitats more commonly than others. This environmental selection of different chemotypes into separate habitats is also apparent to a larger extent along the more tropical coastline of Australia (e.g. north of latitude 16°30’S), where collections made from any one site always contain only one chemotype. When the type of habitat was considered it was found that collections from Princess Charlotte Bay (a vast area of clay pan) contained sekikaic acid, whereas thalli collected from the exposed shorelines of the small tropical islands of Ingram I., Turtle I., and Sue I. all contained cryptochlorophaeic acid, thalli collected from Lizard I. all contained salazinic acid. Although the collections from these northern regions may not have been as numerous as those to the south, enough material was collected at each site to make these findings plausible. RAMALINA IN AUSTRALIA 135 Table 3 Medullary metabolites expressed as a percentage of total number of specimens collected at various sites in relation to type of habitat and latitude. 2 o e = = 2) ) ) ae o & = = ) 3 <2 5 a a) Z & 6 s 3 = 8 = ae ae S — [=| — hd par) _— Le Be) oO = S d. ey ele ey. Bien ee Kees a 5 e Be Onna Bee ere oe ae 16 Daintree River sheltered —_ — 23 37 40 87 16 Mossman sheltered — 29 63 — 8 — 38 17 Mission Beach exposed — a> 10 — 15 — 20 18 Murray River exposed — 40 — 33 Ps | i 18 Hinchbrook I. sheltered 1 42 24 15: 16 2 82 19 Townsville exposed — a9 Pan, 12 — 7, 20 Bowen exposed 8 94 — 3 — — 53 Zi Hallidays Bay A. sheltered — — 32 68 — — 34 eA Hallidays Bay B. exposed — 94 — 6 — — an 21 Eimeo sheltered 25 35 4 36 — — 28 22 St Lawrence sheltered 9 8 14 69 — — 70 23 Keppel Sands exposed — 96 — 4 — — 46 23 Gladstone sheltered 3 42 30 Zo — — 60 24 Turkey sheltered 4 48 48 — — — 33 25 Burnett Hds exposed 10 90 — — — — 23 26 Hervey Bay exposed 36 64 — — — — an 5s Moreton Bay exposed 1 99 — — — — 100 29 Clarence R. exposed 66 34 — —_— — — 23 Distribution patterns in relation to climatic indices The distinct distribution patterns of the Ramalinae indicate that there are differences in the ecological requirements of each taxon. The composite map of the distribution of all of the taxa (Fig. 1) shows that geographically the distributions lie in either the tropical or the temperate zones with some taxa overlapping in the intermediate zones (e.g. subtropical and warm temperate). In the higher plants it has been recognized that plants growing under tropical conditions have different growth responses compared to plants growing under temperate conditions, and that such responses are closely related to thermal optima for photosynthesis (Specht, 1981c; Nix, 1982). Asa knowledge of the major climatic indices which regulate plant response (e.g. temperature, precipitation, and solar radiation) has proved valuable in understanding the distribution of the phanerogram flora of Australia, it seemed reasonable to examine their value in understanding the distribution of the Ramalina taxa throughout Australia. Nix (1982) proposed that Australian plants can be classified in terms of their photosynthetic response to temperature. His three major groups are relevant to the pattern of Ramalinae distribution. (1) Megatherm Group — plants with photosynthetic optimum 26—28°C; lower threshold 10°C, upper threshold 38°C. (2) Mesotherm Group — plants with photosynthetic optimum 19—22°C; lower threshold 5°C, upper threshold 33°C. (3) Microtherm Groups — plants with photosynthetic optimum 10—14°C; lower threshold 0°C, upper threshold 25°C. These three major groupings have geographical equivalents as shown in Table 4. Nix (1982) integrated the climatic factors, temperature, precipitation, and solar radiation into a single multifactor Growth Index (G.I.). The Growth Index is different for plants with different thermal growth responses. Figs 4A-B and 5A indicate the distribution of megatherm, mesotherm, and microtherm plants respectively with G.I. values >0-45 and G.I. values >0-30 136 G. NELL STEVENS Table 4 Thermal divisions of Australian plant biota and their geographical equivalents, as suggested by Nix (1982). Mean annual Geographical Thermal divisions temperature equivalent Megatherm 24°C tropical Megatherm/Mesotherm interzone 20-—24°C subtropical Mesotherm 14—20°C Mesotherm/Microtherm warm temperate interzone 12-14°C cool temperate Microtherm <1 <0-45 in the first two groups and with G.I. values <0-6 to >0-2 in the latter. The multifactor Growth Index (G.I.) has values ranging from zero to unity and can never exceed the value of the single most limiting factor. It is apparent from Fig. 4C, that Ramalina taxa occur only in those areas with a G.I. larger than 0-30 in at least one thermal response category. The distribution pattern of each Ramalina taxon conforms in a general sense to the patterns of G.I. for megathermic, mesothermic, and microthermic phanerogams, indicating that the climatic indices which determine distribution limits for the higher plant species also create similar limits for the species of Ramalina. Biogeographical division of the Australian Ramalinae By superimposing the phanerogam thermal growth response patterns onto the distribution pattern of each Ramalina taxon, the latter are able to be divided into megatherm, mesotherm, and microtherm groups quite satisfactorily (Figs 4C, SB). (1) The megatherm element Nine of the Ramalinae can be regarded as confined to a region with a megatherm Growth Index value >0-30 <0-45. Their southern limits rarely extend below the tropics (latitudes 23°S—24°S). These taxa are R. tropica, R. tenella, R. subfraxinea var. subfraxinea, R. subfraxinea var. norstictica, R. nervulosa var. nervulosa, R. nervulosa var. luciae, and R. litorea. Ramalina subfraxinea var. confirmata and R. subfraxinea var. leiodea are exceptions as their southern limit reaches latitudes 33°S and 29°S respectively (Table 5). The occurrence of R. subfraxinea var. confirmata on Sue Island (latitude 9°20’S) confirms its placement in the megatherm group; R subfraxinea var. leiodea extends to 18°S (Figs 4A, 28). (2) The mesotherm element Other Ramalina taxa which have an extended range along the eastern coastline are regarded as mesotherm elements as they have a southern distribution reaching the warm temperate region of Australia. These taxa are R. exiguella, R. inflata subsp. perpusilla, R. peruviana, and R. pacifica (Table 5). Although R. pacifica extends into the tropics to latitude 17°S it has been regarded asa mesotherm element because of its world distribution pattern which indicates its overall range is in the subtropics (Stevens, 1983). The presence of R. exiguella, R. peruviana, and R. inflata subsp. perpusilla in the tropical regions at higher elevations than sea-level is similar to the occurrence of subtropical or even temperate phanerogams which are reported to find an optimal thermal environment with increase in elevation at low latitudes (Nix, 1982). Two taxa which have restricted distribution in the mountains at altitudes exceeding 900 m are R. filicaulis and R. nervulosa var. dumeticola. Being rare in occurrence it is difficult to establish in what thermal group they belong. They have been tentatively placed as mesothermic plants as the former occurs in areas where R. celastri subsp. celastri, R. peruviana, and R. inflata subsp. perpusilla (all mesothermic plants) grow; and the latter must be a mesothermic offshoot of the otherwise megathermic R. nervulosa group. The distribution pattern of R. inflata subsp. perpusilla covers a region corresponding to RAMALINA IN AUSTRALIA 137 DISTRIBUTION PATTERN OF MEGATHERM PHANEROGAMS SS Growth index values > 0.45 E53 Growth index values > 0.30 < 0.45 (After Nix, 1981) DISTRIBUTION PATTERN OF MESOTHERM PHANEROGAMS TI) Growth index values > 0.45 anit] Growth index values > 0.30 < 0.45 (After Nix, 1881) Fig. 4 A. Distribution patterns of megatherm phanerogams with G.I. values >0-30; B. Dis- tribution patterns of mesotherm phanerogams with G.I. values >0-30; C. Ramalina distribu- tion pattern with megatherm and mesotherm plant boundaries for plants with G.I. values >0-30, superimposed. G. NELL STEVENS 138 xX Gee, Sa 4 pjjaidsavo "yf >, >, 1AD4AU1YM *Y >. CO, Grp, < XOX SEXO ae syp4aqopiun *Y SX XOTIXS XS FX ppiuquy *y een ee ae syvao “dsqns 14380]99 *Y GP. G.€ XK XX SISUA1ADUDD *Y a ee ey ek, Se vssyf “MY a ee Fe eek oe xX syvaisno ‘dsqns vjopfur “y x =X X pjonpad *Y > > Gee 4 > OR ae ab Ca Same > ae 8 suaosaonv]3 "Y > i Sea So. CP Se. Se ee Am, Gao. Ge x pioyui ‘dsqns vjopful “yy SD Gee 4 DJOINAUNDP "IVA VDSO]NAAIU "Y > 4 dt D4, 4 CX x S1SUdI]DAISND “Y x ok Oy a ko a SO ee a x 14180129 ‘dsqns 14380]99 *Y x >A x synvoiyf “Yy > >. ie ae, ae, Com Cae. Gay ea Cae. 4 vapola] “IeA DaUIXDA{GNS *Y x XEx MN ee Ok XEN ee voyiovd *Yy bee. Ke ee MOK. Oe xX 26 5X > ag. $ puvianiad ‘Y x > Ge tm eae Ge GP the ae Bb ay oes » eee pjjisndaad ‘dsqns viopful “y er Na ee, CoG Cee OC AK ke xX x 4 pjJansixa *y xX Woo OM X OY avion] “IRA DSOjNAsaU *Y xX > ie § x vas0}1] “YM 4 x b GR ae GE CES Ce aero AD Om ee Lap aap. Gad. Sam ¢ x DIDUAYUOD “eA DaUIXDA{GNS *Y x a ae ae 4 Wa SO NN DSO]NAAIU “IRA DSO]NALIU "Y VX. DOEK = FOR Sam anes DININSIOU “IRA DaUIXDA[GNS "Yy ee Xe ee eee ak DaUIXDALGNS “IRA DaUIXDA{GNS *Y ae ae ae 4 » >, eed DIJ9UA] “Y » Oe Ces >< voido4l Y bh fv Ch Ib OF 6E BE LE YE SE HE CE CE TE OE 62 BC LZ 97 ST HT ET CC IZ OC GT BI LI OT ST PT ET Cl ~S.opmney ‘BXE} DUIJDUUDY UeTeISNY 34} JO UONNQUISIpP [RUIPNINeT ¢ aqVyL RAMALINA IN AUSTRALIA 139 mesotherm response patterns with G. I. values above 0-30. The inland areas with G.I. values >0-30 <0-45, correspond to areas where the sekikaic acid race as well as the divaricatic acid race occur. Along the coast, the areas with G.I. >0-45 mainly produce thalli with divaricatic acid + a trace of sekikaic acid. To the south, the mesotherm region with G.I. values >0-45 coincides with the R. inflata subsp. inflata distribution (divaricatic acid alone or rarely + trace sekikaic acid). No specimens of the R. inflata complex occur in the region delimited by the mesotherm elements with G.I. >0-30 <0.45 in inland New South Wales. In the temperate southern and south-western regions of Australia R. inflata subsp. australis occupies this mesotherm region (G.I. >0-30 <0-45) and again the sekikaic acid race is present at inland sites, as well as the divaricatic acid race. Along the coastline only the divaricatic acid race is present which is a similar situation to that found with the taxon R. inflata subsp. perpusilla. Thus in the R. inflata complex, only taxa containing divaricatic acid + trace sekikaic acid are found along the eastern and southern coastlines. In the south-west corner of Western Australia both divaricatic acid and sekikaic acid taxa occur, reflecting the general aridity of this part of Australia even close to the coast. There is a continuation of both mesotherm elements into Tasmania (Fig. 4B) and both R. inflata subsp. inflata and R. inflata subsp. australis occur, but in different regions. The R. inflata subsp. inflata taxa occur in the inland montane region whereas R. inflata subsp. australis taxa are found in the drier eastern areas (Figs 21, 22). Sekikaic acid taxa do not occur in Tasmania (extreme dryness is not a feature of the island). A microtherm influence could be responsible for the distribution pattern of the R. inflata subspecies in Tasmania and in many instances the presence of intermediate morphs causes difficulty in delimiting Tasmanian taxa. Another taxon placed in the mesothermic group is R. celastri subsp. celastri. The majority of collection sites for this taxon lie within the belt with G.I. values >0-45, with a few collection localities to the west in the area covered by the mesotherm belt with G.I. >0-30 <0-45. Specimens from the latter area are always broader than coastal specimens. This taxon extends south to latitude 38°S with lessening frequency. At approximately 38°S, 145°E the mesotherm region of plants with G.I. >0-45 is replaced by the region of plants with G.I. >0-30 <0-45 (Fig. 4B). This latter region is occupied by R. celastri subsp. ovalis, a taxon which extends to Western Australia (Fig. 20). It is also found in northern Tasmania. Ramalina australiensis occurs coastally in the mesotherm pattern with G.I. value >0-45 (e.g. Toogoom in Queensland and Yamba to Twofold Bay in New South Wales, latitudes 25°S and 29°-37°S respectively). The main distribution of R. canariensis is along the southern coastline of Australia in the region with mesotherm G.I. values >0-30 <0-45 which corresponds to a Mediterranean climate region. Such environments are found in South Africa, Chile, and the Mediterranean (Raven, 1973) — all areas where this taxon occurs. The presence of R. canariensis at several sites along the south-east coastline of Australia may be due to suitable microhabitat conditions. The two species found on the Bass Strait islands (viz. R. whinrayi and R. caespitella) cannot be ascribed accurately to any of the three response groups. (3) The microtherm element There are four taxa which occur in the mountainous areas of eastern and/or southeastern Australia at altitudes exceeding 600 m which can be regarded as belonging to the microtherm response group: (1) R. reducta, (2) R. glaucescens, (3) R. unilateralis, and (4) R. fimbriata. (1) Ramalina reducta occurs at elevations over 1000 m in open-forests on the New England tableland. The area experiences very low winter temperatures with snow. Figs SA—B show that it covers a region with G.I. values >0-2 <0-6. (2) Ramalina glaucescens is classed as a microtherm element with its two chemical races occupying two different response groups (i) G.I. >0-4 <0-6 and (ii) G.I. >0-2 <0-4 the former region being occupied by the sekikaic acid race and the latter by the acid-deficient race (Figs 5B, 13). (3) Ramalina unilateralis occurs inland and along the coast at several sites, but its distribution 140 G. NELL STEVENS QV Microtherm GB CI> 0.6 EZZ7ACIt> 10045046 EE9 CI> 0.2<0.4 ee E=a6r 2-6-2 E Cs BZ Fig. 5 0-2 (after Nix, 1981). pattern corresponds with the microtherm element rather than the mesotherm element. Its range covers a region where microthermic G.I. values are >0-2 <0-6 (Fig. 5B). (4) Ramalina fimbriata grows on rock outcrops at high altitudes in the open-forest country of temperate southeast Australia (Fig. 5B); its range covers a region where microthermic G.I. values are >0-2 <0-6. Biogeography Taxa in the genus Ramalina found in Europe and North America are generally distinct from the taxa which occur in Africa, South America, Australasia, and the Pacific Islands. The records of Ramalina taxa which occur in the various regions of the world are scattered through the literature. Only the few present day revisions of the genus can be relied on to contain correct identification of the species; many publications record names of doubtful application. Those regions covered in recent reliable publications are: Canary Islands (Krog & @sthagen, 1980), East Africa (Krog & Swinscow, 1974, 1975, 1976), Fennoscandia and the British Isles (Krog & James, 1977), Europe (Poelt, 1969), and the West Indies (Landrén, 1972). Although the Ramalinae of ‘North America have not been completely revised, many of the taxa have been recently investigated by Hale (1978), Rundel (1978a), Bowler (1977), Bowler & Rundel (1972a, 1972b, 1973, 1974, 1978), Rundel & Bowler (1974, 1976), Hale & Culberson (1970), and Moore (1968). Less recent studies cover the Hawaiian lichens (Magnusson & Zahlbruckner, 1945; Magnusson, 1956); the Chinese Ramalinae (Zahlbruckner, 1930) and the Japanese Ramalinae (Asahina, 1938, 1939). The South American Ramalinae are in need of revision but records of taxa collected in various regions of that continent appear in publications of Vainio (1890), Malme (1934), Follmann RAMALINA IN AUSTRALIA 141 Ci ‘3 3 NG Fig. 5B. Four Ramalina taxa with distribution patterns within the microtherm boundaries: 1 = R. reducta, 2 = R. glaucescens (sekikaic acid race), 3 = R. unilateralis, 4 = R. fimbriata. (1967), Rundel (19785), Osorio (1970a, 1970b, 1972, 1978), and Osorio, Aquila & Zanette (1980), as well as others. Papers with Ramalina taxa found on the Pacific Islands are listed in Appendix A of Lichen Ecology (Hawksworth, 1977). The ‘Catalogue of lichens from India, Nepal, Pakistan and Ceylon’ (Awasthi, 1965) was consulted to obtain species numbers in India, as no information concerning the present revision of Indian Ramalinae was obtainable from India. Nine Ramalina taxa are described for New Zealand (Galloway, 1985). From the above mentioned records and personal knowledge, some idea of the global distribution of the taxa which occur in Australia was obtained (Table 6, Fig. 6). Table 6 shows that Africa has the greatest number of species in common with Australia (50%). This close relationship with African lichens was mentioned by Rogers & Stevens (1981). New Zealand and South America have the next highest percentages of species in common with Australia. It is 142 G. NELL STEVENS Fig. 6 Global distribution of Ramalina species found in Australia. interesting that (as far as could be ascertained) there is little similarity of species between India and Australia (see below). The lack of correlation between Australian taxa and those of North America, Europe, and Asia, confirms that the greatest similarities are found in countries which have a former Gondwanaland origin. Palaeobiogeography Before presenting an interpretation of the likely palaeobiogeography of the Ramalinae, an understanding of some aspects of the Earth’s geological history has to be appreciated. (1) Geological history of Gondwanaland The geological events that took place during the Cretaceous and Tertiary Periods which are so vital to the understanding of the distribution patterns of plants are summarised in Table 7. The early ideas of land bridges between continents and the later theory of Continental Drift have been superseded by the theory of Plate Tectonics. This theory is now generally accepted by most geologists to account for the separation of continents which show evidence of having been once grouped together as a Pangaea or super-continent, or as two large continents, Laurasia (north) and Gondwanaland (south). In this theory, the continents break up by rifting and separate by sea-floor spreading. New oceanic crust is formed at spreading ridges and consumed in subduction zones. The theory of an expanding earth (Carey, 1976) obviates the necessity for subduction zones and produces a more neatly-fitting Pangaea than other models, especially in the closer fit of northern and north-west Australia with Asia, making the distribution of now distant but similar biota more easily understood; unfortunately this theory has not had general acceptance. (2) Plant distribution related to geological events The map of the Gondwanaland continent during the Cretaceous Period (130-100 m.y. B.P.) [Fig. 7] shows the relevant palaeolatitudes at that time (Powell, Johnson & Veevers, 1981). Raven & Axelrod (1974) stated that direct migration from South America and Africa via RAMALINA IN AUSTRALIA 143 Table6 The global distribution of the Ramalina taxa occurring in Australia. Islands of the Pacific Ocean Islands of the Indian Australia Africa South America New Zealand India/Sri Lanka Ocean Tristan da Cunha Europe North America West Indies China Ramalina species * australiensis caespitella canariensis exiguella filicaulis fimbriata glaucescens litorea peruviana reducta tenella unilateralis whinrayi celastri subsp. celastri subsp. ovalis inflata subsp. inflata subsp. perpusilla subsp. australis R. fissa R. nervulosa var. nervulosa var. luciae var. dumeticola R. pacifica R. subfraxinea var. subfraxinea var. leiodea var. confirmata var. norstictica R. tropica *** *£© &©* © & & & & He HR F * & © * * ADDDAAAAAAAAARHA * * * * * * * * * * & & & * * * * * * * * * & & & * * * &* & & & * * TOTALS 28 14 8 9 4 9 3 3 1 3 3 i Antarctica to Australia was possible at this time (Table 7). Even in the late Cretaceous, Australia still had connections with South America through Antarctica, but the migration of plants via the southern route probably became progressively reduced until finally stopped by the break of Australia from Antarctica in the Eocene (53-38 m.y. B.P.). The early break-away of India from the rest of Gondwanaland (125-100 m.y. B.P.), and its rafting north across the equator with the loss of some Gondwanaland species, could account for the lack of taxa similar to the Australian Ramalinae. Two major pathways of dispersal of the higher plants from west Gondwanaland have been postulated by plant biogeographers: (1) a temperate migration route via Antarctica from west Gondwanaland (i.e. South America and Africa), to Australia and New Zealand, and (2) a tropical route from East Africa to Madagascar, the islands of the Indian Ocean and Indo- Malaysia to Australia (Croizat, 1952; Raven & Axelrod, 1974). As the Ramalina species found in Australia can be placed in a range of temperature response groups from cool temperate to 144 G. NELL STEVENS Table 7 The history of the break-up of Gondwanaland, indicating available migration routes (from Raven & Axelrod, 1974 and Powell, Johnson & Veevers, 1981). Million years BP Geological periods Event/geographic situation 125 early Cretaceous Africa and South America begin to split; India begins to split from Australia/Antarctica but remains attached to Madagascar. 110 + 10 Direct migration between west Gondwanaland and Australia ceases. 100 mid Cretaceous Direct migration between Africa and Madagascar still possible. India rafts northwards and is completely separated from Australia and Antarctica; no exchange of species possible. Migration of species between Africa and Australia via east Antarctica still possible. 90 Africa and NE. Brazil separated by a narrow strait; migration possible at this point. Africa and South America separated by 800 km of ocean but numerous islands linked them along the mid-Atlantic Ridge. Australia ceases to have any migration from Africa. 80 late Cretaceous Australia/Antarctica extend from 40°S to the polar region. South America closely connected with Antarctica 80—30 million years BP. Separation of New Zealand and New Caledonia from Australia. 63 early Palaeocene Africa and Europe connected at Spain and possibly with Asia at Arabia, allowing both northward and southward migration. 55 Palaeocene-Eocene Oceanic crusts form between Antarctica and Australia; northern edge of Australia now at 30°S and commencement of separation of Australia from Antarctica. 49 mid Eocene Separation of the continental margins of Australia and Antarctica but another 10 m.y. of migration of species from South America through Antarctica to Tasmania along the South Tasman Rise before final separation. 45 Direct migration between India and Asia. 30 mid Oligocene The leading edge of Australia as it rafts northwards still south of the latitude of the Sunda Arcs. 20 early Miocene Initial contact between Sunda Arc and New Guinea (abundant precipitation). 10 late Miocene Collision of Australia with the Sunda Arcs. Antarctic ice-sheet rapid expansion (lowered precipitation). Present Holocene Australia moves towards the equator at 66 mm year’. tropical, it could be hypothesised that their routes of entry to Australia were from both the south (the microtherm and mesotherm taxa) and from the north (the megatherm taxa). (3) Dispersal and migration Any postulates concerning the migration pattern of the Ramalinae from a centre of origin are purely hypothetical as no fossil records are available and any assumptions made are based on the migration pathways of the higher plants deduced from fossil records. The approach used in this study is similar to that of Jorgensen (1983) and Sipman (1983) although the writer does not agree with all of their conclusions. Biogeographic evidence suggests that many genera and families of gymnosperms and angiosperms had evolved before the breakup of Gondwanaland began in the Cretaceous, and had arrived in Australia from South America and Africa via Antarctica during the middle RAMALINA IN AUSTRALIA 145 Fig. 7 Position of east and west Gondwanaland in the early to mid-Cretaceous (130-100 m.y. B.P.). Ticks show longitudes which have no fixed origin (Adapted from Powell, Johnson & Veevers, 1981). Cretaceous (108-100 m.y. B.P.). Fossil evidence indicates that many of the Gondwanan families and genera had occupied Australia by the early Tertiary (63-55 m.y. B.P.) (Specht, 1981c). (a) Step by step dispersal: Schuster (1979) stated that migration of land plants normally tends to be by short range dissemination as members of structured communities, and he proposed that the dispersal of Hepaticae occurred as part of the dispersal of whole plant communities. It is also reasonable to hypothesise that the Ramalina species also made this step by step spread across the continents of Laurasia and Gondwanaland with the migrating structural communities during the Cretaceous-early Tertiary. This suggested step by step migration of plant communities appears to be a valid means of dispersal as this enables plants which have similar habitat requirements to expand as a whole, in favourable climates, and so migrate when conditions are favourable and disappear from areas where environmental conditions have deteriorated. It seems highly likely that under the circumstances lichens which grow within these communities would migrate at the same rate and in the same direction. If step by step migration halted with sea-floor spreading of the continents, then another means of dispersal has to be proposed. Seeds of the higher plants have been found to be carried by birds and animals over long distances; the lighter seeds being wind dispersed. The Ramalinae produce both spores and vegetative diaspores and it is feasible that these could be wind dispersed or carried by birds. (b) Wind dispersal: If long distance dispersal of spores by the wind is accepted then this means of dispersion would have been in operation in early times. It is reported by Raven & Axelrod (1974) that with the separation of Australia and Antarctica in the Eocene, there was a 146 G. NELL STEVENS strengthening of the circum-Antarctic wind which could have carried spores from South America to Australia, Tasmania, and New Zealand. The success of a spore in a new environment would need the availability of a suitable phycobiont and the lack of competition from plants already in the area. With the vegetative diaspore both partners arrive together so that provided the environment is hospitable to their genetic requirements they will readily become established in the new habitat. Spores can become established after long distance wind dispersal since fertile lichen species have been found on the islands of Hawaii. Jorgensen (1979, 1983) remarked on the efficiency of wind dispersal of lichen spores in the population of the remote island of Tristan da Cunha. If the original dispersal of the Ramalinae is comparable with the original dispersal of angiosperms and gymnosperms, then their migration from west Gondwanaland may have taken place by both step by step migration and wind dispersal, with the latter means continuing after the break up of the continents. (4) Centres of origin The centre of origin of a group is often regarded as that area having the greatest concentration of species. Hale used this method in assessing the centre of origin of the genera Relicina (Hale, 1975), Pseudoparmelia (Hale, 1976a), Parmelina (Hale, 1976b), and Bulbothrix (Hale, 1976c). If the numbers of Ramalina taxa occurring in each landmass that once formed Gondwanaland are estimated (and these numbers must be only approximate as without recent revisions many names could be synonyms or misidentifications) then it is found that 41 Ramalinae occur in South America, 42 in Africa, 36 in India, and 30 in Australasia. South America and Africa (once West Gondwanaland) appear to have the largest number of Ramalinae and could be regarded as the centre of origin of the southern hemisphere taxa. The occurrence of high numbers of taxa within the one genus, in a particular region, can, however be due to a re-radiation of species after the original dispersal of the taxa. When an environment is free of competitive species, newly evolved taxa belonging to the genus will become established. This situation has been found in Western Australia amongst the eucalypts, where secondary radiation of species has occurred in a region where there is little competition from other genera (Pryor, 1981). Such a situation may have arisen in Australia with the evolution of the mangrove flora in this region (Specht, 1981b). Mangroves are the major phorophytes for the present day R. subfrax- inea complex. It is proposed that the ancient stock of the R. subfraxinea group could have entered Australia via the northern route from East Africa (as this group is regarded as a megatherm response group) and become established along the Australian tropical coastline on the newly evolved mangroves. The occurrence of all six chemotypes in the region around latitude 18°S indicates an area of speciation from where secondary radiation could have spread to the east and west. (5) Vicariant taxa The Croizat theory of vicariance biogeography (Croizat, 1952; Croizat, Nelson & Rosen, 1974) postulates that species which are ecologically similar but occur on different land masses have a common ancestry. They are regarded as disjunct populations of once continuously distributed taxa which have undergone speciation as physical disruption occurred and Croizat cited many Gondwanaland taxa to support his theory. Galloway (1979) employed the same ideas in his comments on the distribution of species in the genera Usnea, Pseudocyphellaria, Psoroma, and Stereocaulon. A number of taxa belonging to the Fistularia group in Australia appear to be part of the vicarant taxa of Gondwanaland origin. For example the Australian taxon R. inflata subsp. perpusilla has an ecotype which grows on mangroves and closely resembles the East African species R. consanguinea which also occurs on mangroves with the main difference in the spore size; in some habitats in Australia R. inflata subsp. inflata morphology closely resembles some forms of the East African taxon R. calcarata, but their chemistry differs in the minor constituents. RAMALINA IN AUSTRALIA 147 Chemical difference is the only character which separates the Australian taxon R. filicaulis from the American taxon R. anceps. Material similar to the Australian material, and also lacking medullary acids, has been cited from South Africa. It is proposed that all three continents possess species which could be regarded as vicariant taxa belonging to the R. usnea complex, as they are ecologically and morphologically similar. On the basis of the evidence discussed above it is postulated that Australia has been populated by species of Ramalina since the late Cretaceous. Since west Gondwanaland is accepted as the centre of origin for the angiosperms, based on fossil evidence, there is the possibility of these acting as phorophytes for ancestral taxa of the present day Ramalina, as that same region is a likely centre of origin for Ramalina. The step by step spread of the plant communities would have enabled the Ramalinae to spread by one of two migration routes either (1) the southern route from South America (after Africa had separated) via the Scotia arc, Antarctica to Tasmania and Australia or (2) the northern migration route from East Africa via Madagascar, the Mascarenes, Indo-Malaya to northern Australia, by which means the megathermic Ramali- nae could have reached tropical Australia. Once established in Australia the megatherm Ramalina taxa remained a maritime group as inland habitats would have been inhospitable. Secondary radiation from Australia by some of the megathermic taxa appears to have occurred with the spread of species to the Pacific and Indian ocean islands. The microtherm Ramalinae became established in the mountainous region of southern Australia since these habits would have suited their thermal response patterns as Australia drifted northwards and the temperatures increased. The climatic changes that occurred in the late Miocene Period caused the desiccation of the continent. Subsequent speciation of mesotherm Ramalinae could have occurred producing taxa which occupied the new and dryer habitats of the south-west and west of Australia. In this way the present distribution pattern of the Australian Ramalinae reflects their palaeobiogeographic and palaeoecological history. Taxonomy The delimitation of Ramalina species proved difficult owing to the amount of morphological variation shown by each taxon. A large amount of material had to be examined in order to ascertain what concepts should be applied in the delimitation of each taxon. Three taxonomic ranks have been used. Taxonomic ranks (1) Species: The rank of species has been given only to those taxa which can be recognized by several distinct morphological traits which collectively produce a particular thallus type (allow- ing for an appropriate amount of variation), together with a set of specific anatomical and chemical properties, and which show a distinct habitat and distribution range. Difficulties arose when two or more chemotypes produced a single morphotype; this problem had to be resolved by considering the habitat and distribution range of the taxa, and if the only difference found was in the chemistry then these chemical races were treated as conspecific. The concept of species pairs (Artenpaare), as proposed by Poelt (1970, 1972) is that species which reproduce by vegetative methods (secondary species) have been derived from existing or extinct fertile species (primary species), and that primary and secondary species have identical (or closely related) chemical components. This view is held by most lichenologists and species pairs have been found in many genera e.g. in the Physcia and allied genera (Moberg, 1977), Parmelia sect. Hypotrachyna (Culberson & Hale, 1973), Dirina (Tehler, 1983), and to a lesser extent in Alectoria (Brodo & Hawksworth, 1977). Only a few records of species pairs have been found in the Ramalinae. Rundel & Bowler (1976) discussed R. leptocarpha and R. sublectocarpha as a fertile-sorediate species pair and Krog & @sthagen (1980) have commented that the fertile species R. implectens may be regarded as the parent morph of the sorediate species R. farinacea. 148 G. NELL STEVENS None of the Australian sorediate Ramalinae is regarded as derived from any of the fertile taxa found in Australia today. (ii) Subspecies: The delimitation of a taxon based on well marked discontinuities in characters was not always possible. Some Ramalina taxa were found to intergrade into each other, differing only subtly morphologically yet extending over large areas. When such complexes were examined carefully the most obvious morphological changes were found to coincide with different geographical regions. The term subspecies has therefore been applied to a taxon which could be shown to occupy a particular region within the total distribution range of a large complex (as defined by Hawks- worth, 1976). This taxon can be extremely variable within its geographical confines and can produce several ecotypes which occupy separate habitats. As subspecies intergrade into each other, their delimitation is much more arbitrary than that of a species, and depends more on geographical divisions than on chemical or morphological differences. Imshaug & Brodo (1966) pointed out that the use of the rank of subspecies permitted a classification which is not only practical but one which indicates a possible phylogenetic system. The evolutionary relationships that may exist in some of the Australian taxa have been indicated by the use of the rank of subspecies. (iii) Variety: The rank of variety has been used when more or less morphologically identical taxa have been found to produce different chemistries which correlate with distinctive distri- bution patterns. The production of either divaricatic acid or sekikaic acid in populations of the one taxon posed a nomenclatural problem. Several lichenologists have mentioned the same difficulty e.g. in the genus Dirinaria (Jorgensen, 1974) and in the genus Ramalina (Imshaug, 1972; Rundel, 1978a). The solution has been to regard both acids as representing chemical races within the one taxon. Hybridization It is not easy to establish the genetic homogeneity or heterogenity of lichenized fungi. Jahns (1974) pointed out that thallus or hyphal fusions do occur in lichens, which could enable a form of hybridization to take place. A specimen of R. leiodea recorded by Elix (1982) as being a ‘schizophrenic thallus’, had ‘left-hand branches containing one acid, the right-hand branches another and the central branches containing a mixture of both’. In this case the two acids involved were two depsides, not closely related (cryptochlorophaeic acid and boninic acid). Another such combination of two acids in one specimen belonging to the R. subfraxinea complex has been found during the present study; the specimen was collected at Hinchinbrook I. where all acid races of the R. subfraxinea complex are present. The thallus concerned contained both boninic acid and salazinic acid, one a depside and the other a depsidone. As these two acids are not biosequential metabolites it appears likely that this is a similar form of hybridization to that found by Elix (1982), and that both could be regarded as ‘chimeras’ produced by the fusion of hyphae from two different spore types. It is not known what chemistry the thalli produced from spores of this plant would have. No other thalli with this combination of acids has been found. The occurrence of sekikaic acid in the divaricatic acid race of R. subfraxinea var. subfraxinea is regarded as different from the above. Sekikaic acid and divaricatic acid are reasonably closely related biosequentially and apparently can interconvert (Elix in litt.) which would result in chemical combinants if the conversion was incomplete. However the hypothesis of interspecific hybridization (Brodo, 1978) remains feasible and has to be considered beside the hypothesis that incomplete conversion of a precursor acid occurs during the biosynthetic steps from one acid to another (Bowler & Rundel, 1978). If hybridization is accepted as the cause of this mixture of acids then it would be natural to expect the proportions of each acid to be equal in the thallus. However, this was not the case with the taxa examined; in all instances divaricatic acid occurred in far larger quantities than did the sekikaic acid. It is therefore suggested that incomplete conversion of one acid to another has taken place. RAMALINA IN AUSTRALIA 149 Table 8 Taxonomically important characters in the Australian Ramalina taxa. major acid constituents orcinol depsides B-orcinol depsidones substrate thallus para- meta- cryptochlorophaeic saxicolous inflated sorediate divaricatic evernic sekikaic boninic salazinic norstictic psoromic Ramalina species >| nil acids . australiensis . caespitella . canariensis . exiguella . filicaulis fimbriata glaucescens litorea . peruviana . reducta . tenella . unilateralis whinrayi Species complexes R. celastri subsp. celastri subsp. ovalis R. inflata subsp. inflata subsp. perpusilla subsp. australis R. fissa R. nervulosa var. nervulosa var. luciae var. dumeticola R. pacifica R. subfraxinea var. subfraxinea var. leiodea var. confirmata var. norstictica R. tropica mx * ~ mx * xx > > +! Corticolous Pe ~< ~~ X D4 4 D4 Od DS OS DS DS DS DS DX | Solid ~ mK XM DDD DW WD RWDWD WDA ~*~ x ~*~ x mK mK mK »* mK x ~*~ ~ mK xX ~ KK ~ X Xx mmm MM KKM Mmm MM * * X Description of the genus Ramalina Ramalina Ach., Lichenogr. Univ.: 122,598 (1810), nom. cons. Thallus fruticose, greenish-grey, pale green or greenish-yellow, caespitose, erect, subpendulous to pendulous; branching dichotomous, subdichotomous to irregular or dense intricate; branches few to many, twig-like, strap-like to palmate, flat, subterete or terete, solid or inflated; apices forked, attenuate, rounded; surface shiny or matt, smooth or rugose, pseudocyphellate; holdfast delimited or dispersed; soralia small and punctiform to large and round or elongate; 150 G. NELL STEVENS cortex composed of two layers, the outer cortex hyphae more or less anticlinal or randomly oriented, the inner cortex hyphae periclinally arranged; hyphae surrounded by a large amount of matrix; medullary hyphae loosely interwoven between the upper and lower cortex or confined to the algal zone. Apothecia terminal, subterminal, marginal, lateral, laminal; convex, plane or concave, sometimes incised; disc pale to flesh coloured; margin concolourous with thallus; asci clavate, containing 8 spores; spores hyaline, oblong, ellipsoid or fusiform, straight or curved, 1-septate. Pycnidia inconspicuous, pale, usually rare in Australian species. The algae are Trebouxioid, globose, forming an irregular ring along the edge of the inner cortex; when confined to the upper side only the thallus becomes dichroic with a white lower surface. Type species (conserved): Ramalina fraxinea (L.) Ach. Enumeration of the Australian taxa The Australian Ramalinae comprise 28 taxa which can be grouped into 17 fertile and 11 sorediate taxa. Sixteen species have no subspecific taxa recognized, and four species contain a total of 12 subspecific taxa. Table 8 lists the taxonomically important attributes of each of the 28 taxa; those belonging to species complexes are placed after the alphabetically arranged single species, as is also done in the description of the taxa. The disposition of the other Ramalina taxa recorded by Wetmore (1963) and Weber & Wetmore (1972) is shown in Appendix 1. The specimens examined are cited with location, latitude and longitude in degrees, type of substrate, date, the collector’s name and the herbarium abbreviation (Holmgren, Keuken & Schofield, 1981). Collectors initials have been used for those collectors whose specimens are cited most frequently in this work. These names have been abbreviated as follows: AA A. Archer JC J. Cashin ED E. Dahl JW J. S. Whinray GB GoC. Bratt NCS N. C. Stevens GK G. Kantvilas NS N. Sammy GNS G. N. Stevens RF R. B. Filson HS H. Streimann RS R. D. Seppelt JAE J. A. Elx RWR R. W. Rogers Only one specimen is cited from each location for most of the taxa; where the taxon grows profusely over a wide area, only specimens from selected sites are listed: viz, R. celastri subsp. celastri, R. celastri subsp. ovalis, R. inflata subsp. inflata, R. inflata subsp. perpusilla, R. inflata subsp. australis, and R. fissa. Key to Ramalina in Australia la. Thallusinflated (partially or totally). c0cc5 vse sees step castes 0 ce cn water ieecayenesaoeseee recone 35 ib: Uhalhws mori ated Saka sthek eee es es coe Ree Pe pentcr ties sue oct tee a nan cea nao 2 De (1b): Plant SaxicOlous. os. s-c052. so serccaes so Aipocsecae ccs san ssnecusanae Ropes bene austadnaeenr te duekoads Beeeds 3 2b. PIANECONTICOIOUS «29 oecnetwteciev se tate een tadaccutis des sncuncccoauchrenvnrwerotanereeriets 12 Sa (Zaye Dhalltis SOTECIAtE co5.o coe os seas fens eon sade teks fe emicibei aa pe alot eren ae tise SACRA EMO ONe OU: yamoe nated 4 3b. Ralls MOCSOLECIALG A ecco nse teen oe ot tata ons oe ce nee ach eetigecinis tun oseeeae pe anginne? 5 4a (3a). Soredia eruption through eroded lower surface and at apices ......... 6. R. fimbriata (p. 161) 4b. Soredia contained in well defined soralia, which are marginal and sometimes laminal 2. R. caespitella (p. 155) 5a, (50). Phallus made up Ola few 1OO8e DIGNCHES. 5. ¥,..cs.tacacosscesnery oxioes eventos orenacunadanateczee sts 6 Sb. Thallus composed of numerous branches forming a dense cushion ................0c:0eeeees 8 6a (5a). Branches compressed, flat (apothecia marginal) ...... 14. R. celastri subsp. celastri (p. 180) 6b. IBTABCNES SUDTCTOLE TO LOLOLS fociccdsd.sca cs noinisen stboiaspwansts aesgiseriaees sss enesenrpernentnntsesa i 7a (6b). Branching dense, distal, K— (no medullary acids) .................. 1. R. australiensis (p. 152) 7b. Branching sparse, K+ red (salazinic acid) ...............sscsesessceseseeees 20. R. tropica (p. 210) 8a (5b). 8b. 9a (8a). 9b. 10a (8b). 10b. 11a (10a). 11b. 12a (2b). 12b. 13a (12a). 13b. 14a (13b). 14b. 15a (14b). 15b. 16a (15b). 16b. 17a (16a). 17b. 18a (16b). 18b. 19a (18b). 19b. 20a (19a). 20b. 21a (12b). 21b. 22a (21a). 22b. 23a (21b). 23b. 24a (23a). 24b. 25a (24a). 25b. 26a (25b). 26b. 27a (26b). 210; 28a (27b). RAMALINA IN AUSTRALIA ove Ti oly Verh <6 (8 ONRRp re RES Nee Papen Be Sys anne EE oy Ne a oR Mca RiP ARRON Ey Beh Ae Branches inflated, Omen Sout 2200.74.55 cis pass- eos cu se oey fees sacsseeie eae oncineoncod tavanvnti eg ans Apothecia terminal, large 3—10 mm diam. ..............ccssoeesecteceens Apothecia subterminal (subtending narrow pointed branchlets, apothecia 1-2 mm diam.) Thallus showing some inflation; splitting and tattering of branches common .............. Teak US FOE WEL ACEO cas 2 hs cic ioe es dees atc kd Sans sean Nnnmenec ees un euguae aaiie Lecenedoae WMedulla K+ £60 (SAlAziOtG RCIA) ian ee a vese so sece cs cen sg heer Medulla K— or K+ pink (divaricatic acid or sekikaic acid) 15. R. inflata subsp. australis (p. SH ALTSISOTECIALC eee iae ee career cae ae aoa eee Te EE aes, eerie eats enone PNAS MOCSOTECIALC cee eit a eee ae rere oie a ch aI te eee tata toes Lamemare Lower surface of anastomosing strands resting on the medulla........ Lower surface not of anastomosing strands Only... 7c255. 0 arise .hd scien sa vensevccevnsdee seasons Branches few, usually broad at base (soft textured, soralia apical and marginal) 3. R. canariensis (p. Branches numerous, narrow Branches usually cartilaginous (splitting open at intervals along their length) 12. R. unilateralis (p. Branches not cartilasinous, Completes: co.cc sh sscnes Goenan ave ue eeaonsninda sabesoees aabeds Meédulla K+; red (salazinie acid). 3c aeewa visa sewed oiesace tee ges Dione pnetnee eens ces oa a Medulla K— or K+ pink Soralia punctiform, mainly apical on branchlets (thallus up to 3 cm high) 11. R. tenella (p. Soralia ellipsoidal or round, marginal on elongate branches (thallus usually 4-8 cm long) Soralia ellipsoidal or round, marginal on elongate branches ...................00ccecseeeeee ees Soralia punctiform, marginal and apical (branching dense, intricate, thallus resilient in texture) 9. R. peruviana (p. Medulla K+ pink (sekikaic aggregate acids) Medulla K— (divaricatic acid and stenosporic acid) 17. R. nervulosa var. nervulosa (p. Sekikaic aggregate acids + faint ramalinolic acid Sekikaic aggregate acids + faint homosekikaic acid .... Seem mee meee eee meee meee eee e eee eee eeeeeeeeseeeeeeeesseesssessssEssssessesseee Branches flat, sometimes canaliculate, containing norstictic acid (spores 12-16 x 4—5 MATE Sale Parse erica cies oan ote Ove anon ees Branches usually subterete, never canaliculate, containing salazinic acid (spores 14-22 Me BN MON) oy sascitos atevex sa vestadddiere Gee Miswies ra ere eed ack tale seeae as Branches subterete to terete, containing no medullary acids ..................:ccceceeeeee eens Branches fiat, with'or without medullary acids .22..<..ccscccssancrenvansiwatencteoateasnareaneade PLANS Bo an vores ie tacercd oaalnn Gores combine viouaiek ieyiecusrnpnlaeicus ov owe ptaee Rac IRR AS Cedar e SORE aaa aYe Medulla K+ pink Apothecia lateral and subterminal with branch apex appearing as a long spur (some- TINGS DISK MOVE): cers waco ere toes eyiniendicty ores ceeaasseeencsnntes Apothecia never subterminal nor bearing a spur Distal branchlets occur on otherwise unbranched primary branches 1. R. australiensis (p. Dense distal branchlets not present -[oc6. sec ssecs nose cous thy sen cre edeccdaderanee heevers ene vises Thallus thread-like (with dichotomous branching) Thallus fruticose Apothecia marginal on both edges (branches may be canaliculate, no medullary acids) 14. R. celastri subsp. celastri (p. 13. R. whinrayi (p. 8. R. litorea (p. 18. R. pacifica (p. 17. R. nervulosa var. dumeticola (p. 17. R. nervulosa var. luciae (p. 19. R. subfraxinea var. norstictica (p. 20. R. tropica (p. 4. R. exiguella (p. 152 G. NELL STEVENS 28b. Apothecia not matginal On DOUR CORES «<< tec ceassne sens va ndunvaet ones espe vespaaWagamas tenes ee 29 29a (28b). Apothecialaminal, numerous, and equal in size ......... 14. R. celastri subsp. ovalis (p. 184) 29b. Apothecia not laminal and numerous and equal in Size ...............eceecececeeseceeeeeeeseees 30 30a (29b). Branches broad with apices forked or lacerated, never simple and attenuate [apothecia in axils of branches (up to 5 mm diam.) or terminal on the broad margins of the apices, or small and sparsely spread laminally] ...................... 7. R. glaucescens (p. 163) 30b. Branches usually narrow with apices simple and attenuate or bent to form a spur below BPP OURC CUMIN oi sac wied th ars sates oceamisd cabenay RAMALINA IN AUSTRALIA ~ Ky, rey yrrrye IUNUERUUUVUAUUUINUNNNUUUVUUNIONOONTONUORUONUOUEURTOVOVIMM il OTOL DOL DOU TOOOVUOTDU ¢ Ton TITITITLI LIL 154 G. NELL STEVENS 0-5—1-0(—1-2) mm diam., plane to convex, margin entire; spores ellipsoid to fusiform, straight, 12-16(—20) x 4-6 wm (Plate 4, fig. 2). Chemistry. Usnic acid only. Remarks. The name R. australiensis as the oldest name has been upheld for this taxon, overriding the more commonly used name R. myrioclada. However the location given for the type material - Swan River, Western Australia — is dubious. All recent collections have been made in eastern Australia with no other material being found in the west. Whether the original material from Western Australia was wrongly recorded was not able to be ascertained. The type material of R. australiensis has branchlets with black tips but this feature (as with R. exiguella) is not present in most specimens. The type is a coarse thallus form of this taxon, rather resembling that of the subtropical Queensland collection. There is not sufficient type material to obtain an overall picture of the total thallus. The degree of secondary branching varies from specimen to specimen, those with dense branching resemble R. peruviana whereas others have sparse branching producing an open thallus form resembling the morphology of R. filicaulis N. Stevens. The names R. usneoides and R. usnea have been applied to some specimens of R. australiensis collected in New South Wales but the Australian taxon can be distinguished by its branching pattern and subterete branches instead of the characteristic flattened branches of R. usnea sens. str. Distribution and habitat. This subtropical/warm temperate maritime species does not have a 120 130 140 150 10- 20- a ote a TTS Ras 30- 40- ee | rn L rt Fig. 8 Distribution of R. australiensis @ and R. filicaulis *. RAMALINA IN AUSTRALIA 155 wide distribution (Fig. 8). It occurs along the coast from Yamba 29°S to Twofold Bay 37°S and has a disjunct occurrence on the Queensland coast at Toogoom (25°S). This Queensland occurrence cannot be adequately explained, although the rainfall in this area is similar to that which occurs along the New South Wales coastline. The Queensland material is extremely coarse, it was collected from Callitris columellaris which was growing on exposed sand dunes. The morphology of several Ramalina species collected from this area was also extremely coarse, indicating environmental adaptation. The major phorophyte of R. australiensi: is Avicennia marina, which grows in sheltered positions along the bays and rivers of the New South Wales coastline. However this Ramalina has also been collected from rocks at Port Hacking and Port Macquarie, which indicates it is not substrate specific. This taxon occurs on the offshore islands of the North Island of New Zealand which lie at equivalent latitudes to those where R. australiensis occurs in Australia. The species has not been recorded anywhere else. World distribution: Australia and New Zealand. Queensland: Toogoom, 25°15’S, 152°40’E, on Callitris columellaris, 1983, GNS (BRIU4217NS). New South Wales: Yamba, 29°27’S, 153°20’E, on Avicennia marina, 1977, GNS (BRIU 1978NS); Nambucca Hds. 30°41'S, 153°00’E, on Avicennia marina, 1977, GNS (BRIU1998NS); Port Macquarie, 31°27’S, 152°55'E, on metamorphic rock, 1975, JAE (JAE 1081); Patonga 33°30’S, 151°20’E, on Avicennia marina, 1978, GNS (BRIU2314NS); Gosford, 33°30'S, 151°20’E, 1978, JAE (JAE 4715); Newport, 33°40’S, 151°20’E, on mangroves, 1888, F. R. M. Wilson (MEL 9409); Lilli Pilli Beach, 34°04’S, 151°10’E, on sandstone, 1975, JAE (JAE 1181); Sussex Inlet, 35°10’S, 150°35’E, on Avicenna marina, 1977, RWR (BRIU1902NS); Buckenbowra R. 35°42’S, 150°06’E, on Avicennia marina 1983, HS (CBG8306040); Shoalhaven R. 34°52'S, 150°42’E, on Aegiceras corniculatum, 1975, P. Saenger (BRIU2137NS). 2. Ramalina caespitella N. Stevens Plate. fig, .2. in Lichenologist 18: 183 (1986). Type: Australia, Long Island (Bass Strait), 1969, Whinray s.n. (MEL! — holotype; sekikaic acid, 4’-0-demethylsekikaic acid, + several terpenes, and usnic acid). Thallus saxicolous, pale greenish-grey to stramineous, minute, caespitose, rigid, up to 1-0 cm high; branching subdichotomous, irregular, narrow thalli densely branched, broad thalli sparsely branched; branch width 0-1—2-0 mm, exceptions to 6 mm, branches solid, flat to subterete, narrow branches nodular, apices broad and blunt; surface shiny; holdfast delimited or diffuse, soralia marginal and laminal, round to ellipsoidal, becoming fissural on the upper surface occurring on eroded areas of the lower surface and at the apices. Apothecia rare, laminal near the branch apices, rarely terminal, disc 1-0 mm diam, concave, margin thick, entire; spores not produced. Chemistry: sekikaic acid (M), 4’-0-demethylsekikaic acid (t), + several terpenes and usnic acid. Remarks. Ramalina caespitella produces two morphotypes which intergrade into one another. One morph has narrow subterete to terete nodular branches with apical and marginal soralia, whilst the other morph produces very thick, flat to subterete branches bearing numerous soralia on both upper and lower surfaces. The extreme button-like morphology of both forms is indicative of their exposed, saxicolous habitat. No other taxon has been found to have a close affinity with R. caespitella. Distribution and habitat. This temperate maritime taxon is found on several of the islands in Bass Strait and along parts of the Tasmanian coastline (Fig. 9). It occurs on siliceous rocks (e.g. granite on the islands and on quartzite or dolerite at Tasmanian locations) and often occupies cliff-face habitats. World distribution. Australia. Bass Strait: Hogan Group: Hogans I, 39°13’S, 146°59’E, on granite, 1973, JW (MEL1013070); Kents Group, Dover I, 39°29’S, 147°17’E, on granite, 1970, JW (MEL1012702); Deal I, 39°30’S, 147°22’E, on granite, 1972, JW (MEL1012940); Furneaux Group: West Sister I, 39°45’S, 147°56’E, on granite, 1966, JW 156 G. NELL STEVENS t { 7 1 120 130 140 150 B aj 10 0 800km 8 20+ ye PS aed AQ ¢ 9 1) 30+ = GES a 407 uy \ L L Fig.9 Distribution of R. caespitella. (MEL1026132); Cape Barren I, 40°25’S, 148°10’E, on granite, 1969, JW (MEL1018083); Big Chalky I, 40°06’S, 147°54’E, on granite, 1972, JW (MEL1018053); Isabella Reef, 40°25’S, 148°30’E, on granite 1973, JW (MEL522220); Flinders I, 39°45’S, 147°56’E, on rock, 1965, RF (MEL1026188); Swan I, 40°41’S, 148°05S’E, on rock, 1974, JW (MEL1019119); Tasmania: West Point, 40°55’S, 145°15’E, on quartzite, 1981, GK (362/81); Rocky Cape, 40°51’S, 145°31’E, on quartzite, 1970, GB (H032637); Eddystone Pt., 41°00’S, 148°23’E, on granite, 1973, GB (H032640); Arthurs River, 41°07’S, 144°42’E, on quartzite, 1970, GB (H032639); Bruny I., 43°29’S, 147°09’E, on granite, 1973, GB (H032636). 3. Ramalina canariensis Steiner Plate 7, figs 3-4. in Ost. bot. Z. 9: 8 (1904). Type: Gran Canaria, Tafira, Bornmiiller, Pl. Canar. 3500 (W! — isotype; divaricatic acid and usnic acid). Thallus corticolous, pale green to grey green, erect to subpendulous, 1—3(—5) cm long; branching palmate or irregular; branch width 1-3(-10) mm, apices blunt, usually split with margins separated; surface matt, smooth to rugose, sometimes coarsely reticulately ridged by chondroid strands, pitted distally; holdfast delimited; soralia marginal and apical occurring through the separation of the upper and lower surfaces. Apothecia not seen. Reported as rare. Chemistry. Divaricatic acid, usnic acid. Remarks. Fine thalli of this taxon come close to resembling immature thalli of R. unilateralis RAMALINA IN AUSTRALIA 157 and both species contain divaricatic acid, but the cartilaginous surface and the finely lacerated apices of R. unilateralis enable the two to be separated. Distribution and habitat. The distribution pattern of R. canariensis in Australia is mainly maritime in areas experiencing a Mediterranean type of climate (Fig. 10). The only inland collections are two South Australian locations (Springton and Kuitpo Forest) at elevations of 300 m where frequent mists occur. Krog & @sthagen (1980) mentioned that in the Canary Islands (28°N) this species ‘seems to prefer well lit sites with a certain influence of mist. . . between 400 and 1000 m altitude’. Insome of the drier Australian coastal sites morning fogs are prevalent. Along the southern coastline of Australia this species grows in well-lit situations protected from winds, usually on twigs of Leucopogon parviflora, Casuarina, and Melaleuca spp. It has also been found on mangroves (Avicennia marina) along the eastern coastline south from latitude 35°S. Ramalina canariensis occurs in homogeneous colonies or immixed with R. fissa. Overseas specimens were examined from South Africa (Cape Province, latitude 33°S at 800 m elevation) and Chile (Valparaiso, latitude 33°S at sea level, on dune shrubs). Both of these regions have climates similar to southern Australia and other Mediterranean regions. Ramalina canariensis was previously thought to grow only in the northern hemisphere in the Mediterra- nean region and on the Canary Islands. World distribution. Mediterranean region, Canary Islands, South Africa, Australia, and South America. 120 130 140 160 0 800km sy Uy , m . 1 1 a) & f 20~ Qr 9 a f 20, S v e Cm 9 @ é 40N 1 T i ie Fig. 10 Distribution of R. canariensis. 158 G. NELL STEVENS New South Wales: Sussex Inlet, 35°10’S, 150°35’E, on Avicennia marina, 1977, RWR (BRIU1903NS); Buckenbowra R. 35°42’S, 150°06’E, on Avicennia marina, 1983, HS (CBG8306039); Batemans Bay, 35°45’S, 150°10’E on A vicennia marina, 1978, GNS (BRIU2171NS); Batehaven Beach, 35°45’S, 150°13’E, on Casuarina sp., 1975, JAE (JAE 1061); Burrenwarra Pt., 35°49’S, 150°14’E, on Casuaria sp., 1975, JAE (JAE 1246). Victoria: Cape Conran, 37°48’S, 148°44’E, on Leucopogon sp., 1983, GNS (BRIU3701NS); Westernport Bay, 38°13’S, 145°18’E, on Avicennia marina, 1978, GNS (BRIU2336NS); Gunnamatta Beach, 38°21'S, 144°45’E, on Monotoca elliptica, 1983, GNS (BRIU3692NS); Warrnambool, 38°23’S, 142°29’E, 1886, F. R. M. Wilson (NSW.L4311); Portland, 38°20’S, 141°36’E, 1896, F. R. M. Wilson (MEL 9419); Portsea, 38°25’S, 144°42’E, on tree on sea cliff, 1964, J. Williams (MEL1026126); Somers, 38°24’S, 145°10’E, 1964, Marie (MEL1026173). South Australia: Kuitpo forest, 35°07’S, 139°16’E, on Pinus sp., 1968, RWR (BRIU1426NS); Kangaroo I. Lockwood Corner, 35°39’S, 137°38’E, 1967, G. Jackson (S.A.97647604); Mt Lofty Ra., Carey Gully, 34°45’, 139°00’E, on Exocarpus cupressiformis, 1976, JAE (JAE 2853); Springton, 34°45’S, 139°05’E, on Callitris sp., 1976. JAE (JAE 2807); Eyre Peninsula: Memory Cove, 34°40’S, 135°50’E, on Melaleuca sp., 1970, RF (MEL1018638); Elliston, 32°59’S, 134°30'E, 1967, N. Donner (SA97648163). Western Australia: Gairdner River, 34°14'S, 119°17’E, on Hakea sp., 1980, D. Richardson (WA000943); Recherche Arch. Long I, 34°03’S, 121°57’E, on Melaleuca sp., 1950, JW (MEL10020). Bass Strait: Furneaux Group: Flinders I, 40°00’S, 148°00’E, on shrub, 1977, JW (MEL10225619); East Sister I, on dune shrub, 1972, JW (MEL1025619); West Sister I, 39°42’S, 147°56’E, on Pomaderris apetala, 1969, JW (MEL521577); Hogan Group: Long I, 39°12’S, 147°01'E, on Banksia integrifolia, 1973, JW (MEL1012966). 4. Ramalina exiguella Stirton Plate 7, fig. 5; Plate 13, fig. 3. in Trans. Proc. R. Soc. Vict. 17: 68 (1881). Type: Australia, Queensland, Brisbane, Bailey 91 (BRI! - holotype; BM — isotype; usnic acid). Ramalina pertenuis Magnusson in Ark. Bot. 32A (2): 7 (1945). Type: Oahu, Keeawa awa Valley, 1922, Skottsberg s.n. (GB! — part of holotype; usnic acid). Thallus corticolous, grey-green, caespitose, erect, rigid, up to 3 cm high; branching mainly from the base, squarrose branchlets sometimes present on main branches; branch width 0-3-1-0 mm, branches subterete, narrow, apices attenuate, sometimes black tipped; cortex shiny, longitudinally grooved giving a string-like appearance; pseudocyphellae linear; holdfast delimi- ted; soralia absent. Apothecia few to many, usually marginal along the branches and/or subterminal causing the branch apex to become geniculate with a long, attenuate spur; disc. 0-2—2-0 mm diam., concave, plane to convex; margin entire, thick on immature apothecia; spores ovoid to ellipsoid or gibbous and slightly curved, (12—)14-16 x (S5—)6—8 um (Plate 5, fig. 1). Chemistry. Usnic acid only. Remarks. The affinities of this taxon lie with R. gracilis (Pers.) Nyl.; Nylander (1870) recorded material from Brazil, Madagascar, and Australia as belonging to R. gracilis. The type of R. gracilis was not available from either L or PC, so any relationship between these two taxa was not able to be checked. Some material of R. gracilis held at BM is morphologically similar to R. exiguella but contains salazinic acid. Krog & Swinscow (1976) reported that R. attenuata (Pers. ) Tuck. from Domingo, resembled R. exiguella but contained psoromic acid. Landron (1972) stated that R. gracilis contained psoromic acid whereas R. attenuata contained salazinic acid and that the latter species is synonymous with R. rigida Ach. Howe (1913-14) placed R. rigida and R. gracilis as synonyms of R. attenuata. It could be that this group of taxa is a complex with several chemical races, but until the type material of all of these species can be located and examined, the relationship between these taxa cannot be resolved. Stirton stated in his type description that R. exiguella had affinities with R. melanothrix Laurer, but the latter species is quite distinct and the type material does not resemble Australian material of R. exiguella. Distribution and habitat. The distribution of this coastal species is continuous along the eastern shoreline of Australia from latitudes 23°S—35°S, with additional occurrences at 18°S and 21°S. Its occurrence on the coastal hills is rare, one tropical collection was made from Abbey Peak, 650 m RAMALINA IN AUSTRALIA 159 elevation at latitude 14°S and one subtropical collection from Mt Mothar 320 m elevation at latitude 26°S (Fig. 11). Ramalina exiguella grows on several species of mangrove (viz. Rhizophora stylosa, Ceriops tagal, Avicennia marina, and Aegiceras corniculatum) and on other coastal trees (Casuarina equisitifolia and Callitris columellaris) which grow close to the water. It also occupies very exposed situations in the dune communities facing the ocean, where it is often the only lichen present. This taxon shows a marked difference in size of thallus in different habitats. The thallus may be minute (1-5 cm high) and narrow (as the name implies) or extremely coarse and ropy and up to 3 cm high. South of 28°S the thalli tend to produce lateral branchlets along the branches. In sheltered habitats, R. exiguella grows in association with R. inflata subsp. perpusilla, R. pacifica, R. subfraxinea var. leiodea, and R. subfraxinea var. confirmata. World distribution. East Africa, Australia, and New Zealand. Queensland: Abbey Pk 14°18’S, 144°30’E, on shrubs, 1983, C. McCracken (BRIU4232NS); South Mission Beach, 17°55’S, 146°05’E, on Ficus sp., 1983, GNS (BRIU4159NS); Eimeo, 21°06’S, 149°10’E, on Cerios tagal, 1979, GNS (BRIU1860NS); Yeppoon, 23°08’S, 150°45’E, on Casuarina sp., 1980, RWR (BRIU2221RR); Keppel Sands, 23°21'S, 150°47'E, on Rhizophora stylosa, 1975, RWR (BRIU890RR); Rhodds Peninsula, 24°02’S, 151°40’E, on Rhizophora stylosa, 1975, GNS (BRIU1277NS); Turkey, 24°06’S, 151°37’E, on Ceriops tagal, 1975, GNS (BRIU1156NS); Round Hill Head, 24°10’S, 151°54’E, on Aegiceras corniculatum, 1975, RWR (BRIU1935RR); Burnett Heads, 24°45’S, 152°25’E, on Casuarina, 1976, GNS (BRIU4203NS); Fraser I, Wathumba Ck, 24°50’S, 153°13’E, on Ceriops tagal, 1975, GNS 120 130 140 160 KG, 104 cede L 1 1 800km G e Q 204 5 mae SS : Cite ores y NY 9 e 304 ae GS ee ‘ ey 40 \ ‘ ' ' Fig. 11 Distribution of R. exiguella. 160 G. NELL STEVENS (BRIU642NS); Urangan, 25°19’S, 152°55’E, on Rhizophora stylosa, 1975, GNS (BRIU631NS); Boreen Point, 26°18’S, 152°58’E, on Callitris columellaris, 1982, GNS (BRIU4181NS); Cooran, 26°17’S, 152°50’E, on Araucaria cunninghamii, 1985, NCS (BRIU4335NS); Noosa River mouth, 26°24’'S, 153°04’E, on Casuarina sp., 1980, GNS (BRIU3184NS); Maroochydore, Maroochy R, 26°42’S, 153°05'E, on Avicennia marina, 1976, GNS (BRIU1313NS); Currimundi Lake, 26°45'S, 153°08’E, on Casuarina sp.., 1974, RWR (BRIU3078RR); Bribie I, 27°02’S, 153°10’E, on Avicennia marina, 1983, GNS (BRIU3711NS); Moreton I, Koorigal, 27°20'S, 153°25'E, on Callitris columellaris, 1980, RWR (BRIU4179NS); Stradbroke I, Myora Springs, 27°28’S, 153°24’E, on Avicennia marina 1977, GNS (BRIU1585NS); Thorneside, Tingalpa Ck 27°28’S, 153°11'E, on Rhizophora stylosa, 1977, GNS (BRIU1742NS); Wellington Point, 27°28’S, 153°15'E, on Avicennia marina, 1975, RWR (BRIU5648RR); Ormiston, 27°30’S, 153°16’E, on Avicennia marina, 1975, RWR (BRIUS459RR); Coochie Mudlo I, 27°35'S, 153°20’E, on Callitris columellaris, 1975, GNS (BRIU4579RR); Point Talburpin, 27°35’S, 153°17’E, on Rhizophora stylosa, 1982, GNS (BRIU4189NS); Little Rocky Point, 27°38’S, 153°18’E, on Ceriops tagal, 1977, GNS (BRIU1437NS). New South Wales: Hastings Point, 28°26’S, 153°32’E, on Avicennia marina, 1977, GNS (BRIU1929NS); Brunswick Heads, 28°33’S, 153°33’E, on Callitris columel- laris, 1982, GNS (BRIU3316NS); Byron Bay, 28°40’S, 153°36’E, on Cupaniopsis sp., 1979, RWR (BRIU3179NS); Ballina, 28°54’S, 153°33’E, on Avicennia marina, 1977, GNS (BRIU1961NS); Yamba, 29°27'S, 153°20’E, on Avicennia marina, 1977, GNS (BRIU2125NS); Woolgoolga, 30°06’S, 153°12’E, on dune shrubs, 1982, GNS (BRIU4178NS); Sawtell, Boambee Ck, 30°22’S, 153°06’E, on Avicennia marina, 1977, GNS (BRIU2034NS); Macksville, 30°41’S, 153°00’E, on Avicennia marina, 1982, GNS (BRIU4177NS); Macleay River, 30°54’S, 153°03’E, on Casuarina sp., 1982, GNS (BRIU4175NS); Port Macquarie, 31°27’S, 152°54’E, on shrub, 1975, JAE (JAE 1260); Port Macquarie, 31°27'S, 152°54’E, on Avicennia marina, 1977, GNS (BRIU1983NS); Laurieton, 31°20'S, 152°55’E, on Avicennia marina, 1977, GNS (BRIU4339NS); Taree, Saltwater, 31°54’S, 152°34’E, on tree, 1977, JAE (JAE 4014); Booti Booti, 32°15’S, 152°31'E, on Casuarina sp., 1982, GNS (BRIU4332NS); Patonga, 33°30’S, 151°20’E, on Avicennia marina, 1978, GNS (BRIU2318NS); Towra Point, 34°03’S, 151°10’E, on Avicennia marina, 1977, C. Scarlett (BRIU641CS); Batehaven, 35°45’S, 150°10’E, on dead Casuarina sp., 1974, JAE (JAE 417). 5. Ramalina filicaulis N. Stevens, sp. nov. Plate 7, fig. 6. Thallus corticolus, virido-griseus, pendulus vel decumbens usque ad 20 cm longus, ramificatione pro parte majore dichotoma; rami lati ad 0-2—0-5 mm, tenues, filicaules, subterete vel teretes, apice attenuati et filamentosi, saepe uncinatus; pagina laevigata, fere striata; soralia nulla. Apothecia rara, lateralia, disco plano vel convexo, ad 0-5—1-0 mm in diametro; sporae ellipsoideae vel fusiformes, rectae vel raro curvatae, 12—16(—20) x 4-6 wm. Acidum usnicum tantum continens. Holotype: Australia, Queensland, Lamington Plateau, 9 km N of O’Reillys in rain-forest on Araucaria cunninghamii, alt. 920 m. 27 November 1983, G. N. Stevens 3730 (MEL 1048085!). Thallus corticolous, green-grey, pendulous or decumbent, up to 20 cm long; branching predominantly dichotomous, curving at the axils; branches width 0-2—0-5 mm, narrow, thread- like, subterete to terete, tapering, apices attenuate and filamentous, often hooked; surface smooth, with linear pseudocyphellae; holdfast small, delimited but branches often attached to the substrate along their length; soralia absent. Apothecia uncommon, lateral sessile, disc 0-5—1-0 mm diam., plane to convex; margin entire; spores ellipsoid to fusiform, straight or rarely curved, 12—16(—20) x 4-6 wm. Chemistry. Usnic acid only. Remarks. Ramalina filicaulis may prove to be a member of the R. usnea complex because, apart from a chemical difference, fine specimens of R. filicaulis (no medullary acids) come close to the morphology of the type material of R. anceps (norstictic, + salazinic acids). Landr6én (1972) recorded R. anceps as a species in the Caribbean Islands (10°N—21°N) that occurs in lower montane rain-forest, most abundant on trees along the edge of the forest and in areas of high rainfall. Rundel (1978a) recorded that the R. usnea race with no medullary acids occurs most abundantly in eastern South America (23°S—28°S), a distribution pattern closely resembling the Australian taxon, but the flattened branches of R. usnea do not occur in R. filicaulis. There are morphological similarities between the Canary Island species R. chondrina and the Australian RAMALINA IN AUSTRALIA 161 taxon; both are subtropical species which occur in the mountains. Krog & Msthagen (1980) recorded R. chondrina as sparse and scattered on trees in laurel forests between 550 and 900 m elevation at 27°30'’N to 28°30’N. However, the anatomy of these two species differs; R. chondrina has an even cylinder of supportive tissue, whereas R. filicaulis has an uneven, ribbed cylinder of supportive tissue. Distribution and habitat. This species is rare in occurrence, being known from only a few locations (Fig. 8). It occurs profusely at Jimna State Forest (26°40’S) and at Lamington Plateau (28°15’S) at elevations between 600-900 m. Two early collections may have come from lower elevations in the Rockhampton area (23°S), but little information was available on the herbarium specimens. At both Jimna and Lamington this taxon grows almost exclusively on the trunks and branches of Araucaria cunninghamii (hoop pine) in the rain-forest. The only other phorophyte is Citriobatis pauciflora, a shrub which is a remnant of the rain-forest. The climatic data recorded for the occurrence of the hoop pine is also relevant for R. filicaulis. Webb & Tracey (1967) record that regions supporting Araucaria cunninghamii have a mean annual temperature in the subtropics between 12°—23°C, with occasional frosts; the annual average rainfall varies from 800-1800 mm with fogs and dew augmenting precipitation in upland areas. Hoop pine occurs as part of the climax forest only on soils of relatively low fertility (e.g. Jimna, on metamorphics); on soils of high fertility it does not occur in the climax forest except as rare veteran trees (e.g. Lamington Plateau, on basalt). Because of the rarity of this Ramalina it is difficult to assess what environmental factors are limiting its occurrence elsewhere. World distribution. Australia. Queensland: Jimna, 26°40’S, 152°28’E, on Araucaria cunninghamii, 1980, RWR (BRIU2202RR); Jimna, Marumba View, 26°40’S, 152°28’'E, on Araucaria cunninghamii, 1980, GNS (BRIU3349NS); Lamington Plateau, O’Reillys, 28°15’S, 153°08’E, windfall, 1970, ED (O); O’Reillys, 28°15’S, 153°08’E, on Citrio- batus pauciflora, 1984, P. Merottsy (BRIU4344NS). 6. Ramalina fimbriata Krog & Swinscow Plate:7, fig: 7. in Norw. J. Bot. 21: 117 (1974). Type: Kenya, Central Province, Nyeri District, Krog & Swinscow 2K36/137 (O! — holotype; BM! — isotype; divaricatic acid and usnic acid). Thallus saxicolous, pale green to stramineous, small, caespitose, usually very compact and button-like, often minute; up to 1-0 cm high; branching dense and intricate, distally producing branchlets; branch width 0-3—1-0 mm, branches flat to subterete, variably inflated, some branches with slit-like perforations; surface shiny to matt, smooth; holdfast diffuse; soralia form from disintegration of the lower cortex occurring laminally and at the apices (Plate 1, fig. 5). Apothecia uncommon, but numerous on some thalli, lateral or terminal, disc 0-5—3-0 mm diam., concave, plane or convex; margin entire or incised at maturity; spores ellipsoid, straight or rarely curved, 8-13 x 4-5 um. Chemistry. Divaricatic acid (often in trace amounts). Remarks. Reduced forms of R. unilateralis come close to the morphology of R. fimbriata, but the former species is corticolous in Australia and shows no inflation of the branches. The northern hemisphere species R. pollinaria, when growing on rock, produces thalli closely resembling R. fimbriata, but the former produces evernic acid. It is suggested that convergence of morphology due to similarity of habitats is the reason for the resemblance. Distribution and habitat. Ramalina fimbriata has a distribution range from latitude 35°S—43°S over most of the Victorian highlands extending into eastern Tasmania (Fig. 12). In Australia this species is saxicolous only, although in East Africa it is both saxicolous and corticolous (Krog & Swinscow, 1974). Its occurrence on several types of rock (viz. granite, sandstone, conglomerate, trachyte, and dolerite) shows it is not substrate specific. In some localities R. fimbriata occupies sheltered overhangs and ledges, habitats which would have little moisture available for the thalli, apart 162 G. NELL STEVENS 120 130 140 160 10N dames ; 800km SY as i 204 Ee ae he ea QQ ‘ $ f 304 cy ae ® e e e pe 3 i t } 1 Fig. 12 Distribution of R. fimbriata. from vapour or water droplets from the overhang. Yet large monotypic populations are found in such habitats. Fertile specimens were found at Mt Arapiles under a rock overhang. The production of apothecia in this sorediate species is rare and the ascospores were produced only sparingly, so that mature spores were hard to find. Other habitats where this species occurs are exposed rock surfaces where the small thalli are further reduced in size, indicating the harsh environmental conditions prevailing (e.g. cold winter temperatures with frosts and/or snow and hot summer temperatures). Krog & Swinscow (1974) record R. fimbriata growing on rock in the alpine zone of East Africa at altitudes of 3200-4100 m. The compact button-like thalli commonly collected in Australia are smaller than the cortico- lous Kenyan material recorded by Krog & Swinscow (1974), but in all other characteristics they are identical. World distribution. Africa, Australia, and New Zealand. New South Wales: Budawang Ra. 35°22’S, 150°03’E, on’conglomerate, 1981, W. A. Weber (COLO 616); Australian Capital Territory: Tharwa, 11 km SW of, 35°31’S, 149°04’E, on granite, 1979, JAE (JAE 6195); Booroomba Rocks, 35°32’S, 149°00’E, on granite, 1979, HS (CBG7910804); Mt Clear, 35°43’S, 149°07’E, on granite, 1980, JAE (JAE 9013). Victoria: Mt Arapiles, 36°45’S, 141°50’E, on sandstone, 1981, RWR (BRIU2420RR); 1969, A. C. Beauglehole (MEL1013299); Melville Caves Park, 36°39’S, 143°42’E, on rock, 1964, RF (MEL1516515); Bogong High Plains, 36°44’S, 147°18’E, on rock, 1954, C. Skewes (MEL1025567); Mt Ararat, 37°20’S, 142°52’E, c. 1875, D. Sullivan (MEL 9471); Hanging Rock, 37°23’S, RAMALINA IN AUSTRALIA 163 144°37’E, on trachyte, 1976, GNS (BRIU3127NS); Yea, 37°13’S, 145°26’E, on granite, 1964, RF (MEL1026128); You Yangs, 38°00'S, 144°29’E, on granite, 1966, GB (H032638). Tasmania: Avoca, 41°45'S, 147°42’E, on granite, 1972, GB (H032552); Freycinet National Park, Mt Amos, 42°13’S, 148°18’E, on granite, 1979, JAE (JAE 5522); Woods Quoin, 42°16’S, 147°05’E, on dolerite, 1972, GB & JC (H032336); Lake Tooms Rd, 42°20’S, 147°28’E, on rock, 1974, GB & J C (H032551). 7. Ramalina glaucescens Krempelh. Plate 8, figs 1-12; Plate 13, fig. 4. in Verh. zool.-bot. Ges. Wien 30: 333 (1880). Type: Australia, Victoria, Mount Ararat, 1875, Sullivan s.n. (M! — holotype; sekikaic acid and usnic acid). Ramalina leiodea var. fastigiatula Mill. Arg. in Flora, Jena 66: 21 (1883). Type: Australia, Victoria, The Grampians, Sullivan 14 (G! — holotype; sekikaic acid (trace) and usnic acid). Plate 13, fig. 6. Ramalina lacerata Mill. Arg. in Flora, Jena 66: 20 (1883). Type: Australia, Western Australia, Eucla, Oliver s.n. (G! — holotype; M!, MEL!, UPS! - isotypes; no medullary acids usnic acid only). Plate 13, fig. 5. Ramalina brevis F. Wilson in Victorian Nat. 6: 69 (1889). Type: Australia, Victoria, Warrnambool, 1887, F. R. M. Wilsons.n. (NSW! — holotype (L4091); no medullary acids, usnic acid only). Plate 13, fig. 8. Ramalina calicaris var. australica Rasanen in Suomal. eldin-ja kasvit. Seur. van. Tiedon. Péytiak. 4: 178 (1949). Type: Australia, Victoria, Barwon Heads, 1894, F. R. M. Wilson s.n. (H! (herb. Rasanen) — holotype; no medullary acids, usnic acid only). Plate 13, fig. 7. Thallus corticolous, pale green, caespitose, erect to subpendulous, up to 4 cm long, excep- tions to 6 cm; branching sparse, subdichotomous to irregular, additional side branchlets occur on narrow branches; branch width (0-5—)1—5(—10) mm, branches compressed, narrow and canaliculate or broad and flat, apices usually forked, narrow branches with fine pointed apices broad branches with dissected apices; surface shiny and smooth or matt and rugose, becoming coarse with chondroid strands showing at the surface, often horny in texture; holdfast delimited or diffuse (if growing in colonies); soralia absent. Apothecia common, laminal at the axil of bifurcating branches (to 5 mm diam.) or laminal towards the branch apices (0-5—1-0 mm diam.), marginal and subterminal on narrow canalicu- late branches (0-5—1-5 mm diam.), laminal on broad branches (2-10 mm diam.) at or near the apex margin, with smaller apothecia on the same branch near the centre; disc concave becoming plane, rarely convex; margin thick to thin, persistent, often inrolled; spores ellipsoid, straight or curved, 10—12(—16) x 4—5(-6) um. Chemistry. Sekikaic acid (+ trace homosekikaic acid or trace divaricatic acid), or usnic acid only, or divaricatic acid and usnic acid. Remarks. The type material of R. glaucescens comprises a branch covered with tiny tufted specimens and a card with four specimens attached to it (which are here designated a, b, c, d). Two growth forms are represented on the card; a and d have bifurcate branches with small apothecia lying at the axil of the fork, and b and c have broader branches and bear terminal or subterminal apothecia, the latter with small subtending spurs. All of the thalli tested contain sekikaic acid. The specimens are exceptionally small and are immature. The considerable polymorphism encountered in R. glaucescens presented much difficulty in interpretation, as there was no correlation with the chemical variation; a situation rather similar to that encountered by Kristinsson (1969) in dealing with Cetraria islandica in Iceland. Five different species names already existed for entities now included in the complex: R. glaucescens Krempelh., R. leiodea var. fastigiatula Mill. Arg., R. brevis F. Wilson, R. lacerata Mill. Arg., and R. calicaris var. australica Rasanen. After the examination of large quantities of material belonging to this complex, both in the field and from herbarium collections, it became obvious that the four names additional to that of R. glaucescens (the earliest name) could not be maintained as species, because of the subtle intergradation between each of them (Plate 8, figs 1-15). An examination of spores from all of the morphotypes failed to show any marked differences in spore size or shape. Each apothecium produced spores with considerable variation in both "s1aadq *Y JO 9dAy 9Y} SO[QUIOSAI CT “SII ‘vynijn1sysvf ‘eA vapoiay ‘y Jo adh} oY} SaTquIsal 6 ‘BIJ ‘VIDAISND ‘IA S14vIIVI “yy JO 9dA} DY} SO[QUIOSOI g “SIL ‘DIDABIv] “Y JO adj ay} SaTquiasal ¢ ‘S14 ‘suaosaonn]8 “y JO 9dA} oY} So[QUIOSOI p “SI ‘WU UI oyeds OWes 0} [TY “UOneLeA [eo1Zojoydiow Zurmoys ‘Ave [PUOTIOIIP-1[NUI UI yNO Jos suadsaony{s ‘y Jo sodAyoydIow USAT -§ -3IeId UOVOUOUUOUOOLUQUOLUMTTVETTONIILIL PTTTTITOLIII UL OULU EUUGUT LL TOU Co ! el eb | Lt . %: } > ¥ A wg i7 oO Pe Lhe gM : 22) Z fi > fa = Nn — e fa Z ) RAMALINA IN AUSTRALIA 165 characters, so that the ratio of breadth to length was never constant for any particular morphotype. The ratio of straight to bent spores was also variable with a tendency towards a greater proportion of bent spores in the coastal thalli. The taxonomic interpretation of this complexity of morphotypes offered two alternatives: (1) to regard each existing ecotype as a distinct species which would allow for the five existing species and would produce at least two other new taxa, or (2) to regard the whole group as one extremely variable species showing remarkable phenotypic plasticity in response to environ- mental pressures. The latter would incorporate numerous morphotypes with similar spores and anatomy, which occupied different habitats, yet showed intergradation between their extreme morphological forms (Plate 8, figs 1-15). The second alternative was chosen because variation in chemistry also occurs throughout the group, which made it impossible to satisfactorily find any point for delimitation based on the acids present. Consideration was given to applying varietal rank to the different ecotypes, but each ecotype would then have had a sekikaic acid race and an acid-deficient race. The division of R. glaucescens into two taxa based on presence or absence of medullary acids could be feasible, especially as thalli containing sekikaic acid occur mainly east of longitude 147°, whereas west of this latitude the majority of the specimens contain no acids. If a division of taxa was made this way then a mixed group of morphotypes would be placed in each class. The amount of sekikaic acid present in R. glaucescens specimens is very variable in quantity, and division of a species on so fine a difference as presence or absence of traces of sekikaic acid appeared unwarranted. Usually acid-deficient races are not distinguished taxonomically from their acid-producing partners. The extreme variability of this taxon (as described above) indicates it could be a rapidly expanding species in Australia. As it also occurs in New Zealand, investigation of material here may provide information to clarify how it should be divided, whether on its chemistry or its morphology. Until the matter is finally resolved this complex of morphotypes is regarded as a single, markedly polymorphic species. Two specimens collected from Walcha Road (31°S) contain divaricatic acid; and both sekikaic and divaricatic acids occur in several thalli collected at Nimmitabel (36°S); this anomaly cannot be explained. Both areas are at altitudes of over 1000 m and such habitats may produce physiological changes in the metabolism of R. glaucescens. No taxonomic rank was considered for the atypical divaricatic acid thalli. Distribution and habitat. Ramalina glaucescens has a wide distribution in the warm to cool temperate regions of eastern Australia (Fig. 13), occurring at altitudes from sea level to over 1300 m. A large number of morphotypes are found as illustrated in Plate 8, figs 1-15. A general intergradation of forms can be seen which do not form a continuum but merge from several directions. Thalli with a morphology which is larger, but otherwise similar to that of the type of R. glaucescens, are found along the inland mountains west of the eastern escarpment in New South Wales and in the mountainous area of central Victoria, as well as in the central and eastern mountains of Tasmania (Plate 8, figs 2,3, 4, 5). All of the specimens examined contain sekikaic acid in varying amounts, except for a few from western Victoria and South Australia which are acid-deficient. Intergrading with this morphotype is a narrower, sometimes canaliculate morph which often has subsidiary branchlets extending from the main branches; it has both terminal and sub- terminal apothecia. The distribution of this ecotype is in the Eastern Highlands (with a disjunct appearance in the Grampians — type location of R. leiodea var. fastigiatula) and in the highlands of Tasmania (Plate 8, figs 5, 6, 7,9, 13, 14). In its narrowest form the branches become subterete and very fine, bearing small apothecia; the apices of the branches are sharply pointed (Plate 8, figs 11, 12); this spiky form occurs in mountain habitats above 1400 m elevation in the temperate region, e.g. it occurs at Mt Aggie and the Brindabella Ranges in New South Wales and at Mt Wombargo in Victoria, and at lower altitudes at Buchan Caves, Victoria, and in Tasmania. All collections were from twigs of Epacridaceae bushes, so it is not clear whether low temperatures 166 G. NELL STEVENS 1230 130 140 160 1 1 1 n Fig. 13 Distribution of R. glaucescens: @ sekikaic acid race; O acid-deficient race; * divaricatic acid race. or a particular bark substrate has an effect on this phenotype’s distribution. Sekikaic acid was always present in such thalli. One ecotype of R. glaucescens grows along the southern Victorian coastline; it has a short broad thallus with large terminal or laminal apothecia and is acid-deficient. A few specimens resembling this morphology have been collected from inland in eastern Australia, but that material contains sekikaic acid and is not so coarsely textured as the coastal thalli. This ecotype resembles the type of R. brevis (Plate 8, fig. 15).. Another acid-deficient ecotype occurs inland at low altitudes in Victoria and South Australia, but the thallus is generally narrower than that of the coastal ecotype and the apothecia are not as large; however, intergradation between the two is evident at some locations (Plate 8, fig. 8). The presence of small branchlets on the main branches resembles the eastern highlands ecotype. The type of R. calicaris var. australica belongs to this ecotype. In some habitats the thallus becomes quite coarse in texture with chondroid strands showing at the surface; such characteristics are apparent in the material from the arid zone in Western Australia, including the type collection of R. lacerata. The largest morphotype in the R. glaucescens complex is found in western Victoria and eastern South Australia, with a few specimens collected in western New South Wales. The latter specimens contain sekikaic acid but all other material proved to be acid-deficient. Most of the ecotypes of R. glaucescens have been found to intergrade with this large morphotype (Plate 8, fig. 1). It is normally sterile although pycnidia have been detected in some of these specimens. Four of the type specimens here included in R. glaucescens were initially collected within 100 RAMALINA IN AUSTRALIA 167 km radius of one another in Victoria, which illustrates the ecophenotypic response to different habitats. Table 9 summarises the morphological variation that is found in the different ecotypes. This taxon is entirely corticolous, the most common phorophytes being Acacia spp. and Bursaria spinosa. It occurs as dense colonies along the branches of isolated trees in pastureland or in open-forest communities. In some localities the entire population may be small and immature (as found in the type material of both R. glaucescens and R. leiodea var. fastigiatula), whereas in other areas only large mature specimens occur. World distribution. Australia and New Zealand. New South Wales: Ben Lomond, 30°00’S, 151°42'E, on Acacia sp., 1980, GNS (BRIU3249NS); Sherwood Ck, 30°03’S, 153°03’E, on lemon tree, 1978, HS (CBG7905095); Walcha, 10 km W of, 30°58’S, 151°30’E, on Acacia sp., 1982, GNS (BRIU4100NS); Abercrombie Caves, 33°32’S, 149°15’S, on Leptospermumsp.., 1978, JAE (JAE 4640); Rockley, 33°42’S, 149°34’E, E. Boorman (G); Black Springs 10 km N of, 33°49’S, 149°48’E, on Acacia sp., 1984, NCS (BRIU4292NS); Black Springs 5 km S of, 33°52’S, 149°48’E, on Acacia sp., 1984, NCS (BRIU4283NS); Abercrombie R. 34°12'S, 149°45’E, on Pinus sp., 1984, NCS (BRIU4284NS); Wombeyan Caves, 34°18’S, 150°02'E, E. Cheel (NSW L4294); Taralga, 37 km N of, 34°24'S, 149°49’E, on fence posts, 1977, JAE (JAE 3165); Crookwell, 34°28’S, 149°30’E, on Acacia sp., 1980, GNS (BRIU3489NS); Yass, 26 km S of, 35°00’S, 148°50’E, on Casuarina sp., 1975, HS (CBG058278); Little Peppercorn Ck, 35°35'S, 148°37’E, on Bursaria spinosa, 1979, HS (CBG7906928); Rules Point, 35°43’S, 148°31’E, on shrub, 1979, HS (CBG7907045); Lake Eucumbene, 36°10’S, 148°50’E, on Hymenanthera sp., 1971, L. Craven (MEL1020182); Adaminaby, 36°12’S, 148°50’E, on Briar, 1984, NCS (BRIU4287NS); Countegany, 36°12’S, 149°29’E, on Acacia melanoxylon, 1976, JAE (JAE 1925); Cooma 46 km S of, 36°14’S, 149°05’E, on Acacia sp., 1978, JAE (JAE 5440); Nimmitabel, 36°31’S, 149°15’E, on Prunus sp., 1980, GNS (BRIU3215NS); Holts Flat, 36°40’S, 149°15’E, on Acacia sp., 1980, GNS (BRIU3310NS); Delegate, 37°03’S, 148°57’E, on Acacia sp., 1980, GNS (BRIU3490NS). Australian Capital Territory: Ginninderra Ck, 35°05'S, 149°03’E, 1977, HS (CBG066912); Uriarra Crossing, 35°08’S, 148°57’E, on Casuarina sp., 1974, JAE (JAE 628); Blundells Ck Rd., 35°21'S, 148°50’E, on Acacia sp., 1977, D. Verdon (CBG8008538); Mt Aggie, 35°27'S, 148°46’E, on Hymenanthera dentata, 1979, JAE (JAE 5870). Victoria: Walwa, 9 km S of, 35°55'S, 147°45’E, on Acacia sp., 1984, NCS (BRIU4338NS); SW of Corryong, 36°16’S, 146°50’E, on Briar, 1984, NCS (BRIU4288NS); Myrtleford, 19 km S of, 36°39’S, 146°42'E, on Acacia sp., 1984, NCS (BRIU4290NS); Tawonga, 36°41’S, 147°08’E, on Prunus sp., 1984, NCS (BRIU4293NS); Bright, 36°44’S, 146°58’E, on copper beech, 1984, NCS (BRIU4286NS); Mt Wombergo, 36°36’S, 148°11'S, on Hymenanthera sp., 1966, RF. (MEL1026138); Bonang, 37°12’S, 148°43’E, on Malus, 1980, GNS (BRIU3491NS); Kyneton, 37°15'S, 144°28’E, 1897, F. R. M. Wilson, (MEL 9440); Newlyn, 37°25’S, 144°00’E, on Crataegus sp., 1981, RWR (BRIU2334RR); Buchan Caves, 37°30'S, 148°10’E, 1965, RF (MEL1026151); Omeo, 7 km N of, 37°04'S, 147°38’E, on Acacia sp., 1984, NCS (BRIU4285NS); Skipton, 11 km S of, 37°47’'S, 143°22’E, on Hymenanthera sp., 1984, NCS (BRIU4291NS); Buninyong, 37°42’S, 143°40’E, on trees, F. R. M. Wilson (MEL 9479); Steiglitz, Moorabool R dam, 37°53’S, 144°11’E, on twigs, 1968, GB (H032786); Yallourn, Storage dam, 38°10’S, 146°20'E, 1975, RF (MEL1013745); Koweerup, 12 km NE of, 38°10’S, 145°36’E, on Pinus sp., 1984, NCS (BRIU4336NS); Loy Yang, 38°11’S, 146°37’E, on Acacia sp., 1984, NCS (BRIU4337NS); Billywing Forest, 37°10’S, 142°10'E, Acacia baileyi, 1981, RWR (BRIU2429RR); Dartmoor, Crawford Lake, 37°56'S, 141°25'E, 1978, T. Muir (MEL1025115); Macarthur, 16 km W of, 37°58’S, 141°48’E, on Leptospermum sp., 1969, A. Orchard (SA97647210); Sale, 38°07'S, 147°04’E, on Acacia sp., 1980, GNS (BRIU3208NS); Tyrendarra, 38°15’S, 141°50’E, on tree, 1951, A. C. Beauglehole (MEL 9487); Lake Beeac, 38°12'S, 143°38’E, on Briar, 1984, NCS (BRIU4340NS); Barwon Hds, 38°20’S, 144°30’E, 1894, F. R. M. Wilson (H); Gunnamatta Beach, 38°20'S, 144°45’E, on Pimelea sp., 1983, GNS (BRIU3680NS); Strzelecki South, 38°20’S, 145°58’E, on dead wood, 1971, GB & JC (H032630); Westernport Bay, 38°22’S, 145°20’E, on Avicennia marina, 1978, GNS (BRIU4250NS); Warrnambool, 38°23’S, 142°31’E, on Bursaria spinosa, 1887, F. R. M. Wilson (NSW L4091); Pirron Yalook, 38°21'S, 143°26’E, on shrub, 1952, A. C. Beauglehole (MEL1023376); Yarram, 17 km NNW of, 38°24’S, 146°38’E, on Acacia sp., 1980, HS (CBG8002966); Korumburra, 38°26’S, 145°49’E, on Acacia sp., 1980, GNS (BRIU3301NS); Forrest, upper Barwon R, 38°32’'S, 143°45’E, on ‘prickly moses’ bush, 1952, A. C. Beauglehole (MEL1023362); Gellibrand R. falls, 38°40’S, 143°10’E, on tree, 1952, A. C. Beauglehole (MEL1023385); Skenes Ck, 38°43’S, 143°45’E, on Helicrysum dendroideum, 1983, M. Seaward 104633; Wilsons Promontary, 39°06’S, 146°12’E, on Acacia sp., 1980, GNS (BRIU3207NS). South Australia: Yalata Roadhouse, 31°39’S, 129°03’E, on bushes, 1980, AA (BRIU4341NS); Swan Reach, 24 km S of, 34°37’S, 139°30’E, 1971, K. Czornij (SA97414420); Springton, 6 km W of, 34°39’S, 139°06’E, on Hakea sp., 1975, JAE (JAE 884); — pois}}e} (uoIse1 —_ ouejd = [eUTUIO}qns soorde pu — ejong) wi 9-4 x 0} pue asIvod =—s_- 9} BTMIeued elyersny 9I-ZI €-[ eavou0s yeuTULIS} “AuTYs MOWeU = G-[-T €-Z [tu ISOM WS peains X9AU09 WS 1 ste.13s 0} = [BUTUIO}qns asieod = ayeTMIeued elyeijsny yNos wi (9-)S—-p x oueyd pue yjOoouls Anysiys pur PLIO}OIA 9I-ZI €-[ oavouo0d jeuTUIS} ‘yeu 0} 1ey y-Z p-Z [lu JSOM-YINOS %OS Pears PoytoJ BHOPIA WS Wste1s oueyd jeulsiew sooide [IU Moy SOM [B.1]U99 0} wi ¢-p x 0} pue yjoowls aieynoiyeues Ayureul (sopnynye ysry): , - 9I-ZI1(-OT) Z-[ oaevou0d yeUTWId}qQns ‘AUIYS MOWeU = ZS -0) €-T ore yIyos spuelysiy wsoyseq | iS %OE Peamns (seore Alp) as WOOL Wste1s oueyd pospu por9}}e} oe yIyos elyelsny yynos | D wi ¢-p x 0} esosni Udo soorde MO} YIM PUP PLIOJOIA ° 3 vI-ZI ¢-I 9AReou0d yeurwey] ‘yeu “yey 8-Z ¢ [Iu JSOM-YINOS Z %O8 peas SOTYULIM o WOT 1Ws1e.ys oueyd JeUIWID} + o}e[NONOI peolq wil (g-)S-p x 0} pue quia soorde (91-)rI-ZI 8-[ daeouo0d eure] as1e09 ‘yey 8-Z Z [tu BLIOJOIA [BISVOD %O0S Peains AWS Wste.ys oueryd ososn Poyloy [IU MoJ BLIOWIA wil 9-p x 0} as1B09 soorde ‘AyUTe Ur SOM [P.1}U99 0} 91-Z1(-0T) 8-S:0 sARDU0D [eure] ‘yeu ‘ey aes § €-T = oe yHTes spuelysiy us1o\seq soiodg eloyjode odeys Blooyjode X9}109 soorde wu wid juasoid vole jo ‘weiq osIg jo jo 2 odeys Ypia =- ys] poy jeorydeis0eyH uoIsOg 91N}X9 |, yourig yourlg yourlg *sadA}099 94} JO SIo}ORILYO [eoIWIaY pure [eoso;oydiour oy} UI UONeLIeA BUIMOYS sUadsaony]s Duy d1Y4dIOUIATOd dy} Jo UOHeSoISes s1ydeIsOIH 6 AIGUL 168 RAMALINA IN AUSTRALIA 169 Curramulka 5 km W of, 34°43’'S, 137°45’E, on Casuarina sp., 1977, JAE (JAE 3731); Kuitpo Forest, 35°14’S, 138°36’E, windfall, 1968, RWR (BRIU3032RR); Myponga, 35°24’S, 138°28’E, 1968, V. Cruick- shank (SA97414419); Clarendon, 35°06'S, 138°40’E, 1968, V. Cruickshank (SA97418422); Strathalbyn, 35°20’S, 138°50’E, 1968, V. Cruickshank (SA97414424); Salt Ck, 35°28’S, 139°09’E, on dead gum, 1967, GB (H032783); Deep Ck, 35°38’S, 138°12'E, on Casuarina sp., 1979, RWR (BRIU1863RR); Messent N.P., 35°28’S, 139°10’E, 1969, J. Carrick (SA97416171); Hindmarsh Falls, 35°29’S, 138°38’E, 1968, V. Cruickshank (SA97416168); Kangaroo I. Lockwood Cnr. 35°39'S, 137°38’E, on tree, 1967, G. Jackson (SA97647604); The Gap, 36°41’S, 140°40’E, 1973, M. Beck (H032671); Kingston, 60 km N of, 36°50’S, 139°51'E, P. Copley (SA97648420); Milbrook Reservoir, 34°50'S, 138°47’E, on Hakea sp., 1969, RWR (BRIU1780RR); Penola, 37°23’S, 140°49’E, on twigs, 1971, RWR (BRIU2035RR). Western Australia: Eucla, 31°43’S, 128°55’E, J. Oliver (G, M). Tasmania: Trowutta, 41°02’S, 145°05’E, on Malus sp., 1981, GK 384/81; Beaconfield, 41°12’S, 146°45’E, on Acacia dealbata, 1980, GK 214/80; Prossers Forest Rd, 41°15'S, 147°18’E, on Malus sp., 1966, GB (H032777); Avoca, Lochaver, 41°41’S, 147°50'E, on Prunus sp., 1978, GK 1/78; Cressy, 41°43’S, 147°02'E, on dead Acacia, 1964, JC (H032576); Lake Augusta, 41°50’S, 146°34’E, on twigs, 1973, GB (H032591); Great Lake Plateau, 41°57’S, 146°40’E, on bush, 1969, GB (H032628); Macquarie R, 42°08’S, 147°33’E, on small bushes, 1966, GB & JC (H032780); Lake Tooms Rd, 42°20’S, 147°28’E, on small bushes, 1974, GB & Gilbert, (H032791); Bothwell, 42°23’S, 147°00’E, 1968, GB & JC (H032593); Spring Hill, 42°24’S, 147°15’E, 1963, GB (H032770); Clyde Valley, 42°27’S, 146°52’E, on Acacia sp., 1975, K. Davies (H032629); Ouse, 42°28’S, 146°44’E, on Malus sp., 1980, GNS (BRIU3340NS); Bushy Park, 42°43’S, 146°53’E, on Exocarpus cupressiformis, 1980, GK 31/80. 8. Ramalina litorea N. Stevens Plate 7, fig. 8. in Lichenologist 18: 185 (1986). Type: Australia, Queensland, Mackay, Blacks Beach, 1979, Stevens 3129NS (MEL! — holotype; evernic acid, lecanoric acid and usnic acid). Thallus saxicolous, pale yellow-green, small, caespitose, rigid, 0-5—2-0 cm high; branching subdichotomous to irregular, small branchlets or spinules arise laterally from the main branches; branch width 0-2-1-0 mm, subterete to terete or rarely flattened, narrow, apices pointed, fragile; surface shiny to matt, without pseudocyphellae, smooth; holdfast diffuse; soralia absent. Apothecia common, marginal or lateral below the apices, disc 1-0—2-0 mm diam., plane; margin thin to very thick; spores small, oval or ellipsoid, straight, 8-12 x 4—4-5 wm. Chemistry. Evernic acid (M), lecanoric acid (t), + sekikaic acid (t), and usnic acid. Remarks. Some specimens of R. litorea resemble the Hawaiian species R. microspora Krem- pelh. However, the latter species contains divaricatic acid, whereas R. litorea contains evernic acid. Herbarium material labelled R. microspora from Mauritius and Rodrigues (in the Indian Ocean, 20°S) containing evernic acid and morphologically somewhat similar to the Australian taxon have been treated as R. litorea by the author. Further investigation may prove that R. microspora comprises two chemical races disjunctly distributed in the Pacific and Indian Oceans. Distribution and habitat. The full distribution range of R. litorea (Fig. 14) is unknown as only five collections have been made in an area from latitudes 14°40’—24°20’S. This saxicolous species has been collected from coastal cliffs, composed of a variety of rock types (granite, siltstone, slates, rhyolitic, breccia, and tuff) indicating it is not substrate specific. Only small colonies grow on the cliff faces, usually at 10-20 m above beach level; although on Lizard Island (14°S, the most northern collection) a specimen was collected at 360 m elevation. World distribution. Mauritius, Rodrigues Island, and Australia. Queensland: Lizard I. 14°40’S, 145°28’E, on granite (360 m elev.), 1974, R. L. Specht (BRIU3780NS); Cape Hillsborough, 20°56’S, 149°03’E, on rhyolitic breccia (10 m elev.), 1983, GNS (BRIU4124NS); Mackay, Blacks Beach, 21°04’S, 149°12’E, on siltstone (10 m elev.), 1979, GNS (BRIU3129NS); Great Keppel I, 23°10’S, 150°58’E, substrate unknown, 1976, U. Allen (H032240); Double Head, 23°08’S, 150°47’E, on slates (15 melev.), 1984, NCS (BRIU4307NS); Emu Park, 23°15’S, 150°30’E, on slates (10 m elev.), 1984, NCS (BRIU4308NS); Agnes Water, 24°20’S, 151°58’E, on rhyolitic agglomerate (5 m elev.) 1984, NCS (BRIU4277NS). 170 G. NELL STEVENS 120 130 140 150 fF SS 50 o- i 1 800km a ® f 20+ pee te Sara <2 aria? ay ae eae i ae SNC ees a ee oe ities wee ND q 8 - 304 (oS) 2 bs ce 40N 1 " rt Fig. 14 Distribution of R. litorea. 9. Ramalina peruviana Ach. Plate 9, figs 1-2. Lichenogr. Univ.: 599 (1810). Type: South America, Peru, Lagastas.n. (H-ACH! — holotype; BM!, UPS! — isotypes; sekikaic acid aggregate, and usnic acid). Ramaliria farinacea var. dendroides Mill. Arg. in Flora, Jena 66: 21 (1883). Type: Australia, New South Wales, Ash Island (Hunter River), 1883, Ford s.n. (G! — holotype; sekikaic acid aggregate and usnic acid). Plate 14, fig. 2. Ramalina farinacea var. squarrosa Mill. Arg. in Flora, Jena 66: 21 (1883). Type: Australia, Queensland, Rockhampton, 1883, Thozets.n. (G! — holotype; sekikaic acid aggregate and usnic acid). Plate 14, fig. 1. Thallus corticolous, pale green to grey-green, tufty, resilient, erect to subpendulous, up to 6-0 cm long; branching subdichotomous to irregular, often intricate with the production of dense fragile branchlets; branch width up to 1-0 mm, branches flat, compressed becoming angularly subterete to terete distally, branches often slightly twisted, apices sharp to blunt, often broken; surface matt, rarely shiny, smooth to rugose, weakly pseudocyphellate either basally or along the entire length; holdfast delimited or diffuse; soralia punctiform, numerous, marginal or lateral, sometimes apical (Plate 1, fig. 2), mounds of soredia often produce small fibrils (Plate 3, fig. 7). Plate 9 Thallus growth forms in the Australian Ramalinae. Scale in mm. Fig. 1 R. peruviana (dense form). Fig. 2 R. peruviana (open form). Fig. 3 R. reducta (lower surface). Fig. 4 R. tenella. Fig. 5 R. unilateralis (palmate form). Fig. 6 R. unilateralis (large, intricate form). Fig. 7 R. whinrayi. Fig. 8 R. celastri subsp. celastri (narrow form). > RAMALINA IN AUSTRALIA 3 & : Tn Ko tevtsderertitiileti ted FERESRAREASERESELE . LZ G. NELL STEVENS Apothecia rare, marginal, lateral, disc 2-0 mm diam., concave, margin thick, often crenate; spores narrow, fusiform, straight or rarely curved, 14—16(—18) x 3-5—4-5 wm. Chemistry. Homosekikaic acid (M), sekikaic acid (M), ramalinolic acid (m/t), 4’-0- demethylsekikaic acid (m/t), 4’-O-methylnorsekikaic acid (m/t), and usnic acid. Remarks. The holotype material from H and the isotypes material held at BM and UPS are all small pieces and the morphological characters are not well defined, faint pseudocyphellae occur on the basal portion of the branches, and the soralia are not very clear. The Australian material does not closely resemble the morphology of the type material but Krog & Swinscow (1976) commented that some of the East African specimens appeared to differ from the type specimen. When comparing East African material with Australian no differences could be found. Ramalina peruviana material examined from Uruguay and Brazil closely resembled Australian specimens and also displayed considerable variation in morphology. Two previously accepted infra-specific taxa from Australia are reduced to synonymy with R. peruviana; these are R. farinacea var. dendroides and R. farinacea var. squarrosa, both of which are regarded as ecophenotypes of R. peruviana. Ramalina tenella is the only species in Australia which may be mistaken for R. peruviana, but the acid difference sets them apart, the former containing salazinic acid. The Japanese taxon R. intermediella Vainio closely resembles the morphology of R. peruviana and contains similar acids, but it does not occur in Australia. Distribution and habitat. Ramalina peruviana is mainly subtropical to warm temperate in distribution (Fig. 15). North of the Tropic of Capricorn it occurs inland at 1000 m elevation at both Atherton Tableland (17°S) and Undara Crater (19°S). Such elevations in the tropics correspond with the reported distribution of the species in the West Indies (17°N) at altitudes between 600-1000 m (Landr6n, 1972), and in East Africa (10°N—10°S) between 1000-1800 m (Krog & Swinscow, 1976) allowing for latitudinal adjustment. Four ecotypes can be distinguished in the Australian taxa although intergradation occurs between each form: (i) the coastal form (which includes R. farinacea var. squarrosa type material), (ii) a montane form, (iii) a densely branched form found in open-forest habitats in Queensland and (iv) an open sparsely branched form in the open-forest habitats of south-east Queensland and New South Wales (which includes R. farinacea var. dendroides type material). The differences are shown in Table 10. Specimens of these four ecotypes were chemically analysed to assess if there was any correlation between their chemistry and the morphological differences. None was found, but variation in the quantity of some of the minor metabolites was obvious. This variation could be correlated with inland and coastal locations, i.e. coastal specimens had only trace amounts of 4’-0-methylnorsekikaic acid, but specimens from inland sites contained noticeably more. Jgrgensen (1977) commented that R. peruviana may need to be subdivided because of the number of morphotypes present, but from the survey of Australian material it is felt that any subdivision is unwarranted because of the intergradation between ecotypes. All collections in Australia have been corticolous. The main phorophytes for the coastal ecotype are the mangroves Ceriops tagal and Rhizophora stylosa. Once these species reach their distribution limit at 28°S, R. peruviana is found on various coastal trees. In the mountains this taxon occurs on several rain-forest tree species and in the open-forest habitat, shrubs such as Lantana are the host. This taxon is not substrate specific; in South America it has been found on fence posts (Osorio, in litt.). World distribution. Africa, Australia, New Zealand, several Pacific Islands, South America, and Tristan da Cunha. Queensland: Atherton, 5 km N of, 17°14’S, 145°29’E, 1970, E. Dahl (0); Atherton, 2 km SE of, 17°16’'S, 145°30’E, on Casuarina sp., 1983, HS (CBG8302451); Undara Crater, 18°23’S, 144°43’E, on rain-forest tree, 1977, NCS (BRIU3546NS); Forty Mile Scrub, 18°03’S, 144°53’E, on rain-forest tree, 1983, GNS (BRIU4115NS); St Lawrence, 22°20’S, 149°32’E, on Ceriops tagal, 1978, NCS (BRIU2678NS); Glad- stone, South Trees Inlet, 23°52’S, 151°19’E on Ceriops tagal, 1980, GNS (BRIU3592NS); Tannum Sands, RAMALINA IN AUSTRALIA 173 120 130 140 150 30 40 L | \ i Fig. 15 Distribution of R. peruviana. 23°58’S, 151°23’E, on Rhizophora stylosa, 1975, RWR (BRIU876RR); Turkey, 24°06’S, 151°38’E, on Ceriops tagal, 1975, GNS (BRIU1154NS): Jimna, 26°40'S, 152°28’E, on Araucaria cunninghamii, 1980, RWR (BRIU2197RR); Maleny, 26°46’S, 152°51’E, on dead tree, 1982, GNS (BRIU3708NS); Bunya Mts, 26°50’S, 151°40’E, on shrubs, 1979, GNS (BRIU2971NS); Mt Mee, 27°05’S, 152°42’E, on Flindersia sp.., 1979, GNS (BRIU3324NS); Mt Byron, W of 27°06’S, 152°37'E, on Araucaria cunninghamii 1982, GNS (BRIU3624NS); Brisbane, Serpentine Ck, 27°22’S, 153°08’E, on Ceriops tagal, 1975, GNS (BRIU1097NS); Stradbroke I. Amity Pt, 27°24’S, 153°25’E, on Lumnitzera racemosa, 1977, GNS (BRIU2050NS); Brookfield, 27°30’S, 152°55’E, on Citriobatus pauciflora, 1983, GNS (BRIU3694NS); St Lucia, 27°30’S, 153°00’E, on Bauhinia alba, 1976, GNS (BRIU3322NS); Worlds End Pocket, 27°31'S, 152°45’E, on Mallotus celaoxyloides, 1983, P. Forster (BRIU3705NS); Pt Talburpin, 27°39'S, 153°18’E, on Avicennia marina, 1980, GNS (BRIU3548NS); Little Rocky Pt, 27°43’S, 153°21'E, on Ceriops tagal, 1977, GNS (BRIU480NS); Stradbroke I. south point, 27°44’S, 153°25’E, on mangrove, 1975, H. T. Clifford (BRIU3074RR); Cunninghams Gap, 28°03’S, 152°24’E, on rain-forest tree, 1975, NCS (BRIU3695NS); Binna Burra, 28°12’S, 153°11’E, on rain-forest tree, 1976, GNS (BRIU3693NS); O’Reillys, 10 km N of, 28°13'S, 153°06'E, on Gympie tree, 1983, GNS (BRIU3724NS). New South Wales: Wilsons Ck, 28°34’S, 153°27’E, on rain-forest tree, 1980, GNS (BRIU3320NS); Goonengarry, 28°37’S, 153°28’E, on Citriobatus pauciflora, 1980, GNS (BRIU3319NS); The Channons, 28°39'S, 153°16’E, on Lantana, 1982, GNS (BRIU3526NS); Lismore, 28°48’S, 153°16’E, on Acacia sp., 1976, R. Moodie (BRIU2855NS); Nambucca, 12 km SSW. of, 30°41’S, 153°00’E, on Avicennia marina, 1982, GNS (BRIU3670NS); Taree, Saltwater, 31°54'S, 152°34’E, on tree, 1977, JAE (JAE 3999); Hunter R. Ash I. 32°56’S, 151°46’E, 1883, E. Forde (G); Broulee, 35°52’S, 150°09’E, on shrubs, 1970, JAE (JAE 128); Wagonga Inlet, 36°12’S, 150°08’E, 1908, G. H. Halligan NSW L4304). G. NELL STEVENS 174 soyis uodo uns ][NJ sous Apeys *$901} po}eIOSI Ul SOAOISURU JSo1OJUTeI uns [[NJ IO JSoIOJUTLI [eqseoo ouejuoUul UI }Sso10} Uddo juoueUeL ‘AIP eVIQey doeyIns JO sTuqy sonpoid spunowl SUTWIO} [OA9] BAoge doRjAINs 94} UO yor spunow ‘gsuop poste jou UOISNI}XO [[VWIS SUTWIO} ‘asuop BIPoIOS O¥l]-R20eULIRI, oB1R] ode] [jews Apystys 0} WInIpoul ‘wuo0jyyound wuojnound JO waoynound ‘wioyound ye101e] ‘yeuIsIeUl [[euls ‘[e193e] jeulsiew pue [eulsiew BI[e1OS yoursq jo y3u9] o1WUS uOIsal [eseq yOUTSIpUr SuoyTe JOUNSIPUI UI JOUN)SIP Io Juasqe o}~UNSIP oseijoydAdopnesg JJO UdYOIq JjO udyoIq JjO udyo1q jJJO UdyOIG IO poyooy owos ‘9}eNU9}}2 oulOs ‘d}eNUD}}e Io poyooy sooide Jo odeys soyourlg wu ¢-0-7:0 wul Q-T—-S-0 wu ¢-0-7:0 JO UIPIM a}010}qQNS 0} 91019} 93019}qNS 0} 91019}qNs 0} soyoursq yey ‘possorduios 0} 9}019}qns yey ‘possoiduios yey ‘possoiduros jo odeys Ie[NSOLII 0} snowoj}OYsIpqns snowojoYysIpqns 9} e9LI]UI AT[eISIP suryoursq Ie[Nso ‘osuop ‘uodo ‘uodo Ie[Nso1I ‘osuop jo odky, q edAjoydiop J edAyoydiopy gq odAjoydiopy Vv edAjoydiopy JayoRleyy ‘pupianiad pulpy JO S9dhj099 INO} UI puNOJ sIajoVIEYS [edIdojoydiow jo uostedwoD OT F142 RAMALINA IN AUSTRALIA 175 10. Ramalina reducta Krog & Swinscow Plate 9, fig. 3. in Norw. J. Bot. 23: 172 (1976). Type: Ethiopia, Bale Province, Dinshu, Tapper 933 (BM! - holotype; O! — isotype; psoromic acid and usnic acid). Thallus corticolous, grey-green or pale green, erect, flaccid, up to 4 cm long; branching sparse, initially a palmate lobe producing several multi-divided lacinae; branch width 10 mm at base, separate lacinae to 5 mm, branches compressed, flat, apices tattered or capillaceous; surface matt, rugose, often cracking; a continuous lower cortex lacking with the medulla overlaid by cartilaginous strands forming a reticular pattern, some strands becoming detached at one end; holdfast delimited; soredia pale green, produced from the exposed medulla. Apothecia rare, marginal, disc to 2-5 mm diam, concave to plane, margin entire; spores few, ellipsoid, curved, 13-18 x 4—5(-—6) um. Chemistry. Psoromic acid and usnic acid. Remarks. The Australian specimens are much larger than those examined from East and South Africa and cannot be regarded as ‘inconspicuous’ as described by Krog & Swinscow (1976). This species is easily recognised by the anastomosing strands of cartilaginous hyphae on the exposed lower surface. The production of psoromic acid distinguishes it from any other Ramalinae in Australia. Distribution and habitat. This taxon has a restricted distribution range (Fig. 16). It occurs in the open-forest uplands of the New England Plateau, a granitic/basaltic tableland rising to over 1 T 120 130 140 150 1 n i i Fig. 16 Distribution of R. reducta. 176 G. NELL STEVENS 1300 m in the subtropics. The narrow ranges of latitude and longitude in R. reducta distribution coincide with a 12°C mean maximum annual temperature isotherm. The distribution area experiences wet summers and drier winters, indicating that the summer may be the growing period for this species. This particular set of environmental factors is not found in other areas. To the north only individual peaks reach an elevation of over 1300 m and these are covered in rain-forest; and to the south, where the Australian Alps attain a height of over 2000 m, the rainfall is either uniform throughout the year or occurs during the winter. Seasonality may be a factor which inhibits the occurrence of R. reducta further south. Krog & Swinscow (1976) recorded East African material from the upper montane forests between 2600 m and 3100 m altitude, which when adjusted for latitude is comparable with the occurrence of R. reducta in Australia. Ramalina reducta was collected from Acacia, Casuarina, and Banksia spp. growing in well lit open-forest communities. A large monospecific community of this taxon was found at Ebor Falls, New South Wales, growing on all three phorophytes, in an area exposed to spray from the falls under certain wind conditions. In this environment thalli reached 4 cm in length and many produced apothecia, indicating that such a habitat promoted optimal growth. No fertile thalli were reported from the East African material. World distribution. Africa and Australia. New South Wales: Glen Innes, 12 km E of, 29°42’S, 151°50’E, on Acacia parvipinula, 1983, GNS (BRIU3782NS); Ben Lomond, 30°00'S, 151°42’E, on dead Acacia sp., 1980, GNS (BRIU3237NS); 1976, JAE (JAE 2435); Coutts Water, Ebor-Dorrigo Rd, 30°22'S, 152°30’E, on shrubs, 1982, GNS (BRIU3633NS); Ebor Falls, 30°25’S, 152°20’E, on Banksia sp., 1982, GNS (BRIU3632NS); on Acacia sp., 1982, GNS (BRIU3698NS); on Casuarina sp., 1982, GNS (BRIU3699NS); Walcha Road, 30°59’S, 151°30’E, on Acacia sp., 1982, GNS (BRIU3700NS); Topdale, 25 km E of Nundle, 31°30’S, 151°25’E, on Acacia sp., 1984, NCS (BRIU4297NS); Glenbawn Dam, 32°05’S, 151°00’E, on Acacia sp., 1981, AA (BRIU3611NS). 11. Ramalina tenella Mill. Arg. Plate 9, fig. 4 in Flora, Jena 62: 162 (1879). Type: Brazil, Sao Paulo, Apiahy, 1877, Puiggari 152 (G! — holotype; salazinic acid, protocetraric acid (trace), three unknowns, usnic acid). Thallus corticolous, pale green to yellow-green, erect, rigid, delicate, up to 3-0 cm high; branching subdichotomous with short lateral branches sometimes producing numerous distal branchlets; branch width to 1-0 mm, flat basally thereafter subterete; surface matt; short, linear pseudocyphellae; holdfast delimited; soralia apical and lateral on main branches and apical on secondary branches. Apothecia unknown. Chemistry. Salazinic acid + protocetraric acid, three unknowns, and usnic acid. Remarks. This taxon has a distinctive growth form when the branching is sparse; when densely branched it resembles R. peruviana. The presence of salazinic acid sets it apart from the sekikaic acid taxon R. peruviana. Krog & Swinscow (1976) stated that R. tenella resembled R. dendriscoides Nyl., a species which does not occur in Australia, although some early collections of R. peruviana were given the name R. dendriscoides (Appendix 1). Distribution and habitat. Ramalina tenella occurs in isolated, high rainfall pockets along the Queensland coast between 16°S and 24°S (Fig. 17). Uncommon in occurrence, it grows profusely in the areas it inhabits. Material collected from the more northern latitudes always contained three unknown compounds which did not occur in the material from latitudes 21°S—24°S. Lack of bulk material of this species prevented the identification of the unknowns by chemical analysis. Specimens lacking the unknowns were usually found to produce lateral soralia in addition to apical soralia, in contrast to the more northern collections which rarely produced lateral soralia. Ramalina tenella grows on the twigs of Ceriops tagal and Rhizophora stylosa in sheltered but well-lit habitats within the mangrove community. It is found in association with R. subfraxinea RAMALINA IN AUSTRALIA 77 ¥ 7 ui m ip 120 130 140 150 a 104 r) B00km SZ | = i J e L 204 » 9 g : 30 og” Cm p 2 o : Ly 404 i 1 L 4 Fig. 17 Distribution of R. tenella. var. confirmata, R. pacifica, R. exiguella, R. nervulosa var. nervulosa, and R. nervulosa var. luciae. World distribution. Africa, Australia, New Hebrides, and South America. Queensland: Daintree R, 16°15’S, 145°21'E, on Ceriops tagal, 1975, D. Tarte (BRIU2253NS); Mossman, Saltwater Ck, 16°25’S, 145°25’E, on Ceriops tagal, 1983, C. McCracken (BRIU4300NS); Mossman, Rifle Club, 16°28’S, 145°25’E, on Ceriops tagal, 1983, GNS (BRIU4127NS); Cairns, 16°45’S, 145°45’E, 1974, D. Tarte (BRIU2470NS); Murray R, 18°05’, 146°01’E, on Ceriops tagal, 1978, D. Tarte (BRIU3122NS); Hinchinbrook I, 18°20’S, 146°15’E, on Ceriops tagal, 1979, GNS (BRIU3134NS); Eimeo, 21°09’S, 149°10’E, on Ceriops tagal, 1977, GNS (BRIU2791NS); on Rhizophora stylosa, 1976, GNS (BRIU2469NS); Corio Bay, 23°00'S, 150°45’E, on Ceriops tagal, 1975, D. Tarte ‘BRIU1172NS); Gladstone, Wiggins Ck, 23°50’S, 151°15'E, on Ceriops tagal, 1980, GNS (BRIU3549NS); Turkey, 24°06’S, 151°38’E, on Rhizophora stylosa, 1975, GNS (BRIU1282NS). 12. Ramalina unilateralis F. Wilson Plate 9, figs 5—6. in Victorian Nat. 6: 69 (1889). Type: Australia, Victoria, Warrnambool, F. R. M. Wilson 432 (G! - lectotype herein designated; divaricatic acid and usnic acid). Thallus corticolous, pale green, caespitose to subpendulous, up to 4 cm long, exceptionally to 7 cm; branching irregular, dense; branch width 1-3 mm, flat to subterete, sometimes palmate at the base to 8 mm broad, distally producing numerous fine lacinae; surface cartilaginous, smooth and shiny, rarely matt; holdfast delimited; soralia occur in eroded patches at intervals along the 178 G. NELL STEVENS lower surface where the branches have split and flattened out (Plate 1, fig. 6) and also at the apices. Apothecia rare, disc 2-0 mm diam., concave to plane; margin entire; spores ellipsoidal, straight or rarely curved, 10-12 « 4-4-5 wm. Chemistry. Divaricatic acid and usnic acid. Remarks. The specimens held at G are regarded as the lectotype of R. unilateralis as no other Wilson material appears to exist. It is thought that the type may have been lost with numerous other Wilson types in 1907 (Filson, 1976). The morphology of some R. pollinaria specimens resembles the morphology of this taxon, and it was this name that was given to R. unilateralis specimens by early lichenologists. However, R. pollinaria contains evernic acid, whereas R. unilateralis contains divaricatic acid. Some specimens of R. canariensis resemble R. unilateralis; however, the texture of the surface is different. Reduced growth forms of the latter taxon may be mistaken for R. fimbriata, but this species is saxicolous in Australia whereas R. unilateralis is corticolous and generally much larger in size. All three species contain divaricatic acid. Distribution and habitat. Collections of this cool temperate species in Australia have been made from sea-level to elevations of up to 1150 m (Fig. 18). Although corticolous in Australia R. unilateralis becomes saxicolous at colder latitudes, e.g. Macquarie I. (55°S) and Tierra del Fuego (54°S). It appears to tolerate a broad range of environments. The type material was collected at Fig. 18 Distribution of R. unilateralis. RAMALINA IN AUSTRALIA 179 sea-level, and other coastal sites are Lakes Entrance, Westernport Bay, and Portland; inland collections above 1000 m were at Nimmitabel, New South Wales, and Old Man’s Head, Tasmania. Material cited from South Africa was collected at elevations of 2000 m at latitude 34°S, and from Chile, latitude 53°30’S at an elevation of 800 m. World distribution. South Africa, Australia, and South America. New South Wales: Nimmitabel, 36°31'S, 149°17’E, on Banksia sp., 1976, JAE (JAE 1623); Nimmitabel, 36°31'S, 149°15’E, on Prunus. 1980, GNS (BRIU3781NS). Australian Capital Territory: Mt Coree, 35°20’S, 148°50’E, on tree, 1970, E. Dahl (0). Victoria: Bonang, 17 km N of, 37°03’S, 148°49’E, on Acacia sp., 1978, D. Verdon (CBG7811003); Billywing Forest, 37°10’S, 142°30’E, on Acacia baileyi, 1981, RWR (BRIU2451RR); Spring Ck Gap, 37°15’S, 148°30’E, on Acacia sp., 1978, JAE (JAE 5137); Metung, 37°53'S, 147°51'E, on tree, 1889, F. R. M. Wilson (MEL 9420); Upper Maffra, 37°54'S, 146°51’E, on conglomerate, 1886, F. R. M. Wilson (NSW L4312); Hastings 5 km N of, 38°18’S, 145°11’E, on Avicennia marina, 1975, RWR (BRIU973RR); Warrnambool, 38°23'S, 142°20'E, 1886, F. R. M. Wilson (NSW L4311); Curdies Ck, 38°25’S, 143°00’E, on bush, 1885, F. R. M. Wilson (NSW L4315). South Australia: Springton 6-5 km W of, 34°43’S, 139°0S’E, on Hakea sp., 1976 JAE (JAE 2261); Mt Barker, 35°06’S, 138°51'E, on dead twig, 1981, RWR (BRIU2427RR). Tasmania: Trowutta, 41°02’S, 145°05’E, on Pyrrhus sp., 1981, GK 391/81; Mt Barrow Chalet, 41°22’S, 147°27’E, on Acacia dealbata, 1969, GB & JC (H032547); Upper Esk, 41°26'S, 147°44’E, on Acacia dealbata, 1981, GK 60/81; Cressy, 41°43’S, 147°02'E, on dead Acacia, 1964, JC (H032561); Central Plateau, 41°45’S, 146°50’E, on Athrotaxis cupressoides, 1980, GK 508/80; Friendly Beach, 42°00’S, 145°15’E, on Acacia sp., 1968, GB & JC (H032332); Shannon Lagoon, 42°08’S, 146°51’E, on dead eucalypt, 1980, GK 153/80; Lake Crescent, 42°10’S, 147°09’E, on dead Acacia, 1969, GB (H032544); Old Mans Head, 42°12’S, 147°13’E, on dead Acacia, 1964, GB &JC (H032333); Ouse, 29 km N of, 42°14’S, 146°44’E, on Acacia sp., 1964, GB & JC (H032557); Dee, 6km E of, 42°17’S, 146°38’E, on Acacia sp., 1970, GB (H032548); Tarraleah, 42°18’S, 146°26’E, on Acacia dealbata, 1980, GK 346/80; Ellendale, 42°38'S, 146°43'E, on old oak, 1980, GNS (BRIU3231NS); Sorrel-Nugent Rd, 42°46’S, 147°39’E, on briar and Acacia sp., 1964, GB (H032330); Cygnet, Bradleys Property, 43°09’S, 147°05’E, on dead willow, 1973, GB (H032322); Port Arthur, 43°13’S, 147°50’E, 1908, Mrs Townsend (NSW L4318). 13. Ramalina whinrayi [‘whinrayi? ] N. Stevens Plate 9, fig. 7. in Lichenologist 18: 187 (1986). Type: Australia, Hogans Island (Bass Strait), 1973, Whinray s.n. (MEL! — holotype; divaricatic acid + nordivaricatic acid). Thallus saxicolous, grey-green, caespitose, rigid, up to 1-0 cm high, branching sparse, branches arising from the base, branch width 1-0—2-5 mm, subterete to terete, rarely flattened, apices attenuate or blunt; surface matt; pseudocyphellae coarse, large, punctiform, numerous; holdfast delimited or diffuse when in colonies; soralia absent. Apothecia numerous, disc 3-10 mm diam., concave becoming plane to convex, terminal and lateral, or at the axils of bifurcating branches, spores ellipsoid, straight 12 x 4-5—5-0 wm. Chemistry. Divaricatic acid + nordivaricatic acid or sekikaic acid 4’-0-demethylsekikaic acid and usnic acid. Remarks. The anatomy of R. whinrayi differs from that of any of the other Australian taxa, but it is similar to the anatomy of R. capitata (= R. strepsilis) from the northern hemisphere. Distribution and habitat. Ramalina whinrayi is restricted in distribution to some of the Bass Strait islands and one site on the Tasmanian north coast (Fig. 19). This saxicolous taxon grows on granitic boulders at altitudes from 36-84 m above sea-level. The islands where this taxon occurs lie along a southeast line — from the Hogan Group, the Kents Group to Craggy Island between latitudes 39°12’S and 40°45’S. Ramalina whinrayi also occurs at The Nut, Stanley, Tasmania, where it occurs in overhangs on a basaltic cliff facing the ocean and exposed to strong winds. World distribution. Australia. Bass Strait: Hogans Group: Long Islet, 39°12’S, 147°00’E, on granite, 1973, JW (MEL1012965); Hogans I, 39°13'S, 146°59’E, on granite, 1973, JW (MEL1012948) Type; Kents Group: North East I, 39°27’'S, 180 G. NELL STEVENS a Bia 120 130 140 160 104 20~ 30+ 40% L | \ t Fig. 19 Distribution of R. whinrayi. 147°25'E, on granite, 1971, JW (MEL1012636); Craggy I, 39°41’S, 147°40’E, on granite, 1972, JW (MEL1013031). Tasmania: The Nut, near Stanley, 40°45’S, 145°18’E, on basalt, 1973, GB (H032796). 14. Ramalina celastri (Sprengel) Krog & Swinscow in Norw. J. Bot. 23: 159 (1976). Parmelia celastri Sprengel, Syst. Veget. 4 (2): 328 (1827). Type: Cap. B. Spei, Eklon (S! — holotype; FH! - isotype; usnic acid). subsp. celastri Plate 9, fig. 8; Plate 10, fig. 1; Plate 15, fig. 4. Ramalina fraxinea B yemensis Ach., Lichenogr. Univ.: 602 (1810). Type: Arabia, Forskal, herb. Acharius (H! — holotype). — Ramalina yemensis (Ach.) Nyl. in Bull. Soc. Linn. Normandie II, 4: 144 (1870). Ramalina ecklonii auct., non Sprengel. Thallus corticolous, rarely saxicolous, pale green to grey-green, erect, subpendulous to pendulous, up to 12 cm long, exceptionally to 30 cm; branching initially trichotomous, either Plate10 Thallus growth forms in the Australian Ramalinae. Scale in mm. Fig. 1 R. celastri subsp. celastri (common form). Fig. 2 R. celastri subsp. ovalis (broadest form). Fig. 3 R. inflata subsp. perpusilla (coastal ecotype). Fig. 4 R. inflata subsp. perpusilla (inland ecotypes). Fig. 5 R. inflata subsp. inflata (open-forest ecotype). Fig. 6 R. inflata subsp. inflata (closed-forest ecotype). Fig. 7 R. inflata subsp. australis: LHS divaricatic acid specimen RHS sekikaic acid specimen. Fig. 8 R. fissa (branch at centre top of photo shows medulla due to splitting apart). RAMALINA IN AUSTRALIA TTTITITIFIDT eee 182 G. NELL STEVENS remaining as a three-lobed thallus or thereafter dividing subdichotomously or irregularly, often producing densely branched thalli, rarely producing small branchlets at right angles to the primary branches; branch width (1—)2—5 mm exceptionally to 20 mm, branches compressed ranging from narrow and canaliculate with apices attenuate to broad and flat with apices acute or blunt; surface matt, smooth to rugose, membraneous or coarse and ridged with strands of chondroid tissue, cracks or holes may appear between these strands; pseudocyphellae usually present; holdfast delimited; soralia absent. Apothecia numerous, marginal on narrow branches, laminal and marginal on broader branches; apothecia occasionally appear on both upper and lower surfaces of the branch; disc 0-5-2-0 mm diam., plane to convex; margin narrow, entire, smooth or crennate, almost disappearing at maturity; spores oval, ellipsoid to slightly fusiform, mainly straight to slightly curved, (10—)12-16 x 4-6 wm. Chemistry. Usnic acid only. Remarks. The name R. ecklonii (Sprengel) Meyen & Flotow was incorrectly applied to this species for a long time. Krog & Swinscow (1976) rejected the name R. ecklonii, applying Article 69 of the Code, and chose R. sprengelii as a nomen novum for it. Material collected by Ecklon from Cape of Good Hope was cited as the holotype. This material is very small, whereas an isotype held at FH in Tuckerman herbarium comprises a number of thalli in good condition. All of these specimens are narrow and canaliculate in form. Comparison of the type material with the narrow ecotype found in Australia showed them to be morphologically similar. Specimens bearing this narrow morphology are often confused with R. linearis Swartz, but there are distinct differences which easily separate the two (Stevens, 19835). Broad thalli which belong to R. celastri subsp. celastri resemble the morphology of the holotype of R. yemensis (Ach.) Nyl. (R. fraxinea B yemensis Ach.). This variability in the width of branches caused many new taxa to be described, some of which Hue (1890: 265) listed as synonyms of R. yemensis. Weber & Wetmore (1963, 1972) listed the names R. ecklonii Sprengel, R. ecklonii var. membranacea (Laurer) Mill. Arg., R. ecklonii var. tenuissima Meyen & Flotow, R. linearis Sw.., and R. yemensis (Ach.) Nyl. as occurring in Australia, but examination of these specimens has identified them as belonging to R. celastri subsp. celastri. Examination of some South American material indicates several of these species may be morphotypes of R. celastri subsp. celastri. R. ecklonii var. lobulifera Malme closely resembles the broad morphotype found in Australia. Other species, e.g. R. laevigata and R. continentalis, appear to be morphotypes of R. celastri subsp. celastri, but more research into the South American taxa is necessary before taxonomic conclusions can be established. The position of R. cumanensis Fée in this group of taxa has yet to be determined; narrow canaliculate morphotypes in Australia have been found which closely resemble it. Krog & Swinscow (1976) commented that a specimen of R. celastri from Mt Kenya resembled R. cumanensis, and Landroén (1972) was convinced that R. cumanensis and R. ecklonii were the same species and reduced the latter name to synonomy. This problem is still to be resolved. Distribution and habitat. Ramalina celastri subsp. celastri morphotypes extend from the subtro- pics (26°S) to the temperate region (38°S) (Fig. 20), with the coastal occurrence more restricted in latitude (28°S to 35°S). The inland distribution is extensive in eastern Australia, occurring up to 200 km from the coast. It also has a wide altitudinal range, extending from sea-level to the mountains. Narrow canaliculate morphs occur in montane forest habitats on a variety on tree species. In these habitats the rainfall usually exceeds 1500 mm per annum and mists are common throughout the year. Broad, flat thalli (some reaching widths of over 10 mm) occur in the drier open-forest areas on several phorophytes, species of Acacia being the most common, e.g. A. leiocalyx, A. concurrens, A. irrorata, A. orites, A. melanoxylon, and A. mearnsi. The lichen also occurs on Alphitonia excelsior and on the thorny shrubs Citriobatus pauciflora and Capparis arborea. ; Ramalina celastri subsp. celastri is usually found growing in association with the fistulose taxa RAMALINA IN AUSTRALIA 183 120 130 140 160 10+ 30~ nN l Fig. 20 Distribution of R. celastri subsp. celastri @ and subsp. ovalis O. R. inflata subsp. perpusilla and R. inflata subsp. inflata up to 1000 m, and with R. reducta at altitudes above 1300 m. World distribution. Pan-subtropical. Queensland: Narayen, 25°48’S, 151°10’E, on scrub tree, 1980, N. Gibson (BRIU3287NS); Kingaroy, 3 km SW of, 26°34’S, 151°48’E, on dead shrub, 1979, HS (CBG8000062); Jimna, 26°40’S, 152°28’E, on Acacia sp., 1980, GNS (BRIU3354NS); Bunya Mts, 26°50’S, 151°40’E, on Acacia sp., 1979, GNS (BRIU3289NS); Mt Mee, 27°06’S, 152°42’E, on Citriobatus pauciflora, 1971, RWR (BRIU1919RR); Maclagan, 27°10’S, 151°38’E, on Acacia sp., 1979, GNS (BRIU3290NS); Toowoomba, Highfields, 27°23'S, 151°58’E, on Alphitonia sp., 1983, GNS (BRIU4066NS); Rosewood, 27°39’S, 152°36’E, on Alphitonia sp., 1982, GNS (BRIU3559NS); Pittsworth, 20 km W of, 27°45’S, 151°27’E, on Carissa ovata, 1983, M. Olsen, (BRIU3761NS); Tamborine Mt 27°55'S, 153°12’E, on shrub, 1979, GNS (BRIU3012NS); Mt Cordeaux, 28°03’S, 152°23’E, on rainforest tree, 1978, GNS (BRIU3099NS); Warwick, Apex Park, 28°13'S, 152°02’E, on oak tree, 1980, GNS (BRIU3235NS); Binna Burra, 28°12’S, 153°11'E, on Acacia sp., 1979, RWR (BRIU2153RR); O’Reillys, 28°15’S, 153°08’E, on Eugenia sp., 1983, GNS (BRIU3735NS); Springbrook, 28°14'S, 153°16’E, on Acacia orrites, 1983, GNS (BRIU3759NS); Wilsons Peak, 28°15’S, 152°30’E, on Acacia sp., 1978, GNS (BRIU3829A.NS); Kelvin Falls, 28°27'S, 152°06’E, on Casuarina sp., 1980, NCS (BRIU3286NS). New South Wales: Undercliff Falls, 28°38’S, 152°10’E, on Acacia melanoxylon, 1981, GNS (BRIU3652NS); Lismore, 28°48’S, 153°16’E, on Acacia sp., 1976, R. Moodie (BRIU2850NS); Yamba, 29°27’S, 153°20’E, on Avicennia marina, 1977, GNS (BRIU2124A.NS); Glen Innes, 35 km S of, 30°00’S, 151°40’E, on Acacia sp., 1980, GNS (BRIU3232NS); Clouds Ck, 30°10’S, 152°35’E, on small tree, 1978, HS (CBG7905564); Wollomombi Falls, 30°32’S, 152°03’E, on Acacia sp., 184 G. NELL STEVENS 1982, GNS (BRIU3776NS); Port Macquarie, 31°26’S, 152°54’E, on Avicennia marina, 1977, GNS (BRIU2124NS); Walcha Road, 31°25’S, 151°30’E, on Acacia sp., 1982, GNS (BRIU3764NS); Cobak R. 31°57’S, 151°48’E, on Casuarina sp., 1982, GNS (BRIU3770NS); Gosford, Erina Ck, 33°25’S, 151°20’E, on Casuarina glauca, 1978, JAE (JAE 4722); Jenolan Caves, 33°48’S, 150°01'E, on Acacia sp., 1977, JAE (JAE 3180); Wombeyan Caves, 34°19’S, 149°58’E, on dead shrub, 1979, HS (CBG7907795); Nowra, 14 km N of, 34°48’S, 150°40’E, on tree, 1976, HS (3456); Bundanoon, 34°39'S, 150°18’E, on tree, 1954, M. Tindale (MEL9457); Bungendore, 12 km E of, 35°20’S, 149°28’E, on old bones, 1978, JAE (JAE 5465); Batemans Bay, 35°45’S, 150°10’E, on Avicennia marina, 1978, GNS (BRIU2170NS); Murunna Pt, 36°24'S, 150°05’E, on dead coastal shrub, 1978, JAE (JAE 4591); Twofold Bay, 37°05’S, 149°54'E, White (MEL 9456). Victoria: Lakes Entrance, 37°53’S, 148°00’E, 1901 (MEL 9465); Frankston, 38°11'S, 145°03’E, 1900 (MEL 9462); Traralgon, 8 km E of, 38°12'S, 146°32’E, on Melaleuca sp., 1975 RF (MEL1013204); Yarragon, Strzelecki Ra. 38°12’S, 146°04’E, on trees, 1973, GB & R. C. Weeks (H032292); Trida, Grand Ridge Rd, 38°19’S, 145°58’E, on trees, 1971, GB & R. C. Weeks (H032290); Gippsland, Foster 6-4 km NW of, 38°40’S, 146°13’E, on dead wood, 1968, GB (H032297). Tasmania: Trowutta, 41°02'S, 145°05’E, on Pyrrhus sp., 1981, GK 386/81. subsp. ovalis (J. D. Hook. & Taylor) N. Stevens, comb. nov. _— Plate 10, fig. 2; Plate 15, fig. 5. Ramalina ovalis J.D. Hook. & Taylor in Lond. J. Bot. 3: 655 (1844). Type: Van Dieman’s Land, Gunns.n. (BM! - holotype; usnic acid). — Ramalina calicaris var. ovalis (J. D. Hook. & Taylor) Church. Bab. in J. D. Hook., Flora Tasmaniae 3: 344 (1860). — Ramalina ecklonii var. ovalis (J. D. Hook. & Taylor) F. Wilson in Pap. Proc. R. Soc. Tasm. 1892: 159 (1893). — Ramalina yemensis var. ovalis (J. D. Hook. & Taylor) Zahlbr., Cat. Lich. Univ. 6: 529 (1930). Thallus corticolous, green-grey, erect to subpendulous, 2—4(—5) cm long, exceptionally to 8 cm; branching sparse with one or two branches often constituting the whole thallus; branch width 8-20 mm, compressed, flat, apices blunt or rounded, small branchlets may occur at the apices; surface matt, membraneous and thin, wrinkled, or thick and ribbed with chondroid strands; pseudocyphellae sometimes present; holdfast delimited; soralia absent. Apothecia laminal only, numerous, disc to 2-0 mm diam., plane to convex; margin entire, thin; spores ellipsoid or slightly fusiform, mostly curved, (10—)12—16 x 4-4-5 wm. Chemistry. Usnic acid only. Remarks. There is no record of the exact location of the collection of the type material of R. ovalis apart from Van Diemans Land. The obovate form of the branches was commented on by Hooker & Taylor (1844) in the type description — ‘has the outline of Fucus ovalis’ — and this morphology is common in Tasmanian material, but mainland specimens are more variable. From the taxonomic nomenclature applied to this taxon in the past it appears most authors appreciated its affinity to the species R. celastri (= R. yemensis, = R. ecklonii). Distribution and habitat. This temperate taxon occupies a broad band across Victoria, South Australia, and Western Australia in the latitudes 33°S to 38°S, and is found on the Bass Strait islands and in Tasmania along the north and eastern coast, latitudes 40°S—43°S (Fig. 20). It virtually replaces R. celastri subsp. celastri in these latitudes and is confined to areas of winter rainfall. It grows mainly in open-forest and tall shrubland communities. Krog (in litt.) detected the occurrence of this broad taxon in collections of R. celastri from East Africa. Examination of the BM material from Tanzania, Moshi district (1440 m alt.) and Uganda (1950-2100 m alt.) showed that thalli with the same morphology as R. celastri subsp. ovalis grow in these open-forest areas. This taxon also occurs in New Zealand at comparable latitudes to those where it is found in Australia. The main phorophytes of R. celastri subsp. ovalis are Acacia pycnantha, A. melanoxylon, A. acuminata, A. longifolia, A. dealbata, and Lycium ferocissimum, Bursaria spinosa, Banksia marginata, Casuarina stricta, Callitris sp., and Pyrrhus, Populus, Malus, and Prunus spp. World distribution. Africa and Australia. New South Wales: Neville, 33°43’S, 149°13’E, on Acacia sp., 1980, GNS (BRIU3302NS); Crookwell, 34°28'S, 149°29’E, on Acacia sp., 1980, GNS (BRIU3233NS). Victoria: Marlo, 37°48’S, 148°32’E, on Leucopogon sp., 1980, GNS (BRIU3221NS); Korumburra, Coal Ck, 38°25’S, 145°48’S, on Acacia sp., RAMALINA IN AUSTRALIA 185 1980, GNS (BRIU3248NS); Wilsons Prom, N of, 38°45’S, 146°08’E, on Acacia sp., 1980, GNS (BRIU3219NS); Crib Pt, Westernport Bay, 38°22’S, 145°10’E, on Avicennia marina, 1978, GNS (BRIU2346NS); Gunnamatta Beach, 38°23'S, 144°45’E, on dune shrub, 1976, RF (MEL1025871); Ballarat, 20 km WNW of, 37°30’S, 143°44’E, on Pinus sp., 1975, E. M. Canning (CBG8100893); Portland, 38°21'S, 141°36’E, 1947, A. C. Beauglehole (MEL1020517). South Australia: Oodlawirra, 5 km W of, 32°53'S, 139°03’E, on Muellenbeckia sp., 1969, RWR (BRIU1632RR); Eyre Peninsula, W of Lake Newland, 33°22’S, 134°53’E, 1967, N. N. Donner (AD97539274); Yorke Peninsula, Cape Elizabeth, 34°08'S, 137°27’E, 1969, B. Copley (AD97414422); Port Rickaby, S. of, 34°41’S, 137°29’E, 1974, J. Z. Weber (AD97648159); Springton, 14 km SE of, 34°39’S, 139°06’E, on Callitris sp., 1974, JAE (JAE 481); Mt Lofty Ra., Clarendon, 35°07'S, 138°38’E, 1968, V. Cruickshank (AD97418421); Trent Hill, 35°36’S, 138°12'E, on Casuarina sp., 1968, RWR (BRIU1458RR); Naracoorte, 36°58’S, 140°45’E, D. Hunt (CBG067824); Kangaroo I, W of Kingscote airport, 35°39’S, 137°38’E, on Eucalyptus sp., R. D. Seppelt (MEL1024502). Western Australia: Boyup Brook, 33°5S0’S, 116°22’E, on Malus sp., 1974, K. M. Allan, Lake King, 33°05’S, 119°40’E, on twigs, 1980, D. Richardson (PER000996); Northcliffe, Beedalup Falls Rd, 34°38’S, 116°07’E, on dead tree, 1973, B. Muir (N. Sammy Herb); Recherche Archipelago, Peak I, 34°00’S, 122°50’E, on Acacia acuminata, 1950, J. H. Willis (MEL 10025); Long I, 34°03’S, 121°58’E, on Melaleuca sp., 1950, J. H. Willis (MEL 10020). Bass Strait: Curtis I, 39°29’S, 146°39’E, 1971, McCoy Expedition, (MEL40274); Furneaux Group: Forsyth I, 40°38’S, 148°25’E, on Acacia sp., 1976, JW (MEL1516801). Tasmania: Stanley, The Nut, 40°45’S, 145°15’E, on Bursaria spinosa, 1965, J. H. Willis (MEL 7588); Devonport airport, 41°10'S, 145°20’E, on Poplar sp., 1980, GNS (BRIU3236); Avoca, 41°48'S, 147°40’E, on Prunus sp., 1978, GK 2/78; Sorrell-Nugent Rd, 42°46’S, 147°39’E, on briars, 1964, GB & JC (H032314). 15. Ramalina inflata (J. D. Hook. & Taylor) J. D. Hook. & Taylor in J. D. Hook., Flora Antarctica 1: 194 (1845). Cetraria inflata J.D. Hook. & Taylor in Lond. J. Bot. 3: 646 (1844). Type: Lord Auckland Islands, Hooker s.n. (FH! —holotype; BM! — isotype; divaricatic acid and usnic acid). — Fistulariella inflata (J. D. Hook. & Taylor) Bowler & Rundel in Mycotaxon 6: 195 (1977). subsp. inflata Plate 10, figs 5—6; Plate 14, fig. 5. Ramalina tasmanica Ny). in Bull. Soc. Linn. Normandie II, 4: 162 (1870) type: Australia, Tasmania, J. D. Hooker s.n. (H! — holotype [H-NYL 36849 ex hb. Churchill Babington]; divaricatic acid and usnic acid). Plate 14, fig. 6. Ramalina knightiana Zahlbr., Cat. Lich. Univ. 6: 494 (1930); Ramalina subgeniculata Knight in Trans. Linn. Soc. Lond. (Bot.) 2: 50 (1882), non Nyl. (1870) (Art 64.1). Type: Australia, New South Wales, near Sydney, Knight s.n. (WELT! — holotype; divaricatic acid, sekikaic acid, and usnic acid). Thallus corticolous, green to pale green, caespitose, erect 1—-2(-—3) cm high; branching subdichotomous or irregular; branch width 1-3 mm, branches hollow, terete, inflated and perforate; perforations round to elongate, medullary hyphae continuous, loosely woven across central cavity or compressed against the inner cortex wall either loosely or densely; surface matt to shiny, smooth, rarely pseudocyphellate; holdfast delimited; soralia absent. Apothecia common, terminal on main branches and on short subapical lateral branches, often spurred; disc 2-5 mm diam., always concave initially, innate at branch apices, becoming plane at maturity; margin entire, thin often indistinct; spores broadly ellipsoid, straight or curved, 12-16 x 4—-5(-—6) wm. Chemistry. Divaricatic acid, + nordivaricatic acid, + sekikaic acid, and usnic acid. Remarks. Sterile specimens of this taxon are usually larger than fertile thalli, occurring either as greatly inflated, pulvinate, thalli or having elongate branches with perforations evenly distrib- uted along the lower surface and extending to the attenuate apices. A small morphotype of R. inflata subsp. inflata found in Tasmania was named R. tasmanica Nyl.; the type material contains divaricatic acid. Subsequent collections from Tasmania bearing this name have been found to contain salazinic acid and belong to R. fissa. The type material of R. knightiana Zahlbr. is a very finely branched specimen of R. inflata subsp. inflata collected in the neighbourhood of Sydney; its morphology closely resembles the New Zealand species R. geniculata and was originally named R. subgeniculata by Knight. 186 G. NELL STEVENS The synonomy of R. inflata and R. geniculata has been the subject of debate by several lichenologists (Howe, 1913-14; Landr6én, 1972; Bowler & Rundel, 1977). An examination of the types of both of these species during the present study has shown them to contain the same acid (divaricatic) but to differ morphologically. More material from New Zealand (from which country the type of R. geniculata was collected) would need to be examined to evaluate the amount of phenotypic plasticity displayed by this taxon and to compare the morphotypes with R. inflata ecophenotypes, in order to establish if both taxa intergrade with one another. In many respects the morphology and habitat requirements of the East African taxon R. calcarata Krog & Swinscow come close to the Australian specimens of R. inflata subsp. inflata, but the presence of salazinic acid in the hymenium of the former taxon distinguishes the two. A tan coloration on the lower surface of some specimens of R. calcarata was mentioned by Krog & Swinscow (1974) and this colour occasionally occurs on the Australian material of R. inflata subsp. inflata. However discoloration of the thallus is not common and is not regarded as of any taxonomic significance. The original protologue of R. inflata contains the comment: ‘sometimes stained with dark blood red’, but this colour is rarely evident on specimens. However black patches on the thallus of R. pusilla Le Prev. ex Duby., have been commented upon by Krog & @sthagen (1980). Distribution and habitat. The distribution of R. inflata subsp. inflata (Figs 21, 22) extends along the warm to cool temperate section of the eastern and south-eastern Australian coastal belt and into the cool/cold temperate region of Tasmania. The most northern occurrence of this taxon is T 120 130 140 160 \ \ 1 | Fig. 21 Distribution of R. inflata subsp: inflata O, subsp. perpusilla @, and subsp. australis ®. RAMALINA IN AUSTRALIA 187 120 130 140 1bo % ss 104 ° Ae er es da GB © e \ e f 20, ere a oe ae Merl Wen ea reese Stee ot Ly te fa a 6 e/ es reps! “ Fe % Ce 3 304 eo. e fe) é rx Cen ; oe RO e @ (one) Cele® o e ® @ © @60e e @ SS » 6% i an men \ 1 | 1 Fig. 22 Distribution of R. inflata complex: @ divaricatic acid race; O sekikaic acid race. in the mountains at elevations of 1000 m in the area 28°—30°S latitude. This is an area where intermediates between the morphology of the rain-forest ecotype of R. inflata subsp. perpusilla and the morphology of R. inflata subsp. inflata occur. At 30°S latitude (Wollomombi Falls, elevation 1000 m) R. inflata subsp. inflata grows on Acacia in open forest communities growing on metamorphic rock, and the thalli are typical; at 32°S latitude (Barrington Tops, a basalt range) thalli collected from rain-forest areas were difficult to place in either taxon, indicating that the microhabitat as well as climatic conditions may be influencing these two subspecies. The most common phorophytes for R. inflata subsp. inflata are Acacia and Casuarina spp. World distribution. South Africa, Australia, Auckland Island, Brazil, and Venezuela. [Uncon- firmed reports are from China (Zahlbruckner, 1930), North America (Howe, 1914), West Indies (Landron, 1972), and Hawaii (Magnusson, 1955)]. New South Wales: Undercliff Falls, 28°38’S, 152°10’E, on Acacia sp., 1980, GNS (BRIU3028A.NS); Wollomombi Falls, 30°30’S, 152°03’E, on Acacia sp., 1982, GNS (BRIU3639NS); Barrington Tops, 32°00’S, 151°30’E, on rain-forest tree, 1982, GNS (BRIU3636NS); Scone, 12 km W of, 32°05’S, 150°55’E, on bush, 1982, GNS (BRIU3635NS); Tuncurry, 32°10’S, 152°27’E, on Casuarina sp., 1982, GNS (BRIU3648NS); Erina Ck, 33°25’S, 151°21’E, on Casuarina glauca, 1978, JAE (JAE4705); Towra Pt, 34°02'S, 151°13’E, on Casuarina sp., 1977, CS (BRIU63CS); Barrengarry Mt, 34°41’S, 150°34’E, on tree, 1976, JAE (JAE2714); Currowan State Forest, 35°39’S, 150°08’E, on Acacia sp., 1977, JAE (JAE3576); Batemans Bay, 35°42'S, 150°10’E, 1970, ED (0); Batemans Bay, on Avicennia marina, 1978, GNS (BRIU2172NS); Wallaga Lake, 36°23’S, 150°03’E, on Casuarina sp. n., 1978, JAE (JAE 4543); Spring Ck 188 G. NELL STEVENS Gap, 37°30’S, 148°20’E, on Acacia sp., 1978, JAE (JAE5156); Marlo, 37°48'S, 148°32'E, on Acacia sp., 1978, GNS (BRIU3282NS); Cape Conran, 37°49'S, 148°45’E, on Leptospermum sp., 1983, GNS (BRIU3676NS); Rodondo I, 39°15’S, 146°23’E, 1970, J. Kirkpatrick (MEL39969); Cape Patterson, 38°40'S, 145°35’E, on Acacia pycnantha, 1963, RF & Smith 3413; French I, Tankerton Rd, 38°20’S, 145°20’E, S. Filson (MEL1026170); Healesville, 37°39’S, 145°32’E, on twigs, 1974, RB (MEL1013089); Mt Richmond Nat. Park, 38°18’S, 141°30’E, 1978, AA (BRIU4312NS). Bass Strait: Kents Group: Erith I, 39°28'S, 147°13’E, on Beyeria leschenaultia, 1970, JW (MEL1012899). Tasmania: Rocky Cape, 40°51’S, 145°31'E, 1975, D & M. Cook (H032566); Yolla, 41°09’S, 145°40’E, on Malus sp., 1980, GNS (BRIU3239NS); Mt Barrow, 41°23’S, 147°25’E, on Acacia sp., 1966, GB (H032678); Liffey Falls, 41°41’S, 146°47'E, on Acacia sp., 1969, GB & JC (H032752); Tarraleah, 42°15'S, 146°26’E, on Acacia dealbata, 1980, GK 347/80; Mt Faulkner, 42°48’S, 147°12'E, on Acacia sp., 1965, GB & JC (H032762). subsp. perpusilla (Stirton) N. Stevens, comb. nov. Plate 10, figs 3—4; Plate 14, fig. 3. Ramalina perpusilla Stirton in Proc. R. Soc. Vict. 17: 68 (1881). Type: Australia, Queensland, Brisbane, Bailey 113 (BRI! — lectotype; herein designated, left hand specimen; divaricatic acid and usnic acid). Ramalina minuscula var. alba C. Knight ex Shirley in Proc. R. Soc. Qd 5: 103 (1888). Type: Australia, Queensland, Mt Perry, Keys 91 (WELT! — lectotype; sekikaic acid and usnic acid). Plate 15, fig. 7. Ramalina geniculata var. compacta Mill. Arg. in Bull. Herb. Boissier 4: 88 (1896). Type: Australia, Queensland, Toowoomba, Hartmann s.n. (G! — lectotype designated here; divaricatic acid and usnic acid). Thallus corticolous, green to pale green, often with white lower surface, caespitose, erect, 1—2(-3) cm high; branching subdichotomous, branch width 0-5—2-0 mm, varying in the amount of inflation, terete to almost flat, sparingly to moderately perforate, the perforations on the lower surface only, small and round, extending to the apices on sterile branches; medullary hyphae sparse and adhering to the inner cortex wall either loosely or densely compacted in totally inflated specimens or thick and felty in compressed thalli; surface matt or shiny, smooth to rugose; very rarely pseudocyphellate; holdfast delimited or diffuse; soralia absent. Apothecia sparse to numerous, subterminal on geniculate branches or terminal on main and short lateral branches, spurred, or in the axils of bifurcating branches, disc usually less than 2-0 mm diam., plane to convex; margin entire, well defined, rarely indistinct; spores ellipsoid, straight, rarely curved, (8—)10—14(—16) x 4-5 um. Chemistry. Divaricatic acid + trace sekikaic acid, + nordivaricatic acid or sekikaic acid only, or rarely sekikaic acid, 4’-0-demethylsekikaic acid, an unknown and usnic acid. Remarks. The type material of R. inflata subsp. perpusilla comprises four specimens with the number 113 apparently covering three of the specimens and the number 259 covering the fourth specimen. The left hand specimen contains divaricatic acid and is the most likely single specimen that number 113 could refer to (Plate 14, fig. 3). The middle and right hand top specimens contain sekikaic acid as does specimen 259. The left-hand specimen has been selected as lectotype for this taxon because it appears to conform best with the original protologue. As the majority of specimens collected during this study contain divaricatic acid as the major acid, the choice of a lectotype which contained divaricatic acid seemed appropriate. The only locality given for the type collection is Brisbane River. In describing this species later, Shirley (1888) gave the information ‘Hocking’s Nursery, Brisbane River’. It is not clear however, whether the specimens were growing on native trees in the area, or on the mangroves, or were found on one of the nursery trees, which could have come from elsewhere. Specimens which produce sekikaic acid only have not been found on mangroves; they are usually found in dry inland situations, which would indicate that the type material possibly grew on a plant which had come from another location. This problem remains unsolved and the true type locality is not known. Ramalina minuscula var. alba was collected at Mt Perry, a dry open-forest area in southern Queensland. The type material consists of 11 carded specimens all of which contain sekikaic acid. Ramalina geniculata var. compacta was collected in the Toowoomba area. The material comprises six specimens on a card (a, b, c, d, e, f); b and d contain divaricatic acid whereas a, c, € RAMALINA IN AUSTRALIA 189 and f contain sekikaic acid. Specimen b is designated lectotype as it conforms best with the original protologue and contains divaricatic acid, consistent with the type of R. inflata subsp. perpusilla with which it is synonymous. As Toowoomba is an inland region about 130 km from the coast, the presence of sekikaic acid specimens as well as divaricatic acid specimens is not unusual because fresh collections made during the present study from inland areas always contained sekikaic acid thalli as well as divaricatic acid thalli. The sekikaic acid race has not been given any taxonomic rank because of the co-existence of both divaricatic and sekikaic acids in many of the thalli. The morphology of the East African taxon R. consanguinea Mill Arg. from the mangroves resembles R. inflata subsp. perpusilla specimens from the mangrove communities. Both produce apothecia which are plane to convex but never concave, and the former taxon has short slender side branches and spinules on the main branches, a feature sometimes present on Australian specimens. Ramalina consanguinea, however, contains both divaricatic acid and norstictic acid and produces large spores 20—25(—28) x 4—5(—6) wm, which sets it apart from R. inflata subsp. perpusilla. Ramalina pumila Mont., an Asian taxon, appears to be close to the Australian taxon morphologically, but its chemistry (evernic acid with trace amounts of sekikaic acid) sets the two taxa apart. Distribution and habitat. Ramalina inflata subsp. perpusilla has a distribution range from the tropics to the subtropics (Figs 21, 22). The highly variable morphology found in this taxon over this range is ecophenotypic. Three distinct habitats are considered to explain the ecotypes that occur, e.g. (1) the mangrove habitat, (2) the open-forest habitat, and (3) the rain-forest (closed-forest) habitat. (1) The mangrove habitat: Specimens of R. inflata subsp. perpusilla collected from mangrove communities along the Queensland coastline from Hallidays Bay (latitude 20°56’S) to Moreton Bay (latitude 28°S) are usually small and caespitose with elongate perforations spread irregu- larly along the lower surface; short spinules are sometimes present on the main branches and the margins of the apothecia are very distinct. North of latitude 23°50’S (Gladstone), 80% of the specimens were found to contain only divaricatic acid whereas 45% of specimens south of Gladstone contained divaricatic acid and trace amounts of either sekikaic acid or nordivaricatic acid. The main mangroves which act as phorophytes for this taxon are Ceriops tagal and Rhiz- ophora stylosa, and when these mangroves reach their southern distribution limit at latitude 28°S, R. inflata subsp. perpusilla ceases to occur along the coastline. (2) The open-forest habitat (including disturbed habitats): These dry inland areas lie between the maritime and the montane-rain-forest environments and collections of R. inflata subsp. perpusilla from such habitats are scattered from latitudes 18°S—27°S. The thalli are very small with a maximum size of less than 1-0 cm high; they are usually green or pale green with a smooth, shiny surface, wholly inflated; perforations are round and evenly distributed on the lower surface of the thallus; the branches are either tightly packed together with numerous terminal apothecia covering the thallus (Plate 10, fig. 3) or are palmate at the base, dividing into several branches which bear subterminal apothecia or apothecia situated in the axils of bifurcating branches. In this open-forest habitat more than 50% of the population contains sekikaic acid only with the remainder containing divaricatic acid + trace amounts of sekikaic acid. The most common phorophytes in such habitats are the small thorny shrubs Citriobatus pauciflora and Acalypha capillipes. They grow on basaltic soils in open areas experiencing full sunlight and varying humidity; areas which in some respects resemble the open mangrove community where Ceriops tagal grows. (3) The rain-forest habitat: Pockets of rain-forest where this particular ecotype occurs extend from the tropics to the subtropics along the Eastern Highlands (viz. Atherton Tableland (17°S), Mt Surprise (19°S), Bunya Mts (27°S), McPherson Ra. (28°S), and Dorrigo Plateau (30°S). Specimens of R. inflata subsp. perpusilla from this habitat tend to be loosely branched, 2—3 cm 190 G. NELL STEVENS long, with a pale lower surface; branching is subdichotomous with the branches less inflated, even partially compressed, and more rugose than those of the taxa from more exposed habitats; the perforations are small, slightly elongate and sparse; the apothecia are usually terminal and subterminal, with some tendency to be innate when immature but becoming plane at maturity. The majority of the rain-forest specimens contain only divaricatic acid, with a small pro- portion producing trace amounts of either sekikaic acid or nordivaricatic acid in addition to divaricatic acid. The main phorophytes are various species of rain-forest tree, including Araucaria cunninghamii and Eugenia sp., and also Acacia melanoxylon and A. orrites which fringe the rain-forest borders. Although this taxon occupies three distinct habitats, there is total intergradation of morpho- types from one environment to another which prevented any division of the group being made based on microhabitat morphology. Chemistry was not able to be used taxonomically, as variation from divaricatic acid only in the thallus to divaricatic acid + trace sekikaic acid to sekikaic acid only was found to occur. One inflated specimen collected from Lake Eacham, Atherton Tableland (JAE 2427), has not been included in this taxon; it is a large specimen (4 cm long), partially inflated, partially compressed (which is normal for the rain-forest habitat in which it grew) with white punctiform pseudocyphellae along the branches. The acids present are sekikaic, homosekikaic, 4’-0- demethylsekikaic, and an unknown. This chemistry differs from that produced by the majority of the taxa in this group but it is similar to the acids found in a few inland specimens of R. inflata subsp. perpusilla. The presence of white pseudocyphellae, however, is quite distinct. As no other specimens have been collected from Atherton Tableland, or elsewhere, the specimen remains unnamed at present. World distribution. Australia. Queensland: Tinaroo Dam, 17°16’S, 145°34’E, on Araucaria cunninghamii, 1983, GNS (BRIU4167NS); Forty Mile Scrub, 18°03’S, 144°53’E, on thorny shrub, 1983, GNS (BRIU4114NS); Burdekin R, Big Bend, 19°15’S, 146°12’E, on thorny bush, 1983, GNS (BRIU4154NS); Hallidays Bay, 20°56’S, 149°01’E, on Ceriops tagal, 1983, GNS (BRIU4118); Eimeo Beach, 21°05’S, 149°13’E, on Ceriops tagal, 1976, GNS (BRIU2221NS); St Lawrence, 22°20'S, 149°32'E, on Ceriops tagal, 1983, GNS (BRIU4105NS); Marl- borough, 22°50’S, 149°51’E, on thorny bush, 1983, M. Olsen (BRIU4157NS); Port Alma, 23°35’S, 150°42’E, on Rhizophora stylosa, 1975, RWR (BRIU696RR); Gladstone, 23°50’S, 151°16’E, on Ceriops tagal, 1980, GNS (BRIU3653NS); Cania Gorge, 24°38'S, 150°56’E, on Casuarina sp., 1984, NCS (BRIU4343NS); Bundaberg, The Hummock, 24°51'S, 152°26’E, rainforest tree, 1977, I. R. Telford (CBG7702460); Goodnight Scrub, 25°15’S, 151°50’E, on thorny bush, 1981, P. Forster (BRIU3640NS); Toogoom, 25°15’S, 152°40’E, on Callitris sp., 1983, GNS (BRIU4150NS); Coalston Lakes, 25°35’S, 151°55’E, on thorny bush, 1980, RWR (BRIU3586RR); Narayan, 25°48’S, 151°10’E, on shrub, 1980, RWR (BRIU2197RR); Barambah Ck, 25°40’S, 151°45’E, on shrub, 1982, NCS (BRIU3606NS); Mt Blandy, 25°40'S, 151°38’E, on thorny bush, 1980, P. Forster (BRIU3641NS); Lake Cootharaba, 26°18’S, 152°58’E, on Bruguiera sp., 1978, GNS (BRIU3620NS); Maleny, 26°46’S, 152°51'E, on Acacia sp., 1982, GNS (BRIU3568NS); Bunya Mts, 26°50’S, 151°40’E, on rainforest shrub, 1979, GNS (BRIU3009NS); Mt Mee, 27°06'S, 152°42’E, on Citriobatus pauciflora, 1979, GNS (BRIU3654NS); Stradbroke I, 27°25'S, 153°27'E, on Lumnitzera racemosa, 1977, GNS (BRIU2118NS); Lake Broadwater Park, 27°21'S, 151°05’E, 1983, B. Ballingall 981; Worlds End Pocket, 27°37'S, 152°55'E, on Acalypha capillipes, 1983, P. Forster (BRIU3709NS); Toowoomba, Cooby Ck Rd, 27°23’'S, 151°38’E, on Citriobatus pauciflora, 1983, GNS (BRIU3685NS); Pittsworth, 27°45'S, 151°38’E, on Carissa ovata, 1983, M. Olsen (BRIU3762NS); Coochiemudlo I, 27°35'S, 153°20'E, on Rhizophora stylosa, 1981, GNS (BRIU3623NS); Coomera I, 27°55'S, 153°25'E, on Rhizophora stylosa, 1979, GNS (BRIU3649NS); Tamborine Mt, 27°55’S, 153°12’E, on rain-forest tree, 1979, GNS (BRIU3010NS); Mt Cordeaux, 28°03’S, 152°23’E, on rain-forest tree, 1979, GNS (BRIU3029NS); Springbrook, 28°15’S, 153°16’E, on rain-forest tree, 1982, RWR (BRIU2567RR); O’Reillys, 28°15’S, 153°08’E, on rain-forest tree, 1983,°GNS (BRIU3718NS); The Head, 28°18’S, 152°27’E, on rain-forest tree, 1980, GNS (BRIU3621NS); Kelvin Falls, 28°23’S, 152°06’E, on Casuarina sp., 1980, NCS (BRIU3286A.NS). New South Wales: Wiangaree, 28°25’S, 153°08’E, on rainforest tree, 1978, GNS (BRIU3028NS); Undercliff Falls, 28°38’S, 152°10’E, on Acacia sp., 1980, GNS (BRIU3651NS); Cherry Tree State Forest, 28°54’S, 153°01’E, on vine, 1983, B. Beallingall (BRIU3646NS); Clouds Ck, State Forest, 30°10’S, 152°35’E, on shrub, 1978, HS (CBG9705565); Mt Dorrigo, 30°20’S, 152°40’E, on rain-forest tree, JAE (JAE 3447). RAMALINA IN AUSTRALIA 191 subsp. australis N. Stevens, subsp. nov. Plate 10, fig. 7. Thallus corticolus et saxicolus, virido-griseus parvus caespitosus rigidus erectus, usque ad 1-0 cm altus, ramificatione nulla vel sparsa, ramis e basi exorientibus; rami lati usque ad 10 mm, valde cavi; superficies impolita, laevigata et tenui vel rugosa et crassi; soralia nulla. Apothecia numerosa, terminalia, non calcarata, primum immersa, demum plana; discus usque ad 10 mm in diametro; sporae late ellipsoideae, curvatae aut rectae, (10—)13-14 x 4-5(—6) wm. Acidum divaricaticum + sekikaicum aut sekikaicum et usnicum continens. Holotype: Australia, Western Australia, Moora-Perth Road on Acacia sp. 6 August 1984, G. N. Stevens 4315 (MEL 1048089!). Thallus corticolous and saxicolous, greenish-grey, small caespitose, rigid, erect, usually to 1-0 cm high, branching sparse, at the base only; branches broad, up to 10 mm wide, highly inflated, perforations large, round to elongate and often splitting apart revealing a sparse medulla and internal chondroid strands forming a reticulate fibrose surface on the inner cortex wall; surface matt, smooth and thin to rugose and extremely coarse, rarely pseudocyphellate; holdfast delimited; soralia absent, although some rock specimens with eroded apices appeared pseudo- sorediate. Apothecia numerous, terminal, not spurred, sometimes laminal; disc up to 5-0 mm, excep- tionally to 10 mm diam., concave initially, remaining so or becoming plane; margin entire, inrolled; spores broadly ellipsoid, mainly curved, some straight, (10—)12—14 x 4—5(—6) wm. Chemistry. Divaricatic acid + sekikaic acid or sekikaic acid only and usnic acid. Remarks. Two acid races occur in R. inflata subsp. australis, (1) a divaricatic acid race and (2) a sekikaic acid race. The only name which has been applied to this taxon in the past is R. pusilla, a northern hemisphere Ramalina which contains sekikaic acid with salazinic acid in small amounts (Krog & Swinscow, 1974). Krog & @sthagen (1980) reported that one specimen of R. pusilla from Hierro, Canary Islands, contained salazinic acid only, which indicates that some relationship could exist between the taxa which produce sekikaic acid and R. fissa, which produces salazinic acid only. Partly because of this, R. fissa is placed in the R. inflata complex. Morphologically the sekikaic acid thalli of R. inflata subsp. australis seem closer to the salazinic acid R. fissa thalli than to the divaricatic acid thalli. However, the presence of divaricatic acid and sekikaic acid in the same thallus shows the close connection between the two depside races, thus preventing the sekikaic acid race being separated into a separate taxon from the divaricatic acid race. See Table 11 for a comparison of the three subspecies. Distribution and habitat. Ramalina inflata subsp. australis occurs in the temperate region of mainland Australia and in the drier north and east coastal areas of cool-temperate Tasmania (Figs 21, 22). The two chemical races which occur in this taxon are sympatric over most of their range, although only the divaricatic acid race occurs coastally. No sekikaic acid thalli were found in Tasmania. Within its range R. inflata subsp. australis exhibits a variety of ecophenotypes. In the drier inland regions of South Australia extremely small thalli are found bearing large terminal apothecia which obscure the branches (Plate 10, fig. 7). Both divaricatic acid and sekikaic acid races occur in such areas, with a predominance of the latter race. In some habitats this taxon produces a single inflated branch resembling a small round balloon. The divaricatic acid race also occurs on rock substrates, e.g. sandstone (Mt Arapiles and the Grampians) and on trachyte (Hanging Rock). Although the morphology of the saxicolous thalli differs from the corticolous thalli, the differences are regarded as the result of change in substrate rather than genotypic, e.g. the thalli are much reduced in size, occurring as small button-like growth forms in colonies; they produce prominent apothecia between sterile branches. This rock morphotype resembles the northern hemisphere saxicolous taxon R. mediterranea Magnusson, but the Mediterranean species contains evernic acid. It could be argued that this change in substrate and morphology warrants the erection of a new species for this rock morphotype. However, it is regarded as an ecophenotype of R. inflata subsp. australis. G. NELL STEVENS 192 YOOI IO y1eq yieq yieq 9}e1}sqng ATUO plore oIeyIYas 10 Ajuo plore s1eyryos ploe o1eyryas + 10 sploe o1eyLYyos 10 IO ploe o1eyryos ploe oeoLIeAIpIoU IBILIVAIPIOU I9Y}I9 F plow oeoeaip F plow onevoreaip F plow oeoeAaip Anstey wir! (9-)S—p x win! ¢—p x 9T-Z1(-OT) wi (9-)¢S-p x 9I-ZI (91-)rI-01(-8) azis o10dg poamns A[su01}s poains 0} JY43IeNs PAIN IO JYSsIeIYS Ajeres ‘yys1e1}s odeys o10ds peT[orut pouygep pue pouyep [Jom ][2M 0} JOUT}SIpUI pouyop [Jom UISIVU JO WI04 pozinds jou poiinds souijowios posinds OU IO pozinds oueyd Ajores SARDUOD oueyd 0} savou09 X9AU09 Ayorel ‘oueyd elooyjode jo odeys “WeIp WU (QT—)0-S ‘weIp Wu ¢—Z U90M}0q ‘wWeIp WU 9-7 0} dn B1ooyjode Jo 3ZIS JeUIWIO} eloyjode [eUIWIe] pue [eUTWIO} yeurws} Ajurew ‘jeuruo}qns ATUTeW jo uOnIsog JopuljAd plorpuoyo Ayoyed pue jo ]]en Ayoyed pue usaom UdAO0M ATASOO] 0} 0} Sulloype ‘asieds AJasoo] ‘asuop Ajore1 snonuljuos pue ssuop oeydAy Areyjnpoyy uodo yds 9}e3u0]9 OYT]-3]S Ayyensn ‘[eao ‘asie] ‘punol ‘o31R] Jo ‘punol ‘]jeus SUONPIOJIOg poyeyur Ay[ny uonepur UO}eYUr OWI9I}x9 poyegur Any 0} possoidwios jsowlye jo 90139q suo] suo] snqyey} SuO] Wd ()-| wd (0-p-)0-€-0-Z wid (0-€-)0-7-S-T jo ozis [ens() sypajsno *dsqns pjoyfu1 ‘dsqns pyisndiad ‘dsqns Joyoeleyy *xo]dui09 pour DuYUUDY BY} Ul satdadsqns 9aIY} dy} Jo oUdIOJoId ajeIWsqns puUe IoJOBIeYO [eoIWIDYO pue [eoIsojoydiow jo uosuedwoD I] qe] RAMALINA IN AUSTRALIA 193 Ramalina inflata subsp. australis is mainly corticolous and grows on various tree species. Exotic phorophytes are Pinus, Populus, and Malus; native phorophytes include Acacia, Casuarina, Leucopogon, and Melaleuca spp. World distribution: Australia. New South Wales: Batemans Bay, 35°42’S, 150°10’E, on Avicennia marina, 1978, GNS (BRIU2172NS); Murunna Pt, 36°23’S, 150°03’E, on coastal shrub, 1978, JAE (JAE 4592). Victoria: Marlo, 37°48’S, 148°32'E, on Melaleuca sp., 1978, GNS (BRIU3201NS); Wilsons Prom., N of, 38°45’S, 146°08’E, on Acacia sp., 1980, GNS (BRIU3218NS); Mt Macedon, Hanging Rock, 37°23'S, 144°36’E, on trachyte, 1976, GNS (BRIU3128NS); Spargo Ck, Korwerigaboora Springs, 37°30’S, 144°00’E, 1965, M. Allender (MEL1026135); Triodea Hill, 37°57’S, 141°19’E, on Banksia sp., 1965, RF (MEL1026193); Drik Drik, 38°00’S, 141°18’E, 1952, A. C. Beauglehole (MEL544971); Rocket Lake, 34°39’S, 141°48’E, 1970, ED (0); Big Desert, 35°22’S, 141°13’E, 1970, ED (0); The Grampians, 37°08’S, 142°26’E, on rock, RF (MEL1026047); Dimboola Nat. Park, 36°20’S, 142°01'E, 1949, A. C. Beauglehole (MEL1020493). South Australia: Oodlawirra, 5 km W of, 32°53’S, 139°04’E, on Muellenbeckia sp., 1969, RWR (AD97649267); Yunta, 40 km S of, 32°59’S, 138°50’E, 1971, N. N. Donner (AD97407411); Blanchetown, 13 km W of, 34°21'S, 139°29’E, 1973, J. Z. Weber (AD97647391); Naracoorte, 25 km SE of, 36°58’S, 140°45’E, 1969, J. Z. Weber (AD97417116); York Peninsula, Treasure Cove, 35°00'S, 137°25’E, on dead shrub, 1979, JAE (JAE 6367); Eyre Peninsula, Streaky Bay, 40 km SE of, 32°48’S, 134°13’E, 1967, N. N. Donner (AD97648422); Colona Homestead, 31°38’S, 132°04’E, 1947, J. H. Willis (MEL1026176); Yalata Road House, 5 km W of, 31°39’S, 129°03’E, on bushes, 1980, AA (BRIU4345NS). Western Australia: Three Springs, 5 km from, 29°32'S, 115°46’E, 1969, on edge of salt lake, R. A. Saffrey (PER000427); Winchester, 8 km W of, 29°46’S, 115°56’E, on tree, 1966, P. G. Wilson (PER000426); Toodyay-Bindoon Rd, 31°25’S, 116°15’E, on Acacia sp., 1984, GNS (BRIU4316NS); Pt Mt Henry, Perth, 32°02’S, 116°05'E, on Dryanda sessilis, 1970, N. Sammy; Beverley, 17 km S of, 32°10’S, 116°56’E, on Acacia sp., 1984, GNS (BRIU4319NS); Hyden, The Humps, 32°19’S, 118°57’E, on granite rock, 1972, N. Sammy; Kondinin Forestry Reserve, 32°30’S, 118°24’E, on bushes, 1967, GB 67/366; Cranbrook 17 km N of, 34°15’S, 117°37’E, on Casuarina sp., 1984, GNS (BRIU4314NS); Lake Muir, E side, 34°29’S, 116°40’E, 1980, E. A. Griffin (PER001630); Albany-Borden off Chillinup Rd, 34°20’S, 118°12’E, on Melaleuca sp., 1970, N. Sammy; Lake Biddy, near Lake Grace, 33°00’S, 118°56’E, on tree, 1976, A. C. Smith (N. Sammy); Gairdner R., 34°14’S, 119°17’E, on Hakea sp., 1980, D. Richardson (PER000941); Cape Arid Nat. Park, 34°02’S, 123°09’E, on Hakea clovata, 1971, R. D. Royce (PER000428). Tasmania: Rocky Cape, 40°53’S, 145°30’E, on quartzite rock, 1980, GK 743/80; Ansons Bay Nth, 41°02’S, 148°16’E, on Banksia sp., 1975, GB (H032680); Moores Hill, Beaconsfield, 41°13’S, 146°45’E, on Acacia dealbata, 1980, GK 216/80; Buckland, Pitts Hill, 42°36’S, 147°43’E, on Bursaris spinosa, 1980, GK 450/80; Hobart, Gunners Quoin, 42°54’S, 147°19’E, on dolerite rock, 1980, GK 600/80. 16. Ramalina fissa (Mill. Arg.) Vainio Plate 10, fig. 8. in Mem. Herb. Boissier 5: 2 (1900). Ramalina inflata var. fissa Mill. Arg. in Flora, Jena 71: 203 (1888). Type: Australia, King Island, Bass Strait, 1888, French s.n. (G! — holotype; salazinic acid and usnic acid). Plate 14, fig. 4. Ramalina brevis var. brevissima F. Wilson in Pap. Proc. R. Soc. Tasm. 1892: 159 (1893). Type: Australia, Tasmania, Ulverstone, Antill Ponds, Weymouth s.n. (not located). — Ramalina fraxinea var. brevissima (F. Wilson) Zahlbr., Cat. Lich. Univ. 6: 486 (1930). Thallus corticolous and saxicolous, greenish-grey, erect, rigid to flaccid, inflated and shrubby or flattened and almost foliose, 1-3 cm high; branching from the base only; branches variable in width, often bursting open to reveal the medulla, giving a false effect of gross inflation, with flattening of the thallus the areas directly below the apothecia remain the only part of the thallus with a central cavity; perforations when present, large, round to elongate, splitting to reveal the medulla; medulla loosely woven, sparse, adhering in patches to the inner cortex wall, the wall often marked with reticulate strands of chondroid tissue; sometimes pseudocyphellate; holdfast delimited; soralia absent. Apothecia numerous, terminal, not spurred; sometimes laminal; disc 2—5(—8) mm diam., initially concave becoming flat; margin entire, inrolled; spores ellipsoid, mainly curved, some straight, (10—)12—16 x 4—5(-6) um. Chemistry. Salazinic acid and usnic acid. 194 G. NELL STEVENS Remarks. Ramalina brevis var. brevissima F. Wilson has been placed in the synonymy. Although the type material was not located (it is possibly lost), other specimens identified by Wilson as R. brevis var. brevissima contain salazinic acid and are small caespitose morphotypes of R. fissa. Wilson identified large coarse morphotypes of this taxon as R. fraxinea f. platyna Ny]. Material held at G under the latter name consists of two thalli which contain divaricatic acid and one which contains salazinic acid. Magnusson examined these specimens and found different colour reactions, K— (divaricatic acid), and K + y (salazinic acid). These specimens have been found to belong to R. inflata subsp. australis and R. fissa. Ramalina subpusilla comes near to R. fissa because of its similar chemistry, but morphologi- cally it differs by producing spurred apothecia, a characteristic not found in the latter taxon. Ramalina calcarata Krog & Swinscow contains divaricatic acid and small amounts of salazinic acid in the hymenium; Krog & Swinscow (1974) compared it to R. fissa because of the reticulate-fibrose underside of both and the straight to reniform spores produced by both, but the absence of spurred apothecia in the latter taxon separates the two. Galloway (1985) recorded that R. geniculata (a New Zealand species often confused with R. inflata) contains sekikaic and salazinic (tr.) acids; but the type material (held at FH) contains divaricatic acid. Such variation in acid content of a single species adds weight to the suggestion that in Australia there is an evolutionary relationship between the depside taxon R. inflata subsp. australis and the depsi- done taxon R. fissa, especially as similarities in morphology are also found in some specimens. However, as no Australian material was found to contain both depside and depsidone chemistry, R. fissa is regarded as a distinct, endemic species. Distribution and habitat. The distribution range of R. fissa (Fig. 23) lies within the larger range of the two chemical races of R. inflata subsp. australis in the temperate region of Australia. It is mostly coastal in occurrence, occupying sections of the coastline in both Tasmania and the mainland not usually occupied by R. inflata subsp. australis (divaricatic acid race). In western Victoria and South Australia the inland locations of R. fissa are mainly different from the sites occupied by R. inflata subsp. australis (Figs 22, 23). Variation in morphology appears to be ecophenotypic; in some coastal habitats extreme flattening of the thallus occurs, producing a dorsi-ventral form resembling a parmeloid thallus. In the dry inland regions of southern Australia the thalli are inflated, small, and very coarsely textured. On the islands of Bass Strait and Kangaroo Island, the thalli produce inflated pustules or small branchlets on the surface. A lower ‘pseudo-cortex’ is often produced to cover the exposed medulla in some thalli with completely split-open branches. Occasionally R. fissa occurs on rocks and produces a much reduced growth form. On Hunter Island, Bass Strait, specimens grow on quartzite; Tasmanian collections from Cape Deslacs and Bruny Island do not record the rock type. Some of these saxicolous specimens are tiny, single, balloon-shaped thalli which closely resemble thalli of R. inflata subsp. australis (divaricatic acid race) collected off fine twigs of Leptospermum sp. in Western Australia. Usually R. fissa is corticolous, occurring on a number of different tree species (Acacia spp., Leucopogon par- viflora, Leptospermum laevigatum, Pimelea serpyllifolia, and Melaleuca lanceolata); it occurs on the mangrove Avicennia marina and has been found growing on fence posts at Marlo, Victoria. It does not show any substrate preference. World distribution. Australia. Victoria: Linga, 13 km N of, Pink Lakes, 35°06’S, 141°40’E, on dead wood, 1970, GB & JC (H032663); Casterton, 51-5 km N of, 37°20’S, 141°20’E, on Acacia sp., 1973, RF (MEL1036009); Creswick, 37°26'S, 143°54’E, on Pinus sp., 1963, J. H. Willis (MEL9335); Lake Burrumbeet, 37°28’S, 143°40'E, on Pinus sp., 1945, P. Bibby (MEL10018); Marlo, 37°48’S, 148°42’E, on fence post, 1978, JAE (JAES195); Troidea Hill, 37°57’S, 141°19’E, on Banksia sp., 1965, RF (MEL1026193); Portsea, 38°20’S, 144°42’E, on shrub, 1964, J. Williams (MEL1516513); Portland, 38°21’S, 141°36’E, on Malus sp., 1965, RF (MEL1516516); Western Port Bay, 38°26’S, 145°08’E, on Avicennia marina, 1978, GNS (BRIU2347NS); Wilsons Prom. N of, 39°05’S, 146°25’E, on Acacia sp., 1980, GNS (BRIU3198NS). Bass Strait: Hogans I, 39°13’S, 146°59’E, on Leucopogon sp., 1973, JW (MEL10130013); Little Dog i, RAMALINA IN AUSTRALIA 195 130 130 140 150 Ss? 104 0. 800km | 2 i = i i eee CB is 204 4) ’ 304 t) e@ (§ ca we @ Se ce ee ee @@ ee 404 1 1 ! ‘ Fig. 23 Distribution of R. fissa. 40°15'S, 148°17’E, on twigs, 1977, JW (MEL1026160); Flinders I, 40°01’S, 148°02’E, on Leptospermum laevigatum, 1969, JW (MEL1019938); Hunter I, 40°31’S, 144°45’E, on quartzite, 1973, T. V. Muir (MEL1021245). Tasmania: Table Cape, 40°57’S, 145°43'E, on Acacia melanoxylon, 1970, GB (HO32296); Eddystone Beach, 40°59’S, 148°19’E, on dead tree, 1973, GB (H032679); Devonport, 41°11'S, 146°21’E, on Acacia sp., 1980, GNS (BRIU3240NS); Chain of Lagoons R., 41°39’S, 148°18’E, on bush, 1973, GB & JC (H032667); Cape Deslacs, 42°57’S, 147°33’E, on rock, 1965, GB & JC (H032647); Bruny I, Isthmus Bay, 43°23’S, 147°17'E, on tree, 1982, AA (BRIU3707NS). South Australia: Lock, 25 km WSW of, 33°40’S, 135°20’E, 1967, N. N. Donner (AD97648157); Yararoo Gorge, 34°20’S, 138°20’E, 1967, B. Copley (AD97419099); Kangaroo I, 35°40’S, 137°38’E, on Mallee bush, 1971, G. Jackson (AD97410372); Coorong, 36°12’S, 139°40’E, 1965, A. C. Beauglehole (MEL1020684); Naracoorte, 36°58’S, 140°45’E, on dead wood, 1973, M. Beck (H32559); Eyre Peninsula, Cape Catastrophe, 34°40’S, 135°50’E, on Melaleuca sp., 1970, RF (MEL1020693). 17. Ramalina nervulosa (Mill. Arg.) des Abb. in Bull. Inst. fr. Afr. noire 14: 25 (1952). Ramalina farinacea var. nervulosa Mill. Arg. in Flora, Jena 66: 21 (1883). Type: Australia, Queensland, Daintree River, Pentzke s.n. (G! —lectotype; divaricatic acid, stenosporic acid, and usnic acid). var. nervulosa Plate 11, fig. 3; Plate 14, fig. 7. Thallus corticolous, pale green to stramineous, subpendulous to pendulous, usually 4-10 cm long, exceptionally to 30 cm; branching sparse with unequal dichotomy producing long attenuate branches, often tangled about one another; branch width (0-3—)1-0—1-5(—2-0) mm, 196 G. NELL STEVENS HEE ETTTTITITITTITTTTTITINITITTT LLLP STIPE II ISTP Tet iiitiiiiiiiiien RAMALINA IN AUSTRALIA 197 branches flat and narrow, when extremely narrow becoming subterete and twisted or undulate along the entire length, branches occasionally fuse; cortex matt, marked with longitudinal pseudocyphellae which cause the ‘nervulose’ appearance of this species; several branches arise from a well defined holdfast; soralia marginal, round to ellipsoid. Apothecia not seen. Chemistry. Divaricatic acid and stenosporic acid and usnic acid. Remarks. Miller [Arg.] (1883) applied the name R. farinacea var. nervulosa to specimens which had been sent to him from Norfolk Island, Australia, Tahiti, the Marianas, and India which closely resembled R. farinacea. Des Abbayes (1952) collected and examined sorediate specimens from the Ivory Coast (West Africa) and thought they belonged to the same species that Miiller had described, but he believed that they were sufficiently distinct from R. farinacea to warrant species rank as R. nervulosa (Mill. Arg.) des Abb. (Stevens, 1983a). When the five specimens held at G in the Miller [Arg.] herbarium were examined, it was found that, although their morphology was similar in some instances, each had a quite different chemistry and belonged to separate taxa. The majority of the thalli bear the ‘nervulose markings’ (linear pseudocyphellae and longitudinal cracks) which caused the name ‘nervulosa’ to be applied to them; this trait is quite distinctive as remarked upon by des Abbayes. Of the five specimens cited by Miller (1883) only the Australian material from Daintree River Queensland contains divaricatic acid and stenosporic acid. This specimen was designated as lectotype by Filson in 1971. Distribution and habitat. Ramalina nervulosa var. nervulosa occurs along the Queensland coastline from latitude 14°—28°S (Fig. 24), and at several inland locations (Stewart River (14°S), Barrow’s Range (15°S), and Atherton Tableland (17°S)). This taxon is sympatric with its variety (var. luciae) for much of its distribution in Australia. The largest specimens collected, which exceed 20 cm, grew at Stewart River (14°S) and Turkey (24°S), the reason for such exceptional growth is not known. The mangroves Ceriops tagal and Rhizophora stylosa are the usual phorophytes for this taxon. World distribution: Burma/Bangladesh, Indonesia (Java and Bali), South Africa, and Austra- lia. Queensland: Stewart R., 14°05’S, 143°38’E, 1900, Johnston (MEL 4443); Barrons Range, 15°36’S, 145°04’E, on shrub, 1983, M. Godwin (C2436); Mossman Rifle Range, 16°20’S, 145°45’E, on Ceriops tagal, 1983, GNS (BRIU4128NS); Trinity Inlet (Cairns); 16°56’S, 145°46’E, on Ceriops tagal, 1976, D. Tarte (BRIU2589NS); Hinchinbrook I, 18°17'S, 146°13’E, on Ceriops tagal, 1979, GNS (BRIU3131NS); Eimeo, 21°09’S, 149°09’E, on Rhizophora sp., 1976, GNS (BRIU3544NS); St Lawrence, 22°21’S, 149°32'E, on Excoecaria agallocha, 1976, NCS (BRIU2131NS); Port Alma, 23°35’S, 150°51'E, on Rhizophora stylosa, 1975, RWR (BRIU698RR); Gladstone, 23°50’S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3348NS); Rodds Peninsula, 24°02’S, 151°40’E, on Ceriops tagal, 1975, GNS (BRIU1292NS); South Stradbroke I, 27°55’S, 153°30’E, on Ceriops tagal, 1975, H. T. Clifford (BRIU76RR). var. dumeticola (Krog & Swinscow) N. Stevens Plate 11 ,fig. 2: in Lichenologist 15: 228 (1983). Ramalina dumeticola Krog & Swinscow in Norw. J. Bot. 23: 163 (1976). Type: Kenya, Coast Province, Kwale District, Krog 3K30/127 (O! — holotype, BM! — isotype; homosekikaic acid (M), sekikaic acid (M), 4’-0-methylnorhomosekikaic acid (M), ramalinolic (t), 4’-0-demethylsekikaic acid (t), 4’-0- methylnorsekikaic acid (t) and usnic acid). Thallus corticolous, pale green to yellow-green, tufty, subpendulous, 4—6 cm long; branching dichotomous, moderate to dense but never intricately branched; branch width (0-2)0-5—1-0(— < Plate 11. Thallus growth forms in the Australian Ramalinae. Scale in mm. Fig. 1 R. pacifica. Fig. 2 R. nervulosa var. dumeticola. Fig. 3 R. nervulosa var. nervulosa. Fig. 4 R. nervulosa var. luciae. Fig. 5 R. subfraxinea var. confirmata. Fig. 6 R. subfraxinea var. leiodea. Fig. 7 R. subfraxinea var. subfraxinea (narrow form). Fig. 8 R. subfraxinea var. norstictica (type specimen). 198 G. NELL STEVENS 120 130 140 150 5 104 , SR ; _, 800km WS, as Ee Nas ak a a a \& _ 30 oH L LS 404 1 Ll 1 1 Fig. 24 Distribution of R. nervulosa var. nervulosa. 2-0) mm, branches compressed, flat, usually canaliculate for part of their length, some branches subterete distally, apices attenuate, forked or broken off and appearing blunt; surface matt or shiny, smooth, faint pseudocyphellae sometimes present; holdfast delimited; soralia marginal, round, ellipsoid or elongate. Apothecia not seen. Chemistry. Homosekikaic acid (M), sekikaic acid (M), 4’-0-methylnorhomosekikaic acid (m), 4’-0-demethylsekikaic acid (t), ramalinolic acid (t), 4’-0-methylnorsekikaic acid (t) [the last two acids are rarely detected with tlc] and usnic acid. Remarks. Initially the Australian material of this taxon was thought to be a morphotype of R. peruviana as the two are sympatric in occurrence. However, the shape of the soralia and branching pattern of the two sets them apart. The morphology of R. nervulosa var. dumeticola closely resembles that of R. farinacea. Although both R. peruviana and R. nervulosa var. dumeticola contain acids in the sekikaic aggregate, the presence of 4’-0-methylnorhomosekikaic acid as a minor metabolite in the latter taxon and not in the former, and the constant occurrence of ramalinolic acid in R. peruviana but not in R. nervulosa var. dumeticola distinguishes the two on chemical grounds. Distribution and habitat. The occurrence of R. nervulosa var. dumeticola in Australia at subtropical latitudes appears to be an anomaly (Fig. 25). Previous research (Stevens, 1983a) has shown its distribution to be mostly equatorial and northern hemisphere tropical (to latitude 15°N), penetrating to latitude 8°S (based on known collections). Its presence in the mountains of RAMALINA IN AUSTRALIA 199 t tT Li Li 420 130 140 160 . S? 1O- | a 1 1 ie i: _890km CQ e@ @ é 20 WN * 0 ro * f 304 oe (oe) $ ie 40~ 1 4 ue L Fig. 25 Distribution of R. nervulosa var. luciae @ and var. dumeticola *. New Guinea at latitude 6°S (established during this present research) is in accord with this distribution pattern. The subtropical occurrence of this taxon in Australia in the mountain areas between latitudes 26°S—30°S cannot be explained, nor can the complete absence of this taxon from any coastal habitats. World distribution. India, Indonesia (Java), Philippines, Marianas Is, Kenya, Ivory Coast, Principe, Ascension, and Australia. Queensland: Bunya Mts, 26°50’S, 151°40’E, on understorey shrub, 1979, GNS (BRIU2971NS); Mt Castle, 2 km SW of, 27°58’S, 152°23’E, on rainforest twig, 1983, NCS (BRIU3677NS); Cunninghams Gap, Mt Cordeaux track, 28°03’S, 152°23’E, on rain-forest twig, 1979, GNS (BRIU3697NS); The Head, Moss Gardens, 28°15'S, 152°30’E, on twigs of rain-forest shrub, 1981, GNS (BRIU3622NS); O’Reillys, Botanical Walk, 28°15’S, 153°08’E, on rain-forest shrub, 1983, GNS (BRIU3731NS); Gamboobal State Forest, 28°14’S, 152°20’E, on dead tree on mountain top, 1983, GNS (BRIU3737NS); Queen Marys Falls, 5 km N of, 28°20’S, 152°22’E, on rain-forest tree, 1983, GNS (BRIU3736NS). New South Wales: Chaelundi Mt, 37 km N of Ebor, 30°04’S, 152°21'E, Acacia orrites, 1978, D. Verdon (CBG7809394); Cattle Ck State Forest, 12 km NNE of Dorrigo, 30°15’S, 152°03’E, on Banksia integrifolia, 1978, D. Verdon (CBG8203532). var. luciae (Molho et al.) N. Stevens Plate 11, fig. 4. in Lichenologist 15: 227 (1983). Ramalina luciae Molho, Bodo, Culb. & C. Culb. in Bryologist 84: 396 (1981). Type: Fiji, Kambara (= 200 G. NELL STEVENS Kabara) Island, Salvat B. 358a (DUKE — holotype; ‘sekikaic, homosekikaic, 4’-0-demethylsekikaic and 4’-0-methylnorsekikaic acids and usnic acid’). Thallus corticolous, pale green to stramineous, subpendulous to pendulous, up to 7 cm long, exceptionally to 18 cm; branching sparsely dichotomous, producing extreme elongation between dichotomies; branch width (0-3)0-5—1-0(—2:0) mm, branches flattened at the base but subterete to terete towards the apices, slightly twisted along the whole length, fusion between branches common, causing a tangled growth form; cortex matt to shiny; longitudinal pseudocyphellae usually present but rarely becoming laminal soralia; simple cracks in the cortex and splitting along the margin occurs; several basal branches arise from a well defined holdfast. Apothecia rare, lateral, disc 2 mm diam, concave to ellipsoid to fusiform, sometimes appearing 3,4-celled; straight or rarely slightly curved (10—)12—16(—18) x 4—5(-—6) um. Chemistry. Sekikaic acid (M), 4’-0-methylnorsekikaic acid (m), 4’-0-demethylsekikaic acid (m or t), homosekikaic acid (t), ramalinolic acid (t), 4’-O-methylnorhomosekikaic acid (t). Remarks. Ramalina nervulosa var. luciae was described by Molho, Bodo, Culberson & C. Culberson (1981) from a specimen collected in Fiji. Similar material held at G as a syntype of R. farinacea var. nervulosa Mill. Arg., was collected from Tahiti last century by Jardin. This taxon is acommon maritime Ramalina along the eastern coastline of tropical Australia. Because of the similarity in morphology between R. nervulosa var. nervulosa and R. nervulosa var. luciae, chemical differences had to be used as the main criteria to separate them. The rank of variety was given to each taxon in accordance with the guidelines set out by Hawksworth (1976). Apart from chemical difference, their distribution range when considered on a global scale shows two distinct patterns (Stevens, 1983a). The chemistry of two sorediate Ramalina species resembles that of R. nervulosa var. luciae (viz. R. peruviana Ach. and R. intermediella Vainio) but both of these have a distinct morphology which distinguishes them from the morphology of R. nervulosa var. luciae. Distribution and habitat. The distribution of R. nervulosa var. luciae (Fig. 25) covers a coastal region from approximately latitudes 15°—27°S in Queensland. For most of its range this taxon is sympatric with both R. nervulosa var. nervulosa and R. pacifica. The majority of the collections was made at sea-level from mangrove communities, the main phorophytes being Ceriops tagal, and to a lesser extent Avicennia marina and Rhizophora stylosa which grow on the sheltered, landward fringe of the mangroves. The most northerly collections in Australia are from inland sites, e.g. Butchers Hill, Lakeland, latitude 15°52’S, and Atherton Tableland, latitude 17°16’S. Both of these sites are approximately 45 km from the coast at an altitude of 900 m, indicating cooler habitat requirements at lower latitudes; a tendency observed in the other members of the tropical R. farinacea complex (Stevens, 1983a). At latitude 21°S (Eimeo Beach, Mackay) this species grows prolifically on the landward fringe of Rhizophora stylosa communities; in this habitat fertile specimens as well as very long specimens occur. World distribution. Sri Lanka, Indonesia (Celebes), Kenya, Australia, New Hebrides, Fiji, Cook Is, and Tahiti. Queensland: Lakeland, Butchers Hill, 15°52’S, 144°55’E, on rain-forest shrub, 1983, GNS (BRIU4108NS); Mossman, 16°20’S, 145°24’E, on Pinus sp. in garden, 1983, C. McCracken (BRIU4241NS); Yungaburra Rd Atherton, 17°16’S, 145°29’E, on Casuarina sp., 1983, HS (CBG8302450); Hinchinbrook I, 18°17’S, 146°13’E, on Ceriops tagal, 1979, GNS (BRIU3137NS); Eimeo, 21°09'S, 149°09’E, on Rhizophora stylosa, 1976, GNS (BRIU3542NS); St Lawrence, 22°21’S, 149°32’E, on Ceriops tagal, 1980, GNS (BRIU3600NS); Corio Bay, 22°58’S, 150°46’E, on Ceriops tagal, 1974, D. Tarte (BRIU1168NS); Gladstone, 23°50’S, 151°16’E, on Ceriops tagal, 1980, GNS (BRIU3566NS); Tannum Sands, 23°57’S, 151°16’E, on Ceriops tagal, 1975, RWR (BRIU833RR); Turkey, 24°06’S, 151°40’E, on Rhizophora stylosa, 1975, GNS (BRIU280NS); Toogoom, 25°15’S, 152°40’E, on Ceriops tagal, 1983, GNS (BRIU4149NS); Serpentine Ck, 27°23’S, 151°39’E, on Ceriops tagal, 1975, GNS (BRIU1044NS). RAMALINA IN AUSTRALIA 201 18. Ramalina pacifica Asah. Plate 11, fig. 1; Plate 14, fig. 8. in J. Jap. Bot. 15: 213 (1939). Type: Micronesia, Marianas, Saipan Island, 1925, Kimura s.n. (TNS! - lectotype; salazinic acid and usnic acid). Ramalina insularum Magnusson in Ark Bot. 3 (10): 361 (1956). Type: Fiji Is., 1862, Daemel s.n. (H! - lectotype; salazinic acid and usnic acid). Thallus corticolous, rarely saxicolous, grey-green, tufty, subpendulous to pendulous, 4-8 cm long, exceptionally to 32 cm; branching dichotomous, sparse to moderately dense but extremely variable; branch width (0-5—)1—4 mm, branches compressed, flat, some splitting along the margins; broad in the basal region but gradually narrowing towards the apices or narrow in the basal region and continuing the same width to the apices; apices attenuate, mostly forked; cortex matt or shiny, smooth or with linear pseudocyphellae slightly or strongly developed; holdfast delimited; soralia marginal and laminal, round to ellipsoid (Plate 1, fig. 3), often coalescing, the edges revolute. Apothecia rare, marginal, disc 1-3 mm diam., concave, plane to convex, with edges sometimes incised at maturity; spores fusiform or ellipsoid, (Plate 4, fig. 3), straight or very slightly curved, 12—16(—20) x 4-6 um. Chemistry. Salazinic acid + protocetraric acid and usnic acid. Remarks. The name Ramalina pacifica was given to sorediate material found on the Saipan Islands (Marianas) and on Hachijo Island (Japan) by Asahina (1939); he stated that lichenol- ogists had identified his species as various varieties of R. farinacea. Asahina also commented that some thalli of R. pacifica contained both salazinic and sekikaic acids. It appears that the sekikaic acid he observed could have come from a contaminant, perhaps R. intermediella or R. nervulosa var. dumeticola, as both of these species are maritime and grow in association with R. pacifica. Stevens (1983a) placed R. pacifica in the tropical-subtropical group of the R. farinacea complex. It is the only taxon in this group which has a depsidone chemistry. In ancient times it may have been contiguous with the protocetraric acid species of R. farinacea found on Macquarie I. (latitude 55°S), but the links are now broken except for the presence of trace amounts of protocetraric acid in some specimens of R. pacifica. Ramalina insular'um Magnusson, a synonym of R. pacifica, was collected from Fiji. Its morphology more closely resembles that of R. nervulosa var. luciae material collected from Fiji than it does the Australian R. pacifica, which illustrates the phenomenon of convergent morphology of these two Fijian specimens which are sympatric. Distribution and habitat. The distribution range of R. pacifica covers an area which is mainly subtropical to warm temperate (Fig. 26), although several collections have been made in tropical Australia, e.g. Hinchinbrook I. (latitude 18°S) and Shute Harbour (latitude 20°S). In these areas the rainfall exceeds 2000 mm p.a. and the thalli are very narrow and delicate with smooth, shiny surfaces, without pseudocyphellae. Collections from the drier latitudes along the Queensland coast (latitudes 22°-24°S which receive 800-1100 mm p.a.) comprise thalli which produce numerous pseudocyphellae along the whole length of the branches; these thalli are generally broader than the thalli from the higher rainfall habitats. Only one inland collection of R. pacifica was made, at Atherton Tableland. The main phorophytes for this taxon are the mangroves Ceriops tagal and Avicennia marina. Ramalina pacifica attains its maximum growth at latitudes 27°-28°S (Moreton Bay), annual rainfall 1100-1500 mm p.a. In two relatively undisturbed well-established communities of Avicennia marina some exceptional specimens 32 cm long with branches 2-3 mm wide were found. Fertile specimens also occurred in this environment. The majority of the thalli produced pseudocyphellae as well as both marginal and laminal soralia. In areas of Moreton Bay where R. pacifica grew on Ceriops tagal, the thalli were small (up to 5 cm long) with narrow branches (to 1-0 mm wide), and had a smooth cortex which lacked pseudocyphellae or produced them at the base of the branches only. From latitude 28°S to the limit of R. pacifica distribution at 35°S, this taxon occurs on 202 G. NELL STEVENS cs rE T 120 130 140 160 i 104 o 800km ST oe ry . ) 1 i e e a e 20 oF 4, ys 30 oe os ie ie 404 L 1 i 1 Fig. 26 Distribution of R. pacifica. Avicennia marina (Ceriops tagal does not extend beyond latitude 28°S). Pseudocyphellae are usually present on the branches which reach 8 cm in length with a maximum width of 2 mm. The annual rainfall in this section of coastline is between 1000-1500 mm p.a. It was not able to be established whether the amount of rainfall or the type of substrate had the greater influence on the variation in morphology of this taxon. Ramalina pacifica is rarely saxicolous but occurs on rock at Norries Head, New South Wales, in association with R. subfraxinea var. confirmata, another taxon which is usually corticolous. Presumably the rocks in this area of coastline are able to produce a microhabitat which is equivalent to that found in tree communities on other parts of the coast, enabling both taxa to become established on the cliff face. World distribution. Burma/Bangladesh, Japan, Indonesia (Java), Marianas Is, South Africa, Australia, New Hebrides, New Zealand, and Cook Is. Queensland: Tinaroo Dam, 17°15’S, 145°36’E, on Plumeria sp., 1983, GNS (BRIU41NS); Hinchinbrook I, 18°17'S, 146°13’E, on Ceriops tagal, 1979, GNS (BRIU3441NS); Bowen, 19°59’S, 148°22’E, on Lumnit- zera racemosa, 1979, GNS (BRIU3284NS); Shute Harbour, 20°16’S, 148°47’E, on Ceriops tagal, 1979, GNS (BRIU3283NS); Brampton I, 20°49’S, 149°18’E, J. M. Gilbert (H032317); Eimeo, 21°09’S, 149°09’E, on Rhizophora stylosa, 1976, GNS (BRIU1369NS); St Lawrence, 22°21'S, 149°32’E, on Excoecaria agallocha, 1976, GNS (BRIU2127NS); Keppel Sands, 23°21’S, 150°47’E, on Excoecaria agallocha, 1975, RWR (BRIU779RR); Port Alma, 23°35’S, 150°51'E, on Rhizophora sp., 1975, RWR (BRIU699RR); Gladstone, 23°50’S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3570NS); Tannum Sands, RAMALINA IN AUSTRALIA 203 23°57'S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3599NS); Turkey, 24°06’S, 151°37’E, on Rhizophora stylosa, 1975, GNS (BRIU1266NS); Rodds Peninsula, 24°02’S, 151°40’E, on Ceriops tagal, 1975, GNS (BRIU1153NS); Round Hill Hd., 24°10’S, 151°54’E, on Rhizophora stylosa, 1975, RWR (BRIU915RR); Fraser I, Wathumba Ck, 24°59’S, 152°55’E, on Rhizophora stylosa, 1975, GNS (BRIU3597NS); Burrum Hds, 25°10’S, 152°35’E, on Rhizophora stylosa, 1975, GNS (BRIU1129NS); Urangan Beach, 25°19’S, 152°55’E, on Callitris columellaris, 1978, GNS (BRIU2841NS); Tin Can Bay, 25°55'S, 153°00’E, GB (H032319); Noosa Hds, 26°24’S, 152°07’E, on Rhizophora stylosa, 1975, GNS (BRIU1807NS); Deception Bay, 27°10’S, 153°02’'E, on Avicennia marina, 1978, GNS (BRIU2561NS); Tingalpa Ck, 27°28’S, 153°11'E, on Avicennia marina 1977, GNS (BRIU1756NS); Coochiemudlo I, 27°35'S, 153°20’E, on Rhizophora stylosa, 1977, GNS (BRIU1654NS); Little Rocky Pt, 27°38’S, 153°18’E, on Ceriops tagal, 1977, GNS (BRIU1438NS); Stradbroke I, Swan Bay, 27°45’S, 153°26’E, on Ceriops tagal, H. T. Clifford (BRIU3075RR). New South Wales: Brunswick Hds, 28°32'S, 153°32'E, on Callitris sp., 1980, GNS (BRIU3486NS); Yamba, 29°27'S, 153°20'E, on Avicennia marina, 1977, GNS (BRIU2220NS); Boambee Ck, 30°22’S, 153°05’E, on Avicennia marina, 1977, GNS (BRIU2032NS); Nambucca Hds, 30°39’S, 153°00’E, on Avicennia marina, 1977, GNS (BRIU2001NS); Port Macquarie, 31°27'S, 152°54’E, on Avicennia marina, 1977, GNS (BRIU1989NS); Patonga, 33°30’S, 151°20’E, on Avicennia marina, 1978, GNS (BRIU2308NS); Towra Pt., 34°01’S, 151°13’E, on Avicennia marina, 1977, C. Scarlett (BRIU634CS); Sussex Inlet, 35°19’S, 150°36’E, on Avicennia marina, 1977, RWR (BRIU2226NS). 19. Ramalina subfraxinea Ny. in Bull. Soc. Linn. Normandie II, 4: 139 (1870). Type: South America, Columbia, Turbaco, Nova Granata, 1844, Goudot s.n. (PC! — lectotype; H! — isotype; sekikaic acid, homosekikaic acid, an unknown and usnic acid). var. subfraxinea Plate 1 es7; Plate 42; fies3,. 7; Plate 15, fies 2: Thallus corticolous, pale green to grey-green, erect to subpendulous, up to 3 cm long, exceptionally to 12 cm; branching subdichotomous, sparse; branch width 0-5—1-0(—3-5) mm, branches compressed, narrow and canaliculate with attenuate apices or broad and flat with blunt apices; surface matt or shiny, marginal tuberculate pseudocyphellae frequent, laminal punc- tiform pseudocyphellae on broad branches; holdfast delimited; soralia absent. Apothecia sparse to numerous, marginal and subterminal; disc 1-4 mm diam., concave, plane to convex; margin entire; spores ellipsoid, straight to slightly curved, 11—12(—16) x 4—5(—6) um. Chemistry (Australian taxa). Sekikaic acid, homosekikaic acid, ramalinolic acid, 4’-0- demethylsekikaic acid, paludosic acid, 4’-0-methylnorsekikaic acid, and usnic acid or divaricatic acid + faint sekikaic acid and usnic acid. Remarks. Nylander (1870) reported material of R. subfraxinea from Nova Granata (Goudot), Mauritius (Gretan ex hb. Lenorm.), Bourbon I. (Boivin), and the Marianas Islands (Gaudichaud). Examination and chemical analysis of material held at H-NYL and PC revealed that the material from Nova Granata contained either the acids of the sekikaic aggregate or divaricatic acid, whereas all of the material from Mauritius, Bourbon I., and the Marianas Is. contained boninic acid. Those collections found to contain boninic acid belong to the species R. leiodea (Stevens, 1982), regarded as R. subfraxinea var. leiodea in the present study. Material from Nova Granata, collected by Goudot, is held at both PC and H; the material from PC comprises several thalli in good condition on a twig, whereas the specimen held at H is small. Every thallus on the twig held at PC was analysed and the thallus herein designated as lectotype contains sekikaic acid, homosekikaic acid, and faint 4’-0-demethylsekikaic acid. The morphology of this thallus fits the description in the protologue: ‘subsimilis Ramalinae fraxineae minori et attenuatae, sed sporis (subcurvulis) rectioribus tenuioribusque (longit 0,011-18 millim. crassit 0,0035—0,0045 millim.)’. It was found that several thalli on the twig contain divaricatic acid, although morphologically identical to those containing sekikaic acid. The thallus containing sekikaic acid was chosen for lectotypification in preference to a thallus containing divaricatic acid because of the large number of sekikaic acid specimens found in Australia compared to the few specimens which 204 G. NELL STEVENS contain divaricatic acid. Another reason for this choice was the record (Krog & Swinscow, 1976) of the presence of the sekikaic acid aggregate in both PC and H collections of R. subfraxinea. Distribution and habitat. Ramalina subfraxinea var. subfraxinea is tropical in distribution, reaching its southern limit at latitude 24°S, on the eastern coast of Australia; on the northern coastline it has been collected from three areas (Darwin, Gunn Pt, and Gove) at latitude 12°S (Fig. 27). This taxon has two chemical races (a) a sekikaic acid aggregate race and (b) a divaricatic acid race. The former is widely distributed from 12°S—24°S, whereas the divaricatic acid race has a restricted range between latitudes 15°51'S and 18°17’S (Fig. 31), where both races are sympatric. The occurrence of R. subfraxinea var. subfraxinea is almost entirely coastal, with only one inland collection from Butchers Hill (15°51'S) where both races occur on shrubs in a low closed-format community. Thalli from this area contain divaricatic acid + trace amounts of sekikaic acid or sekikaic acid only indicating a close affinity between the two acid races and strengthening the argument for not separating each acid as a different variety of R. subfraxinea. The main phorophytes for this taxon are Ceriops tagal, Rhizophora stylosa, Lumnitzera racemosa, Excoecaria agallocha, and several rain-forest trees which occur along the tropical beaches. World distribution. India?, Australia, and South America (Colombia). Northern Territory: Darwin, Stuart Park, 12°27’S, 130°50’E, on Ceriops tagal, 1981, GNS (BRIU3496NS); Gunn Point, 12°08’S, 130°59’E, on Lumnitzera racemosa, 1979, E. Hegerl (BRIU3115NS); Gove Airport, u 120 130 140 160 L 1 1 i Fig. 27 Distribution of R. subfraxinea var. subfraxinea. RAMALINA IN AUSTRALIA 205 12°18’S, 136°50’E, on Croton sp., 1982, N. Sammy 820576. Queensland: Princess Charlotte Bay, Marrett R., 14°25’S, 144°09’E, on Ceriops tagal, 1979, J. Davie (BRIU2973NS); Lakeland, Butchers Hill, 15°51’S, 144°53’E, on rain-forest shrub, 1983, G.N. (BRIU4107NS); Daintree R. (south branch), 16°16’S, 145°25’E, on Ceriops tagal, 1983, GNS (BRIU4136NS); Newell Beach, S Saltwater Ck 16°25’S, 145°25’E, on Ceriops tagal, 1983, C. McCracken (BRIU4234NS); Mossman Rifle Range, 16°27'S, 145°22’E, on Ceriops tagal, 1983, GNS (BRIU4126NS); Trinity Inlet, 16°56’S, 145°45'E, on Ceriops tagal, 1975, D. Tarte (BRIU2464NS); North Mission Beach, 17°52’S, 146°06’E, on Rhizophora sp., 1979, NCS (BRIU2663NS); South Mission Beach, 17°55’S, 146°06’E, on Ficus sp., 1983, GNS (BRIU4161NS); Hinchinbrook I, The Haven, 18°18'S, 146°10’E, on Lumnitzera sp., 1979, GNS (BRIU3469NS); Towns- ville, 19°15’S, 146°50’E, on Ceriops tagal, 1975, B. Ballment (BRIU3119NS); Shute Harbour, 20°15’S, 148°47’E, on rain-forest tree, 1979, GNS (BRIU3139NS); Hallidays Bay, 20°56’S, 149°01’E, on Ceriops tagal, 1983, GNS (BRIU4122NS); Eimeo, 21°09’S, 149°10’E, on Rhizophora stylosa, 1976, GNS (BRIU1353A.NS); St Lawrence, 22°21'S, 149°32’E, on Excoecaria agallocha, 1976, GNS (BRIU2132NS); Corio Bay, 22°58’S, 150°45’E, on Rhizophora sp., 1975, D. Tarte (BRIU1169NS); Rodds Peninsula, 24°02’S, 151°40’E, on Rhizophora stylosa, 1975, GNS (BRIU1278NS); Turkey, 24°06’S, 151°37’E, on Rhizophora stylosa, 1975, GNS (BRIU1150NS). var. confirmata (Nyl.) N. Stevens, comb. nov. Plate 11, fig. 5; Plate 12, figs 4, 10; Plate 15, figs 6, 8. Ramalina fraxinea [subsp.] confirmata Nyl. in Bull. Soc. linn. Normandie Il, 4: 138 (1870). Type: Australia, Swan R., 1846, Verreaux s.n. (H-NYL 37423! — lectotype; cryptochlorophaeic acid and usnic acid). Thallus corticolous, rarely saxicolous, grey-green, rigid, erect to subpendulous, up to 4 cm long, exceptionally to 11 cm; branching subdichotomous, sparse to moderate; branch width (0-5—)2—4(-—10) mm, branches compressed, narrow and canaliculate to broad and flat, apices attenuate or blunt; surface smooth to rugose, short linear pseudocyphellae along branches or basally only; holdfast delimited; soralia absent. Apothecia common, marginal and subterminal, rarely laminal, small thalli produce subter- minal apothecia subtended by a spur; disc 2-3 mm diam., concave or plane; margin entire, slightly crenate, often incised at maturity; spores ellipsoid, straight or curved (8—)10—12(—14) x 4—5 um. Chemistry. Cryptochlorophaeic acid (M) + paludosic acid (m), 4’-0-methylcryptochlorophaeic acid (m) and usnic acid. Remarks. The name R. fraxinea subsp. confirmata Nyl. was given to material sent to Nylander by Verreaux, supposedly from Swan River, Western Australia. However, it is doubtful whether the specimens came from that area as Verreaux did not personally collect in Western Australia (Lamy, in litt.) and no further specimens have been found there. Material held at H-NYL, numbers 37421, 37423, all contain cryptochlorophaeic acid. Until the present study both boninic acid and cryptochlorophaeic acid taxa were included in R. leiodea (Stevens, 1982); however, the policy adopted herein gives varietal status to different chemical races within a complex if the distribution patterns differ from one another. The boninic acid race therefore retains the name R. subfraxinea var. leiodea and the cryptochlorophaeic acid race has been give the name R. subfraxinea var. confirmata. Distribution and habitat. This taxon is the most common maritime Ramalina along the eastern coastline of Australia, occurring from latitude 10°S (Sue I., north of Cape York) to latitude 31°S (Fig. 28). Although mostly found along the shoreline, R. subfraxinea var. confirmata has been collected from the coastal ranges at Montville, Kenilworth, Bartle Frere, and Atherton Tableland. In the mangrove communities it grows on Avicennia marina and Rhizophora stylosa. It has also been collected on many coastal shrubs and on Casuarina equisetifolia and Callitris columellaris, indicating it has no substrate specificity. This taxon was collected off a rock substrate at Norries Head, New South Wales, where metamorphic rocks form a sheer cliff facing 206 G. NELL STEVENS ' ak 120 130 140 e 150 9 ¥ + i \ { | Fig. 28 Distribution of R. subfraxinea var. confirmata @ and var. leiodea O. the ocean. Such exposure is comparable with the seaward fringe of the mangroves, the position usually occupied by this taxon. On a global scale the distribution pattern of the cryptochlorophaeic acid taxon differs from that of the boninic acid taxon. Ramalina subfraxinea var. confirmata appears to extend eastwards from Australia only as far as New Caledonia, but occurs to the north and in the Indian. Ocean, whereas R. subfraxinea var. leiodea is common on most of the islands in the Pacific Ocean, as well as occurring in the Indian Ocean region. World distribution. Timor, Zanzibar, Australia, and New Caledonia. Queensland: Sue Island, 10°12’S, 142°49’E, on dune shrub, 1981, J. Clarkson 3957; Ingram I, 14°26’S, 144°50’E, on Surina maritima, 1984, M. Godwin (BRIU4282NS); Turtle Group I, 14°42’S, 145°11’E, on beach shrub, 1979, J. Davie (BRIU3145NS); Cooktown, 15°28’S, 145°15’E, on Rhizophora sp., 1983, G.N. (BRIU4112NS); Green Island, 16°45’S, 145°59’E, on beach shrub, 1976, JAE (JAE 2593); Trinity Inlet, 16°56’S, 145°46’E, on Ceriops tagal, 1976, D. Tarte (BRIU2252NS); North Mission Beach, 17°52’S, 146°06’E, on Rhizophora sp., 1979, NCS (BRIU2663NS); Hinchinbrook I, 18°14’S, 146°13’E, on Lumnitzera sp., 1979, GNS (BRIU3465NS); Townsville, 19°15’S, 146°50’E, on Rhizophora stylosa, 1975, B. Ballment (BRIU3117NS); Bowen, Dry Ck Rd, 20°04’S, 148°22’E, on Rhizophora apiculata, 1979, GNS (BRIU3133NS); Eimeo, Sunset Beach, 21°09’S, 149°09’E, on Rhizophora sp., 1976, GNS (BRIU1353NS); Keppel Sands, 23°21'S, 150°47’E, on Avicennia marina, 1975, RWR (BRIU900RR); Port Alma, 23°35’S, 150°51’E, on Rhizophora sp., 1975, RWR (BRIU695RR); Gladstone, 23°50’S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3456NS); Turkey, 24°06’S, 151°37'E, on Ceriops tagal, 1975, RAMALINA IN AUSTRALIA 207 GNS (BRIU1271); Round Hill Head, 24°10’S, 151°54’E, on Rhizophora stylosa, 1975, RWR (BRIU916RR); Burnett Hds, 24°45'S, 152°25'E, on Casuarina sp., 1975, GNS (BRIU4296); Burrum Hds, 25°10'S, 152°35’E, on Rhizophora stylosa, 1975, GNS (BRIU1121NS); Urangan, 25°19’S, 152°55’E, on Callitris columellaris, 1978, GNS (BRIU2840NS); Kin Kin Ck, 26°13’S, 153°00’E, on Avicennia marina, 1979, GNS (BRIU3214NS); Maroochy R, 26°42’S, 153°05’E, on Avicennia marina, 1976, GNS (BRIU1335NS); Bribie I, 27°02’S, 153°08’E, on Lumnitzera racemosa, 1978, GNS (BRIU2283NS); Whyte I, 27°23’S, 153°10’E, on Avicennia marina, 1978, GNS (BRIU2594NS); Mud I, 27°20’S, 153°15’E, on Avicennia marina, 1978, GNS (BRIU2760NS); Coochiemudlo I, 27°32'S, 153°17’E, on Callitris columel- Hd, 28°18'S, 153°34’E, on metamorphic rock, 1980, GNS (BRIU3665NS); Byron Bay, 28°39’S, 153°37’E, on dune shrub, 1978, RWR (BRIU3883*NS); Yamba, 29°27’S, 153°20’E, on Avicennia marina, 1977, GNS (BRIU3177NS); Port Macquarie, 31°27'S, 152°54’E, on Ficus sp., 1975, JAE (JAE 1089); Hunter R., Ash 1,32°56'S,-151°46'E, 1845;.J2P. Verreaux (PC): var. leiodea (Nyl.) N. Stevens, comb. nov. Plate 11, figi6; Plate 12, fio.;8; Plate 15; fie 3: Ramalina subfraxinea [subsp.] leiodea Nyl. in Bull. Soc. linn. Normandie II, 4: 141 (1870). — Ramalina leiodea (Nyl.) Nyl., Lich. Nov. Zel.: 22 (1888). Type: New Caledonia, ‘Donné par le Musée Colonial’, 1861, Deplanche s.n. (PC! — lectotype; boninic acid aggregate). Ramalina boninensis Asah. in J. Jap. Bot. 14: 253 (1938). Type: Bonin Islands, Ogasawara I., Hahajima, 1936, Kusaka s.n. (TNS! — holotype; boninic acid aggregate). Ramalina boninensis f. subcalicariformis Asah. in J. Jap. Bot. 15: 211 (1939). Type: Bonin Islands, Ogasawara I., Hahajima, 1936, Kusakas.n. (TNS! — holotype). Thallus corticolous, grey-green, caespitose, rigid, erect to subpendulous, up to 4 cm high, exceptionally to 9 cm; branching subdichotomous, sparse to moderate; branch width (0-5—)2—4 mm, branches compressed, flat or canaliculate when narrow, apices attenuate; surface matt, smooth to rugose, short linear pseudocyphellae sometimes present; holdfast delimited; soralia absent. . Apothecia common, marginal and subterminal, rarely laminal, small thalli produce subter- minal apothecia subtended by a spur; disc 2-3 mm diam., concave to plane; margin entire or incised at maturity; spores ellipsoid, straight or slightly curved; 10—12(—16) x 4-5 um. Chemistry. Boninic acid (M), 2-0-methylsekikaic acid (m), 2,4’-di-0-methylnorsekikaic acid (m), 4’-0-methylpaludosic acid (m), 4,4’-di-0-methylcryptochlorophaeic acid (m), and usnic acid. Remarks. Prior to the present study R. subfraxinea var. leiodea had species status. The history of this taxon, however, is one of name changing since its first identification by Nylander in 1859. Specimens of this taxon have been identified as follows: Ramalina Ecklonii Sprengel (Nylander, 1859); Ramalina calicaris f. Ecklonii (Sprengel) Nyl. (Nylander, 1861); Ramalina subfraxinea [subsp. | /leiodea Nyl. (Nylander, 1870) and Ramalina leiodea Nyl. (Nylander, 1888). A thorough investigation of specimens of this taxon held at PC and H was undertaken by Stevens (1982). It was found that two acid races (viz. boninic acid and cryptochlorophaeic acid) were present in taxa bearing this name. Because the material designated lectotype contained boninic acid, the species R. boninensis Asah. and R. boninensis f. subcalicariformis were reduced to synonomy at that time. Krog & Swinscow (1976) commented on the similarity between the East African taxon R. maritima and material from Bourbon I. and Mauritius. Specimens from the latter two areas had been identified as R. subfraxinea by Nylander (1870). Chemical analysis showed they contained boninic acid (Krog & Swinscow, 1976), which was also confirmed in this study (see comments under R. subfraxinea var. subfraxinea). Morphologically and chemically R. subfraxinea var. leiodea and R. maritima resemble one another but differences in spore size sets them apart, the former taxon produces spores in the range 8-16 X 4—5 wm, whereas R. maritima spores lie in the range 18-25 x 5-6 um. Distribution and habitat. Ramalina subfraxinea var. leiodea occurs from latitudes 18°S—29°S along the eastern coastline of Australia (Fig. 28), with its most frequent occurrence in the 208 G. NELL STEVENS subtropics. The distribution range overlaps that of R. subfraxinea var. confirmata, a mainly tropical taxon which extends into the subtropics so that both taxa are sympatric for a large part of their distribution (Fig. 31). This taxon occurs on mangroves (Avicennia marina and Rhizophora stylosa) and on coastal trees (Callitris columellaris and Casuarina equisetifolia). World distribution. Bonin Is, Marianas Is, Mauritius, Réunion I., Australia, Lord Howe I., Norfolk I., New Caledonia, New Hebrides, Cook Is (Rarotonga), and Hawaii. Queensland: Hinchinbrook I, 18°14’S, 146°50’E, on Ceriops tagal, 1979, GNS (BRIU3479NS); Long I, Whitsunday Group, 20°21’S, 148°51’E, on coastal shrub, 1979, A. B. Cribb (BRIU3225N§); Eimeo, 21°09'S, 149°09’E, on Ceriops tagal, 1977, GNS (BRIU3155NS); St Lawrence, 22°20'S, 149°32’E, on Ceriops tagal, 1976, GNS (BRIU2455NS); Gladstone, 23°51’S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3285NS); Turkey, 24°06’S, 151°38’E, on Rhizophora stylosa, 1975, GNS (BRIU1148NS); Burnett Hds, 24°46’S, 152°25’E, on Rhizophora stylosa, 1976, GNS (BRIU4296NS); Toogoom, 25°15’S, 152°40’E, on Callitris columellaris, 1983, GNS (BRIU4152NS); Urangan, 25°19’S, 152°55’E, on Callitris columellaris, 1978, GNS (BRIU2884NS); Kin Kin Ck, 26°13’S, 153° on Casuarina sp., 1979, GNS (BRIU3223NS); Noosa Hds, 26°23'S, 153°05’E, on rain-forest tree, 1980, GNS (BRIU3547NS); Moreton I, 27°20’S, 153°25’E, on Callitris columellaris, 1980, RWR (BRIU3476NS); Cape Byron, 28°37'S, 153°36’E, on Acacia sp., 1980, RWR (BRIU3883NS). New South Wales: Yamba, 29°27'S, 153°20’E, on Avicennia marina, 1977, GNS (BRIU2122NS); Broomes Hd, 29°37'S, 153°20’E, on Casuarina sp., 1982, GNS (BRIU4311NS). var. norstictica N. Stevens, var. nov. Plate 11, fig. 8. Thallus corticolus, viridi-griseus caespitosus rigidus erectus vel subpedulus, 1-5—6-5 cm longus, ramifi- catione subdichotoma sparsa vel mediocri; rami lati ad 1-1-5 mm, applanati vel canaliculati, apice attenuati; superficies impolita, laevigata vel rugosa et crassa, pseudocyphellis laminalibus linearibus, ad fundamenta, tubercularis pseudocyphellis marginalibus, sparsa vel numerosa; soralia nulla. Apothecia numerosa, marginalia vel subterminalia, calcarata; discus ad 3-5 mm in diametro, concava vel plana vel convexa; sporae ellipsoideae vel raro fusiformae, rectae vel raro curvatae, 9-16(—20) x 3-5—5-5 um. Acidum norsticticum + connorsticticum et usnicum continens. Holotype: Australia, Queensland, St Lawrence, on Ceriops tagal, 11 July 1978, G. N. Stevens 2459 (MEL 1048088!). Thallus corticolous, pale green to green-grey, caespitose, rigid erect to subpendulous, 1-5—6-5 cm long; branching subdichotomous, sparse to moderate; branch width 1-1-5 mm, compressed, flat or canaliculate, some marginal splitting of the branches between the upper and lower surfaces, apices attenuate; surface matt, smooth to rugose and coarse, with or without basal pseudocyphellae, tuberculate pseudocyphellae marginal few to many when present; holdfast delimited; soralia absent. Apothecia common, marginal and subterminal, some spurred; disc 3-5 mm diam., concave, plane to convex; margin thin, with or without lateral branchlets; spores ellipsoid rarely fusiform, straight, rarely curved, 9—16(—20) Xx 3-5—5-5 wm. Chemistry. Norstictic acid, + connorstictic acid, and usnic acid. Remarks. This taxon is an Australian endemic; examination of esorediate maritime taxa from overseas herbaria failed to find any norstictic acid taxon which resembled it. In Australia the taxon closest to it morphologically is R. subfraxinea var. subfraxinea. Both taxa are sympatric for part of their distribution and have similar habitats so that the similarity in morphology could be due to convergent morphology under the same environmental pressures (Fig. 31). Ramalina subfraxinea vat. norstictica produces branchlets around the margins of the apothe- cia at times, but this is not a constant character and cannot be used taxonomically. Splitting along the margins of the branches also occurs, but this is also common in the salazinic acid taxon belonging to the R. subfraxinea complex. The rank of variety was given to this norstictic acid taxon because of its distinct chemistry and its geographical occurrence along the northern Australian coastline (Figs 29, 31). Distribution and habitat. Ramalina subfraxinea var. norstictica occurs along the northern and RAMALINA IN AUSTRALIA 209 , A T T 120 130 140 150 4) bs 107 0 800km SS — i i ae | (i @ 20- Xs 9 i) ¢ 30+ aw 3 oo ps 1 i 1 l Fig. 29 Distribution of R. subfraxinea var. norstictica. eastern coastlines of Australia (Fig. 29). Its absence from an area between latitudes 17°-11°S on the eastern coast is an anomaly since it is found on the northern coastline between these latitudes. It is in fact replaced by the salazinic acid taxon R. tropica in this section of the eastern coastline. It could be inferred that a change from one depsidone to the other came about with change in environmental pressures as norstictic acid is almost certainly a biogenetic precursor of salazinic acid (Elix, in litt.). However, no thalli were found containing both norstictic acid and salazinic acid. The morphology of the norstictic acid taxon is variable and appears to be ecophenotypic, ranging from thalli with narrow, canaliculate branches bearing subterminal apothecia, to broad, flat, branched thalli bearing large, marginal apothecia. Tuberculate pseudocyphellae occur on thalli growing in very dry or polluted areas. The spores of R. subfraxinea var. norstictica are intermediate (in both size and shape) between the small ellipsoid spores of R. subfraxinea var. confirmata and the large fusiform spores of R. tropica. World distribution. Australia. Western Australia: Cygnet Bay, 16°35’S, 123°02'E, on Bruguiera sp., 1980, S. Kenneally (PER001045); Swan Point, 16°25’S, 123°02’E, 1906, W. V. Fitzgerald (NSW L4117). Northern Territory: Port Darwin, 12°26'S, 130°48’E, 1890, M. Holtze (MEL 9477); East Alligator R. 12°08’S, 132°40’E, on Ceriops tagal, 1979, E. Hegerl, (BRIU3114NS); Cobourg Pen., Caiman Ck, 11°25’S, 132°06’E, on mangrove, 1968, J. R. Maconochie (MEL100288); Tomkinson R, 12°13’S, 134°16’E, on Avicennia sp., 1975, D. Grace (MEL1012777); Melville Bay, 12°15’S, 136°40’E, on Ceriops tagal, 1948, R. L. Specht (SA 97528120); 210 G. NELL STEVENS Groote Eylandt, Hemple Bay, 13°49’S, 136°38’E, on mangroves, 1948, R. L. Specht (SA 97527533); Carrington Ck, 15°44’S, 136°36’E, on mangroves, McKay (MEL102350). Queensland: Weipa, 12°38’S, 141°55’'E, on mangroves, 1974, R. L. Specht (BRIU3584NS); Hinchinbrook I, 18°22’S, 146°15’E, on Ceriops tagal, 1979, GNS (BRIU3185NS); Townsville, 19°15’S, 146°50’E, on Ceriops tagal, 1975, B. Ballment (BRIU3120NS); Bowen, Dry Ck Rd, 20°04’S, 148°22'’E, on Ceriops tagal, 1979, GNS (BRIU3141NS); Mackay, Eimeo, 21°09’S, 149°09’E, on Ceriops tagal, 1976, GNS (BRIU3434NS); St Lawrence, 22°21'S, 149°32’E, on Excoecaria agallocha, 1976, GNS (BRIU2463NS); Corio Bay, 22°58’S, 150°46’E, on Rhizophora sp., 1975, D. Tarte (BRIU1170NS); Keppel Sands, 23°21'S, 150°47’E, on Excoecaria agallocha, 1975, RWR (BRIU760RR); Port Alma, 23°35’S, 150°51'E, on Ceriops tagal, 1975, RWR (BRIU1039RR); Gladstone, 23°50’S, 151°16’E, on Rhizophora stylosa, 1980, GNS (BRIU3259NS). 20. Ramalina tropica N. Stevens, sp. nov. Plate 12, figs 1, 2, 6. Thallus corticolus et saxicolus, virentes-pallido caespitosus rigidus erectus, usque ad 4-0 cm altus, ramificatione subdichotoma sparsa vel mediocri; rami lati ad 0-5—1-0 mm, subteretes vel teretes vel raro planares, apice attenuati curvi; superficies impolita vel nitida, laevigata vel rugosa, pseudocyphellis laminalibus linearibus instructa aut sine pseudocyphellis; soralia nulla. Apothecia raro aut vulgariter, subterminalia et calcarata vel marginala, ramis flexuosis insidentes; discus ad (1—)1-5—3-0 mm in diametro, planus vel convexus; sporae fusiformae, rectae, 11-18(-—22) x 3-8-S(-6) wm. Acidum salazinicum + protocetraricum et usnicum continens. Holotype: Australia, Queensland, Lizard Island, 26 December 1974, R. L. Specht (MEL 1048087!). Thallus corticolous, or saxicolous, pale green, caespitose, rigid, erect, usually to 2 cm high, exceptionally to 4 cm; branching subdichotomous, sparse to moderate, branches arising from the base with lateral branchlets common; branch width 0:5—1-0 mm, subterete to almost terete, rarely flat, never canaliculate, some marginal splitting of the branches between the upper and lower surfaces, apices tapering distally, often recurved; surface matt or shiny, smooth to rugose, fine linear pseudocyphellae and tuberculate pseudocyphellae sometimes present; holdfast delimited; soralia absent. Apothecia rare to common, marginal on flexuous branches or subterminal with the branch continuing to grow forming along curved spur up to 10 mm, disc (1—)1-5—3-0 mm diam., plane to convex, some assuming a helmet shape; margin entire; spores fusiform, rarely ellipsoid, straight, rarely bent, some appearing as 3-celled; 12—18(—22) x 3-8-5(—6) wm. Chemistry. Salazinic acid, + protocetraric acid, and usnic acid. Remarks. The erection of this new species has been made cautiously. There were several reasons for making this decision rather than regarding the salazinic acid taxon as another chemical variety of the R. subfraxinea complex. (i) Morphologically the four depside taxa and the norstictic acid taxon closely resemble one another in producing thalli with flat or canaliculate branches, whereas the salazinic acid taxon mainly produces subterete branches and in some saxicolous forms, terete branches; (ii) The large size of the spores is in excess of the size produced by any of the other taxa in the complex; (iii) the distinct depsidone chemistry. Ramalina tropica resembles the East African species R. fecunda Krog & Swinscow, which contains salazinic acid and produces large fusiform spores. However, the latter taxon has branches which are flat or canaliculate, not subterete (Krog, in litt). Other maritime species producing salazinic acid are R. zollingeri Szat. (Java), R. sideriza Magnusson (Hawaii), and undescribed material (Fiji). Before any affinities could be established between these taxa and R. tropica more material would need to be examined. Plate 12 Thallus growth forms in the Australian Ramalinae. Scale in mm. Fig. 1 R. tropica (sterile form). Fig. 2 R. tropica (type specimen). Fig. 3 R. subfraxinea var. subfraxinea (top — narrow form; bottom — broad form). Fig. 4 R. subfraxinea var. confirmata (broad form). Fig. 5 R. subfraxinea var. norstictica (sparse apothecia). Fig. 6 R. tropica (terete, inland form). Fig. 7 R. subfraxinea var. subfraxinea (broad form). Fig. 8 R. subfraxinea var. leiodea. Fig. 9 R. subfraxinea var. norstictica. Fig. 10 R. subfraxinea var. confirmata. RAMALINA IN AUSTRALIA 212 G. NELL STEVENS 120 130 140 150 10) 204 30 | g f ie 40~ l 1 L 4 Fig. 30 Distribution of R. tropica. Distribution and habitat. The distribution of R. tropica is restricted to an area on the eastern coastline of Cape York between latitudes 13°S (Claudie River) and 18°S (Hinchinbrook I.) [Figs 30, 31]. Because of its restricted distribution range compared with that of the other R. subfraxinea taxa, indications are that R. tropica may have certain habitat requirements not found elsewhere along the coast. Ramalina tropica is both corticolous and saxicolous over its range, occurring both coastally and on the coastal hills. Morphologically variable, it produces several morphotypes, e.g. at latitude 18°S it is sympatric with the rest of the R. subfraxinea complex and closely resembles these taxa, illustrating the phenomenon of convergent morphology. At latitude 16°S it is sympatric with R. subfraxinea var. subfraxinea and both produce a straggly, wiry growth form with dense lateral branchlets; however, the characteristic rounding of the branches in R. tropica is detectable when comparing the thalli with flattened branches of R. subfraxinea var. subfrax- inea thalli. Corticolous thalli of R. tropica attain their maximum length of 4-0 cm on Lizard Island in the mangrove environment where Ceriops tagal is the usual phorophyte. The branches are very narrow, subterete, and bear marginal apothecia subtended by markedly attenuate recurved branch apices (Plate 12, fig. 1). Ramalina tropica also grows on granite rock on Lizard Island, but the saxicolous morph differs from the corticolous morph in producing a rigid, sparsely branched thallus, bearing few or no apothecia. Inland collections of this taxon were made at Abbey Peak (650 m altitude) at latitude 14°18’S, RAMALINA IN AUSTRALIA 213 Tropic Fig. 31 Distribution ranges of the six acid taxa in the R. subfraxinea complex: var. subfraxinea (sekikaic acid race) ; var. subfraxinea (divaricatic acid race) ——-; var. norstictica -—-—-—- ; var. confirmata — ; var. leiodea ----; R. tropica ———. where it occurs on windblown shrubs at the summit, and at Lighthouse Mountain (400 m altitude) at latitude 16°38’S where it occurs on granite rock at the summit. These morphotypes are more terete than coastal thalli and produce a thicker cortex. Increased thickness of the cortex is a feature reported by Rundel (1982) as an adaptation to decrease the rate of evaporation from the thallus in extreme conditions where selection for low rates of evaporative loss is more important than rapid rate of water uptake. The terete branches and fasciculate growth form (Plate 12, fig. 6) may be a response to increased moisture stress as discussed by Sheard (1978); these adaptations would ensure that less thallus area was effected by heat in an environment which is exposed to extreme heat and dryness for long periods. World distribution. Australia. Queensland: Claudie River, Charlie Taylors Landing, 12°36’S, 143°42’E, on Ceriops tagal, 1983, GNS (BRIU4111NS); Abbey Peak, 14°18’S, 144°30’E, on shrub on summit, 1983, C. McCracken (BRIU4239NS); Lizard I, 14°41’S, 145°28’E, on granite, 1974, R. L. Specht (BRIU3780NS); Lizard I, 14°41’S, 145°28’E, on mangroves, 1974, R. L. Specht (MEL1048087) Type; Lighthouse Mt, between Mt Carbine & Mt Molloy, 16°35’S, 145°13’E, on granite, 1983, C. McCracken (BRIU4218NS); Daintree R. south branch, 16°18’S, 145°24’E, on Ceriops tagal, 1983, GNS (BRIU4189NS); Mossman Rifle Range, 16°29’S, 145°25’E, on Ceriops tagal, 1983, GNS (BRIU4188NS); South Mission Beach, 17°56’S, 146°05’E, on rain-forest tree, 1983, GNS (BRIU4163NS); Murray R. 18°02’S, 146°03’E, on Ceriops tagal, 1979, D. Tarte (BRIU3159NS); Hinchinbrook I, 18°20’S, 146°15’E, on Ceriops tagal, 1979, GNS (BRIU3160NS). 214 G. NELL STEVENS BERS. MUS. Panis, | PH ays re ele Sp )5 ‘oH OBIE MULLER 4V7O9/ Dn (Brwroar.e front APA re b> Ui ty fin Fi MA lor. Nrrevhan FFF Plate 13. Photographs of Australian Ramalina types. Scale in mm. Fig. 1 R. australiensis (PC). Fig. 2 R. myrioclada (G). Fig. 3 R. exiguella (BRI). Fig. 4 R. glaucescens (M). Fig. 5 R. lacerata (G). Fig. 6 R. leiodea var. fastigiatula (G). Fig. 7 R. calicaris var. australica (H). Fig. 8 R. brevis (NSW). RAMALINA IN AUSTRALIA 215 Ps Pe AD ci a Ce (i PIER MULLER ANB. 1896 Karn atin Du / Bey aeroree. thallus, Kem: KC -PD medulia: Kee Commi Cae PD potent! 4%. Kimurhk Ses vt. Plate 14 Photographs of Australian Ramalina types. Scale in mm. Fig. 1 R. farinacea var. squarrosa (G). Fig. 2 R. farinacea var. dendroides (G). Fig. 3 R. perpusilla (BRI). Fig. 4 R. fissa (G). Fig. 5 R. inflata (FH). Fig. 6 R. tasmanica (H). Fig. 7 R. nervulosa var. nervulosa (G). Fig. 8 R. pacifica (TNS). 216 G. NELL STEVENS GS2 Vea nahaa Flee fo. i dow Shenk. — i gy, fiatgann ptr tads fh. fo ther PRM Woden — pe os NE at pass Kye th Les ach. a ers 2 OkBrs Sh ed save g 7593, RAISUAAING GUVANVE "WAIN VENSH MOTH TE eg oo Pia te - Liotta Plate 15 Photographs of Australian Ramalina types. Scale in mm. Fig. 1 R. unilateralis (G). Fig. 2 R. subfraxinea (PC). Fig. 3 R. leiodea (PC). Fig. 4 R. celastri (FH — isotype). Fig. 5 R. ovalis (BM). Fig. 6 R. fraxinea**R. confirmata (H — isotype). Fig. 7 R. minscula var. alba (WELT). Fig. 8 R. fraxinea**R. confirmata (H — lectotype). RAMALINA IN AUSTRALIA 217 Appendix. Disposition of excluded Ramalina taxa The position of other Ramalina taxa recorded from Australia in Wetmore (1963), and Weber & Wetmore (1972) is indicated below. . anceps = R. australiensis Nyl. or R. filicaulis N. Stevens . angulosa = R. australiensis Nyl. . calicaris = R. glaucescens Krempelh. . calicaris var. canaliculata = R. glaucescens Krempelh. . complanata = R. subfraxinea Ny]. confirmata = R. subfraxinea var. confirmata (Nyl.) N. Stevens. dendriscoides = R. tenella Mill. Arg. or R. peruviana Ach. dendriscoides var. minor = R. tenella Mill. Arg. or R. peruviana Ach. . dilacerata var. alba = R. inflata subsp. perpusilla (Stirton) N. Stevens ecklonii = R. celastri (Sprengel) Krog & Swinscow ecklonii var. membranacea = R. celastri (Sprengel) Krog & Swinscow ecklonii var. ovalis = R. celastri subsp. ovalis (J. D. Hook & Taylor) N. Stevens ecklonii var. tenuissima = R. celastri (Sprengel) Krog & Swinscow . fastigiata = R. glaucescens Krempelh. fraxinea = R. glaucescens Krempelh. fraxinea var. ampliata = R. glaucescens Krempelh. . fraxinea f. platyna = R. fissa (Mill. Arg.) Vainio . fraxinea var. taeniata = R. glaucescens Krempelh. furcellata = R. australiensis Ny). geniculata = R. inflataJ. D. Hook. & Taylor geniculata var. olivacea = R. inflata subsp. perpusilla (Stirton) N. Stevens gracilenta = R. exiguella Stirton gracilis = R. exiguella Stirton or R. filicaulis N. Stevens homalea = unknown? inflata var. gracilis = R. inflata (J. D. Hook. & Taylor) J. D. Hook. & Taylor intermedia = R. inflata subsp. australis N. Stevens javanica = R. peruviana Ach. linearis = R. celastri (Sprengel) Krog & Swinscow . pollinaria = R. unilateralis F. Wilson polymorpha = unknown polymorpha f. emplecta = unknown pusilla = R. inflata subsp. australis N. Stevens scopulorum = R. subfraxinea Ny]. var. ? scopulorum var. cuspidata = R. subfraxinea Nyl. var. ? scopulorum var. subfarinacea = R. pacifica Asah. . tayloriana = R. nervulosa var. luciae (Molho et al.) N. Stevens usnea = R. australiensis Nyl. . usneoides = R. australiensis Nyl. . yemensis = R. celastri (Sprengel) Krog & Swinscow . yemensis var. ovalis = R. celastri subsp. ovalis (J. D. Hook. & Taylor) N. Stevens DD WWD DDD DDD D DD DAD DDR DDD RD DDR WD DR DDD RAZ Acknowledgements This research was carried out for a Doctor of Philosophy thesis at the University of Queensland Botany Department. Some funding was provided from University Research Grant Funds. I thank my supervisors Dr R. W. Rogers and Professor H. T. Clifford for their discussions and guidance during the course of this research, and Dr J. A. Elix for his generous help with chemical problems and Mr R. B. Filson for his help and encouragement. I wish to thank all overseas lichenologists with whom I corresponded during this study, especially Professors H. Krog and D. L. Hawksworth, Drs M. E. Hale and D. J. Galloway, and Mr P. W. James. I would also like to thank the directors and curators of the following herbaria for supplying material on loan: AD, BM, BRI, CBG, EDIN, FH, G, HO, LD, M, MICH, MEL, NSW, O, PC, PER, S, TNS, TUR, UPS, UWA, US, and W. In addition I am grateful to the following persons who placed their private collections at my disposal: C. Smith (Hawaiian Ramalinae), G. & B. Hayward (New Zealand Ramalinae), H. Osorio, M. Pinto (South American Ramalinae), J. Clarkson (Torres Strait Ramalinae), R. Seppelt 218 G. NELL STEVENS (Macquarie Island Ramalinae), R. Specht, J. Berjak (South African Ramalinae), G. Kantvilas (Tas- manian Ramalinae) and A. Archer, E. Dahl, J. Elix, R. Rogers, N. Sammy, C. Scarlett, H. Streimann (mainland Australia Ramalinae). I am grateful to Dr A. Henssen for supplying two of the photographs used in this paper. All SEM and TEM photographs were taken by Mr J. Hardy of the Electron Microscope Centre, University of Queensland. I thank M. Lamy for supplying information from the Verreaux journals. References Abbayes, H. des 1952. Lichens récoltés en Guinée Francaise et en Céte d'Ivoire V-VI. Bull. Inst. fr. Afr. noire B, 14: 19-27. Acharius, E. 1810. Lichenographia universalis. Gottingen. — 1814. Synopsis methodica lichenum. Lund. Agardh, C. A. 1821. Aphorismi botanici. Lund. Asahina, Y. 1938. Lichenologische Notizen. X. J. Jap. Bot. 14: 251-255. — 1939. Ramalina-Arten aus Japan (II). J. Jap. Bot. 15: 205-223. Awasthi, D. D. 1965. Catalogue of the lichens from India, Nepal, Pakistan and Ceylon. Beih. Nova Hedwigia 17: 1-137. Bailey, F. M. 1886. A synopsis of the Queensland flora containing both the phaenogamous and cryptog- amous plants. First supplement: 1—99. Brisbane. Beltman, H. A. 1978. Vegetative Strukturen der Parmeliaceae und ihre Entwicklung. [Bibltheca lich. 11]. Vaduz. Bowler, P. A. 1977. Ramalina thrausta in North America. Bryologist 80: 529-532. — 1981. Cortical diversity in the Ramalinaceae. Can. J. Bot. 59: 427-452. & Rundel, P. W. 1972a. Ramalina bajacalifornica. A new species of lichen from Baja California. Bryologist 75: 365-366. — 1972b. A new species of Ramalina from Sinaloa, Mexico. Bryologist 75: 574-576. —— — 1973. Two new lichens (Ramalina) from Baja California, Mexico. Bryologist 76: 211-213. —— — 1974. The Ramalina intermedia complex in North America. Bryologist 77: 617-623. —— —— 1977. Synopsis of a new lichen genus Fistulariella Bowler & Rundel (Ramalinaceae). Mycotaxon 6: 195-202. 1978. The Ramalina farinacea complex in North America: chemical, ecological and morpho- logical variation. Bryologist 81: 386—403. Brandt, T. 1906. Beitrage zur anatomischen Kenntnis der Flechtengattung Ramalina. Hedwigia 45: 124-158. Brodo, I. M. 1978. Changing concepts regarding chemical diversity in lichens. Lichenologist 10: 1-11. —— & Hawksworth, D. L. 1977. Alectoria and allied genera in North America. Op. bot. Soc. bot. Lund 42: 1-164. Carey, S. W. 1976. The expanding Earth. Amsterdam. Chester, D. O. & Elix, J. A. 1978. The identification of four new metadepsides in the lichen Ramalina asahinae. Aust. J. Chem. 31: 2745-2749. Choisy, M. 1954. Classification des lichens fruiticuleux. Bull. Mens. Soc. linn. Lyon 23: 229-240. Croizat, L. 1952. Manual of phytogeography. Den Haag. , Nelson, G. & Rosen, D. E. 1974. Centres of origin and related concepts. Syst. Zool. 23: 265-287. Crombie, J. M. 1880. Enumeration of Australian lichens in Herb. Robert Brown (Brit. Mus.), with descriptions of new species. J. Linn. Soc. Bot. 17: 390-401. Culberson, C. F. 1972. Improved conditions and new data for the identification of lichen products by a standardized thin-layer chromatographic method. J. Chromat. 72: 113-125. —— & Culberson, W. L. 1977. Chemosyndromic variation in lichens. Syst. Bot. 1: 325-339. —— & Hale, M. E. 1973. Chemical and morphological evolution in Parmelia sect. Hypotrachyna. Product of ancient hybridization? Brittonia 25: 162-173. Culberson, W. L. 1967. Analysis of chemical and morphological variation in the Ramalina siliquosa species complex. Brittonia 19: 333-352. 5 —— 1969a. The behaviour of the species of the Ramalina siliquosa group in Portugal. Ost. bot. Z. 116: 85-94. —— 1969b. The use of chemistry in the systematics of the lichens. Taxon 18: 152-166. —— & Culberson, C. F. 1967. Habitat selection by chemically differentiated races of lichens. Science, N. Y. 158: 1195-1197. —— —— 1968. The lichen genera Cetrelia and Platismatia (Parmeliaceae). Contr. U.S. natn. Herb. 34: 449-558. RAMALINA IN AUSTRALIA 219 —— —— 1970. A phylogenetic view of chemical evolution in the lichens. Bryologist 73: 1-31. —_—-—— & Johnson, A. 1977. Correlation between secondary-product chemistry and ecogeography in the Ramalina siliquosa group (lichens). Plant Syst. Evol. 127: 191-200. Ducker, S. C. 1979. History of Australian phycology: the significance of early French exploration. Brunonia 2: 19-42. Duncan, U. K. 1970. Introduction to British Lichens. Arbroath. Du Rietz, G. E. 1924. Die Soredien und Isidien der Flechten. Svensk. bot. Tidskr. 18: 371-396. — 1926. Morfologi och systematik hos slaktet Ramalina sarskilt dess skandinaviska arter. Svensk. bot. Tidskr. 20: 295-299. Elix, J. A. 1982. Peculiarities of the Australasian lichen flora: accessory metabolites, chemical and hybrid strains. J. Hattori bot. Lab. 52: 407-415. Eschweiler, F. G. 1824. Systema lichenum genera exhibens rite distincta pluribus novis adaucta. Niirnberg. Filson, R. B. 1976. Australian lichenology: a brief history. Muelleria 3 (3): 183-190. —— & Rogers, R. W. 1979. Lichens of South Australia. Adelaide. Follmann, G. 1967. Die Flechtenflora der nordchilenischen Nebeloase Cerro Moreno. Nova Hedwigia 14: 215-281. — & Huneck, S. 1969. Mitteilungen iiber Flechteninhaltsstoffe LXI. Zur Chemotaxonomie der Flechtenfamilie Ramalinaceae. Willdenowia 5 (2): 181-216. Galloway, D. J. 1979. Biogeographical elements in the New Zealand lichen flora. Jn D. Bramwell (Ed.), Plants and islands: 201-224. London. — 1985. Flora of New Zealand. Lichens. Wellington. Hale, M. E. 1973. Fine structure of the cortex in the lichen family Parmeliaceae viewed with the scanning electron microscope. Smithson. Contr. Bot. 10: 1-92. — 1975. A monograph of the lichen genus Relicina (Parmeliaceae). Smithson. Contr. Bot. 26: 1-32. — 1976a. A monograph of the lichen genus Pseudoparmelia Lynge (Parmeliaceae). Smithson. Contr. Bot. 31: 1-62. — 1976b. A monograph of the lichen genus Bulbothrix Hale (Parmeliaceae). Smithson. Contr. Bot. 32: 1-29. — 1976c. A monograph of the lichen genus Parmelina Hale (Parmeliaceae). Smithson. Contr. Bot. 33: 1-60. — 1978. A new species of Ramalina from North America (Lichenes: Ramalinaceae). Bryologist 81: 599-601. 1981. Pseudocyphellae and pored epicortex in the Parmeliaceae: their delimitation and evolutionary significance. Lichenologist 13: 1-10. & Culberson, W. L. 1970. A fourth checklist of the lichens of the continental United States and Canada. Bryologist 73: 499-543. Hawksworth, D. L. 1976. Lichen chemotaxonomy. Jn D. H. Brown, D. L. Hawksworth, & R. H. Bailey (Eds), Lichenology: progress and problems: 139-148. London. — 1977. A bibliographic guide to the lichen floras of the world. Jn M. R. D. Seaward (Ed.) Lichen ecology: 437-502. London. Henssen, A. & Jahns, H. M. 1973 [‘1974’]. Lichenes. Stuttgart. Holmgren, P. K., Keuken, W. & Schofield, E. K. (Eds) 1981. Index herbariorum. 7th ed. [Regnum veg. 106]. Hooker, J. D. & Taylor, T. 1844. Lichenes antarctici; being characters and brief descriptions of the new lichens discovered in the southern circumpolar regions, Van Diemen’s Land and New Zealand, during the voyage of H.M. discovery ships Erebus and Terror. J. Bot. Lond. 3: 634-658. Howe, R. H. 1913-14. North American species of the genus Ramalina I-VIII. Bryologist 16: 65—74, 81—89; 17: 1-7, 17-27, 33-40, 49-52, 65-69, 81-87. Hue, A. M. 1890. Lichenes exoticos. Nouv. Archs Mus. Hist. nat. Paris U1, 2: 209-322. 1901. Lichenes extra-Europaei, a pluribus collectoribus ad museum Parisiense missi. Nouv. Archs Mus. Hist. nat. Paris, V1, 3: 80-122. Imshaug, H. A. 1972. Typification of Ramalina usnea (L.) R. H. Howe. Lichenologist 5: 317-318. — & Brodo, I. M. 1966. Biosystematic studies on Lecanora pallida, and some related lichens in the Americas. Nova Hedwigia 12: 1-59. Jahns, H. M. 1974 [‘1973’]. Anatomy, morphology and development. Jn V. Ahmadjian & M. E. Hale (Eds), The lichens: 3-58. New York. Jergensen, P. M. 1974. Dirinaria applanata (Fée) Awasthi on the Agores and a comment on the taxonomical value of sekikaic acid in the genus. Port. Acta biol. B, 12, 5-9. 220 G. NELL STEVENS 1977. Foliose and fruticose lichens from Tristan da Cunha. Norske vidensk.-Akad. I. Mat.- Naturvitensk. Kl. Skr. 11, 36: 1-40. 1979. Phytogeographical relationship of the lichen flora of Tristan da Cunha (excluding Gough Island). Can. J. Bot. 57: 2279-2282. — 1983. Distribution patterns of lichens in the Pacific region. Aust. J. Bot., supp. ser. 10: 43-66. Keuck, G. 1979. Die systematische stellung der Ramalinaceae. Ber. dt. bot. Ges. 92: 507-518. Kristinsson, H. 1969. Chemical and morphological variation in the Cetraria islandica complex in Iceland. Bryologist 72: 344-357. Krog, H. & James, P. W. 1977. The genus Ramalina in Fennoscandia and the British Isles. Norw. J. Bot. 24: 15-43. — & Osthagen, H. 1980. The genus Ramalina in the Canary Islands. Norw. J. Bot. 27: 255-296. — & Swinscow, T. D. V. 1974. Ramalina species with a hollow thallus (Fistularia) in East Africa. Norw. J. Bot. 21: 111-124. — — 1975. Some Ramalina species with punctiform pseudocyphellae in East Africa. Norw. J. Bot. 22: 269-276. —— — 1976. The genus Ramalina in East Africa. Norw. J. Bot. 23: 153-175. Landron, C. I. 1972. The lichen genus Ramalina Ach. in the West Indies with notes on its role in the vegetation of Puerto Rico. Ph.D. thesis (unpublished), Michigan State University. Magnusson, A. H. 1956. A catalogue of the Hawaiian lichens. Ark. Bot. II, 3: 223-402. —— & Zahlbruckner, A. 1945. Hawaiian Lichens III. The families Usneaceae and Physciaceae. Ark. Bot. 32A (2): 1-89. Malme, G. O. 1934. Die Ramalinen der ersten Regnellschen Expedition. Ark. Bot. 26A (12): 1-9. Massalongo, A. B. 1854. Neagenea lichenum. Verona. Moberg, R. 1977. The lichen genus Physcia and allied genera in Fennoscandia. Symb. Bot. Upsal. 22 (1): 1-108. Molho, D., Bodo, B., Culberson, W. L. & Culberson, C. F. 1981. A chemically distinctive new Ramalina from Fiji. Bryologist 84: 396-398. Montagne, J. F. C. 1852. Lichens. Jn C. Gay, Historia fiscia y politica de Chile 8: 53-228. Moore, B. J. 1968. The macrolichen flora of Florida. Bryologist 71: 161-266. Miller, J. 1883. Lichenologische Beitrage XVIII. Flora, jena 66: 17-25. Nix, H. A. 1981. The environment of Terra Australis. Jn A. Keast (Ed.), Ecological biogeography of Australia: 105-133. Den Haag. — 1982. Environmental determinants of biogeography and evolution in terra australis. Jn W. R. Barker & P. J. M. Greenslade (Eds), Evolution of the flora and fauna of arid Australia: 47-64. Frewville. Nylander, W. 1857. Enumération générale des lichens, avec l’indication sommaire de leur distribution géographique. Mém. Soc. Sci. nat. Cherbourg 5: 85-146. — 1859. Prodromus expositionis lichenum Novae Caledoniae. Annis Sci. nat. (Bot.) IV, 14: 280-283. —— 1861. Expositio lichenum Novae Caledoniae. Annals Sci. nat. (Bot.) IV, 15: 37-54. —— 1870. Recognitio monographica Ramalinarum. Bull. Soc. Linn. Normandie I1, 4: 101-181. —— 1888. Lichenes Novae Zelandiae. Paris. Osorio, H. S. 1970a. Contribution to the lichen flora of Argentina IV. New or additional records. Bryologist 73: 392-394. 1970b. Lichenes from Cantera, South Paraguay. Comun. bot. Mus. Hist. nat. Montev. 4 (50): 1-3. — 1972. Contribution to the lichen flora of Uruguay VII. A preliminary catalogue. Commun. bot. Mus. Hist. nat. Montev. 4 (56); 1-46. 1978. Contribution to the lichen flora of Brazil IV. Lichens from southern Rio Grande do Sul. Bryologist 8: 452-454. , Aguiar, L. W. & Zanette, V. C. 1980. Contribution to the lichen flora of Brazil VII. Lichens from Montenegro and Triunfo, Rio Grande do Sul state. Comun. bot. Mus. Hist. nat. Montev. 4 (62): 1-8. Peveling, E. 1970. Die Darstellung der Oberflachenstrukturen von Flechten mit dem Raster- Elektronenmikroskop. Ber. dt. bot. Ges. II, 4: 89-101. —— 1974 [‘1973’]. Fine structure. Jn V. Ahmadjian & M. E. Hale (Eds), The lichens: 147-182. New York. Poelt, J. 1969. Ramalina. Bestimmungsschliissel Europdischer Flechten: 550-562. Lehre. — 1970. Das Konzept der Artenpaare bei den Flechten. Vortr. GesGeb. Bot. II, 4: 187—198. —— 1972. Die taxonomische Bedhandlung von Arten-paaren bei den Flechten. Bot. Notiser 125: 77-81. — 1974 [‘1973’]. Appendix A classification. In V. Ahmadjian & M. E. Hale (Eds), The lichens: 599-632. New York. Powell, C. McA., Johnson, B. D. & Veevers, J. J. 1981. The early Cretaceous breakup of eastern RAMALINA IN AUSTRALIA Zoe Gondwanaland, the separation of Australia and India, and their interaction with southeast Asia. In A. Keast (Ed.), Ecological biogeography of Australia: 17-29. Den Haag. Pryor, L. D. 1981. Australian endangered species: Eucalypts. Australian National Parks and Wildlife Service. Special publication [5]. Canberra. Rasanen, V. J. P. B. 1943. Das System der Flechten. Acta. bot. fenn. 33: 1-82. — 1949. Lichenes Novi V. Soumal. eldin-ja kaswit. Seur. van. Tiedon. Poytak. 3: 178-189. Raven, P. H. 1973. The evolution of Mediterranean floras. In F. de Castri & H. A. Mooney (Eds), Mediterranean type ecosystems, origin and structure: 213—223. Berlin. —— & Axelrod, D. I. 1974. Angiosperm biogeography and past continental movements. Ann. Mo. bot. Gdn 61: 539-673. Rogers, R. W. & Stevens, G. N. 1981. Lichens. Jn A. Keast (Ed.), Ecological biogeography of Australia: 593-603. Den Haag. Rundel, P. W. 1978a. Evolutionary relationships in the Ramalina usnea complex. Lichenologist 10: 141-156. — 1978b. Ecological relationships of desert fog zone lichens. Bryologist 81: 277-293. — 1982. The role of morphology in the water relations of desert lichens. J. Hattori bot. Lab. 53: 315-320. — & Bowler, P. A. 1974. The lichen genus Trichoramalina. Bryologist 77: 188-194. — —— 1976. Ramalina leptocarpha and R. subleptocarpha: a fertile-sorediate species pair. Bryologist 79: 364-369. —_— —— 1978. Niebla, a new generic name for the lichen genus Desmazieria (Ramalinaceae). Mycotaxon 4 (3): 497-499. Schuster, R. M. 1979. On the persistence and dispersal of transantarctic Hepaticae. Can. J. Bot. 57: 2179-2225. Sheard, J. W. 1978. The comparative ecology and distribution and within-species variation of the lichenized Ascomycetes Ramalina cuspidata and R. siliquosa in the British Isles. Can. J. Bot. 56: 939-952. Shirley, J. 1888. The lichen flora of Queensland, with descriptions of species. Part 1. Proc. R. Soc. Qd 5: 80-110. Sipman, H. J. M. 1983. A monograph of the lichen family Megalosporaceae. [Bibliotheca lichenologica 18.] Vaduz. Smith, A. L. 1921. Lichens. Cambridge. Specht, R. L. 1970. Vegetation. In G. W. Leeper (Ed.), The Australian environment: 44-67. 4th ed. Melbourne. —— 1981a. Major vegetation formations in Australia. In A. Keast (Ed.), Ecological biogeography of Australia: 163—297. Den Haag. —— 1981b. Biogeography of halophytic angiosperms (salt-marsh, mangrove and sea-grass). In A. Keast (Ed.), Ecological biogeography of Australia: 575-589. Den Haag. — 1981c. Evolution of the Australian flora: some generalizations. In A. Keast (Ed.), Ecological biogeography of Australia: 783-805. Den Haag. Steiner, J. 1904. Flechten, auf Madeira und den Kanaren gesammelt von J. Bornmiiller in dem Jahren 1900 und 1901. Ost. bot. Z. 54 (10): 351-365. Stevens, G. N. 1978. Lichens on mangroves along the east coast of Australia. M.Sc. thesis (unpublished), University of Queensland. — 1982. Ramalina leiodea (syn. R. boninensis) — a common maritime species in Oceania. Lichenologist 14: 39-45. — 1983a. Tropical-subtropical Ramalinae in the Ramalina farinacea complex. Lichenologist 15: 213-230. — 1983b. Clarification of the name Ramalina linearis. Lichenologist 15: 99-102. Stizenberger, E. 1862. Beitrag zur Flechtensystematik. Ber. Tat. St Gall. naturwiss. Ges. 3: 124-182. Tehler, A. 1983. The genera Dirina and Roccellina (Rocellaceae). Op. bot. Soc. bot. Lund. 70: 1-86. Vainio, E. A. 1890. Etude sur la classification naturelle et la morphologie des lichens du Brésil. Pars prima (1-29). Acta Soc. Fauna Flora fenn. 7 (1): 1-247. Pars secunda. Acta Soc. Fauna Flora fenn. 7 (2): 1-256. Watson, W. 1929. The classification of lichens. Part 1. New Phytol. 28: 1-36. Webb, L. J. & Tracey, J. G. 1967. An ecological guide to new planting areas and site potential for Hoop Pine. Aust. For. 31 (3): 224-239. Weber, W. A. & Wetmore, C. M. 1972. Catalogue of the lichens of Australia exclusive of Tasmania. Beih. Nova Hedwigia 41: 1-137. Wetmore, C. M. 1963. Catalogue of the lichens of Tasmania. Revue bryol. lichén. 32: 223-264. eee G. NELL STEVENS Wilson, F. R. M. 1889. Australian lichenology. Proc. R. Soc. Od 6: 549-553. Zahlbruckner, A. 1907. Lichenes (Flechten) B. Specieller Teil. In A. Engler & K. Prantl (Eds), Die Natiirlichen Pflanzenfamilien 1: 49-249. Leipzig. Index Accepted names are in roman and synonyms in italic; new names and principal references are in bold. An asterisk (*) denotes a figure. Alectoria 116, 118, 119, 147 anceps 147, 160, 217 angulosa 217 arabum 116 attenuata 158 australiensis 119, 124, 127, 129, 138, 143, 149, 150, 151, 152, 153% lod 9521482217, Bistortae 116 Bitectae 116 boninensis 207 f. subcalicariformis 207 bourgaeana 129 brevis 120, 163, 166, 214* var. brevissima 120, 193, 194 Bulbothrix 146 caespitella 121, 123*, 129, 138, 139, 143, 149, 150, 153*, 155, 156* calcarata 146, 186, 194 calicaris 217 f. Ecklonii 207 var. australica 120, 163, 166, 214* var. canaliculata 217 var. ovalis 184 canariensis 121, 123*, 124, 129, 132, 138, 139, 143, 149, 150, 151, 153*, 156, 157*, 178 capitata 121, 123*, 130, 179 carpathica 115 celastri (Parmelia) 180 celastri (Ramalina) complex 124, 129 subsp. celastri 128*, 129, 131*, 136, 138, 139, 143, 149, 150, 151, 171, 180, 181*, 182, 183*, 184, 216*, 217 subsp. ovalis 108, 129, 131*, 138, 139, 143, 149, 150, 152, 181*, 183*, 184, 217 Cenozosia 115, 116, 117, 118, 119 ceruchis 115, 116 chondrina 160, 161 Ciliate 116 combeiodes 115 complanata 115, 126, 217 Compressiusculae 115, 116 confirmata 119 consanguinea 146, 189 continentalis 182 Corticatae 116 cumanensis 182 curnowii 130 decipiens 129 dendriscoides 176, 217 var. minor 217 Desmazieria 115, 116, 117, 119 Dievernia 117, 119, 129 dilacerata var. alba 217 Dirina 147 Dirinaria 148 dumeticola 197 duriaei 129 ecklonii 180, 182, 184, 207, 217 var. lobulifera 182 var. membranaceae 182, 217 var. ovalis 184, 217 var. tenuissima 182, 217 Ecorticatae 116 Ellipsosporae 116 Euramalina 115, 116 evernioides 116 exiguella 119, 120, 128*, 129, 131, 136, 138, 143, 149, 151, 153*, 154, 158, 159*, 177, 214* 217 farinacea 115, 124, 126, 129, 130*, 132, 134, 147, 197, 198, 200, 201 var. dendroides 120, 170, 172, 25" var. nervulosa 120, 195, 197, 200 var. squarrosa 120, 170, 172, Z5* fastigiata 115, 217 fastigiatus (Lichen) 117 fecunda 210 filicaulis 108, 119, 124, 129, 130*, 136, 138, 143, 149, 151, 153*, 154, 160, 161, 217 fimbriata 119, 121, 122*, 123*, 129, 132, 138, 139, 140, 141, 143, 149; 150:,152,153*2161 1627: 178 fissa 131, 132, 138, 143, 149, 150, 1515152, 1975 1815) 185; 191" 193, 194, 195*, 215* Fistularia Vainio 115, 116, 126 Fistulariae R. Howe 116 Fistulariella Bowler & Rundel 110, 115,116,175 119 Fistulosae Hue 115 flaccescens 115 fraxinea 115, 150, 203, 217 B yemensis 180, 182 f. platyna 194, 217 subsp. confirmata 205 var. ampliata 217 var. brevissima 193 var. taeniata 217 furcellata 217 Fusisporae 116 geniculata 117, 185, 186, 194, 217 var. compacta 120, 188 var. Olivacea 217 glaucescens 119, 129, 131, 133, 138, 139, 141*, 143, 149, 152, 163, 164*, 165, 166*, 167, 168, PAC SERN Wel gracilentia 217 gracilis 119, 158, 217 homalea 115, 116, 217 implectens 147 inanis 115, 116 inflata (Cetraria) 185, 215* inflata (Fistulariella) 117, 119, 185, 187, 194 inflata (Ramalina) complex 129, 130*, 133, 139, 191, 192 subsp. australis 108, 138, 139, 1435149; 150; 151,152,181"; 186*, 191, 193, 194, 217 subsp. inflata 128*, 130, 131, 138, 139, 143, 146, 149, 150, 152, 181*, 183, 185, 186*, 187, 217 subsp. perpusilla 108, 128*, 130*, 136, 138, 143, 146, 149, 150, 152, 159, 181*, 183, 186*, 187, 188, 190, 217 var. fissa 120, 193 var. gracilis 217 insularum 201 intermedia 217 intermediella 172, 200, 201 javanica 217 knightiana 120, 185 lacerata 120, 163, 214* laevigata 182 leiodea 117, 148, 203, 207, 216* var. fastigiatula 119, 163, 165, 167, 214* leptocarpha 147 linearis 115, 182, 271 litorea 128*, 129, 136, 138, 143, 149, 151, 153*, 169, 170* luciae 199 maritima 207 mediterranea 191 melanothrix 115, 158 microspora 169 minuscula var. alba 120, 188, 216* Myelopoea 115, 116 Myelopoeae 116 myrioclada 120, 152, 154, 214* nervulosa var. dumeticola 121, 122", 12575124, 134,130,138, 143, 149, 151, 196, 197, 198, 199*, 201 var. luciae 121, 124, 125, 134, 136, 138, 143, 149, 151, 177, 196, 199* , 200, 201, 217 var. nervulosa 121, 124, 125*, 129, 136, 138, 143, 149, 151, 177, 195, 196, 198*, 200, 215* Niebla 115, 116, 117, 119, 129 ovalis 184, 216* pacifica 121, 124,125", 127", 129; 132, 136, 138, 143, 149, 151, 159, 177, 196, 200, 201, 202*, 21S? S207 Parmelia sect. Hypotrachyna 147 Parmelina 146 perpusilla 120, 188, 215* pertenuis 158 peruviana 115, 120, 121, 122*, 124, 125, 129, 131, 136, 138, 143, 149, 151, 154, 170, 171, 172, 173*, 174, 176, 198, 200, 217 Physcia 147 pollinaria 115, 161, 178, 217 polymorpha 115, 217 f. emplecta 217 populina 130 Pseudocyphellaria 146 Pseudoparmelia 146 Psoroma 146 RAMALINA IN AUSTRALIA pumila 189 pusilla 186, 191, 217 Ramalina 109*, 115, 149 Ramalinaceae 115, 116, 117, 118, 119 Ramalinopsis 117, 119 reductai]21--122*. 123*-129) 1132" 138, 139, 141*, 143, 149, 151, 177575176; 183 Relicina 146 rigida 158 scopulorum 115, 217 var. cuspidata 217 var. subfarinacea 217 sideriza 210 siliquosa 133, 134 sinaloensis 126 Solidae 116 Speerschneidera 117, 119 sprengelii 182 Sterecaulon 146 strepsilis 130, 179 subfraxinea 203, 204, 207, 216 complex 129, 134, 146, 148, 208, 210; 217 subsp. leiodea 207 var. confirmata 108, 125*, 127*, 136, 138, 143, 149, 152, 159, 176, 196, 202, 205, 206*, 208, 211, 213, 216" 27. var. leiodea 108, 127*, 136, 138, 143, 149, 152, 159, 196, 203, 205,.206% , 2072 2115.213 var. norstictica 108, 127*, 136, 138, 143, 149, 151, 196, 208, LiL) 209* > 211,213 var. subfraxinea 125*, 127*, 136, 138, 143, 148, 149, 152, 196, 203, 204* , 207, 208, 211, 212, 213; 21687217 subgeniculata Knight 120, 185 subleptocarpha 147 subpusilla 194 tasmanica 117, 185, 215* tayloriana 217 tenella 121, 122*, 123*, 129, 132, 136, 138, 143, 149, 151, 171, L727 On Neel, Tenuicarticatae 116 Tenuicorticate 116 Teretiusculae 115, 116 testudinaria 115 Trichoramalina 115, 117, 119 tropica 108, 136, 138, 143, 149, 150, 151,209; 210;.211:,212** 213 Tubulosae 116 unilateralis 120, 121, 122*, 123*, 124, 129, 132, 138, 139, 141*, 14351495 151156/1575 1618 17 ATT Ste Ome leh usnea 108, 147, 154, 160, 217 Usnea 116, 118, 119, 146 Usneaceae 117, 118, 119 usneoides 154, 217 whinrayi 130*, 132, 133, 138, 139, 143, 149, 151, 171, 179, 180* yemensis 180, 182, 184, 217 var. ovalis 184, 217 zollingeri 210 _ hy 7 - — 7 i a a we 7 7 . {* . 0 f a, 7 a 7 a a eee ~~ 7 ee foto ar 4a Se OT ; 7 Wy oe : a AA” = Aol a: - : 7 iin i ype 7 7 co - * 7 > a - a : 7 : i - a — - i a Gy — ; 7 a: a a 7 > . 7 7 - a 7 = o an i‘ me 7 : : oe - > we M~d ee = 7 Ss Se ae -2, 1) -= oa i = = 7 - ; _ > eae 2 ae), a 7 eo —_ : a 7 nia Tay > = | er ae _ a ae : ee _ |) (CA ae Seer a ee oP. cas O20 A ee aa ae : 7 ~“ . 7 oo i. ce oar PL me - RS a ee ee >a = oc on ts. Si eee De a 4 a, i _ a a 7 a” yp Ee ——* eo 7 2 eo aera wr 4 ne 7 — as _ 7 °h ™- : —. a : - - z ar Cony ay a) ra! ar | Pee, a a ha a ee = Le os _ Z ae a 7 : a ae ee << - - L 7 a ake aigedl’ Sheree. 7 ae Dees andere. ic : ee ew a sy “Sere ih a 7 i ay a 2 a : a ne St. oe a Wo ee ee Rie 2 ee «== etizetem's «= DS. ee > ae | eee 1. 7 ie mG — ee % if - A —_ -* 0 ir ee ee ee ee Le oh “ree ae _ eS ele : = a a aie ae, Oh PES gs Soe ea Stee Sere oe eee > “ae, Der ya op wn Cl) ee eee Ay ah Sa ee ee ane - oe ae : ae - dieters i. ae aro one, ae Aaa : = 7 Re ee ee ee 8a [oa Meee es : = ay wes ae SS oy = al ae ari h 7 == ne a. _ 7 > by - a 7 , ® a > a ja : i Da ae ie San Y 7 aan => - ee aD, & 7 i: _ 7 == S 7s — sc. WJ ee ti _ or es an we oe 5 aaa we -y) ay : a oS - i 4 . im ae = vo rn —- . dnl ae. 7 7 a nae? 7 ad Sey 7 a -_ 7 S 7 : ne -_ we. - en a. ot os Loe I< rer fr, ay Gee a 7 : iad n - 7 x - : en eae 7 ioe oe ss ==» = ee ar iia & a, | = 7 , 7 _ _ ee ee oe ee” | lo ee Laem - a - in 7 : — 5 - : es : 7 . el aaa se, ee) > ee ee oe =e Sa 2 he Ta i 7 - = = —_ ro - 7 7 a Geakke!|™ Qe Sh 6 ga Fe ee a . 7 a. - Wrelares 7 ee Soe ; — - — teow i Pee 2 a ek nn I Te le : ~ - an - - - : \ “4 . Oo Oman a be aad o Xe = 2 oa = a iro Mf : - il 7 : e : - 7 + ls 7 - _ i ST 7 a ae a oe, a ote PT he al a : : _— 7 7 oe 7 i a _ a. : a a ; 7 ee a i ee a ce P a 5 aes : oO — a ~- 1 * _ x = 7 ei ee | - 7 an - ay > >; 1 - 7 a 7 a - eS = . a 7 = > Aa oe : a: a a CALE. . , Se " = 7 7 a ’ a om 7 a ae | ae = ae: ; ae ; -_)> = ‘Se Go) Gavan, . ; AY- aye - : : at f : 7 De ie : | ton = oe a. _ : —_— * OS “a 7 = 7 7 - nes aay G = _ 7 - 7 - a i a — 7 : aan aa 2 rat — _— = = 9s ee ae — > = - 7 a _ 4 7 7 5 7 : - - a = _ 2 _ _ 7 7 - 7 — - : a ov = = > oe 7 7 oo or Oy Ge. Wa Ae o = _ i. + : 7 7 7 ” —. Oe Ge oP Sp ey SESS a) eS.) * : a ols a ae : a: ; - cc av i aon = a et i ol i. oe - Bes oe 7 7 ~ 7 —_ : : a 7 a - 7 ie ef 7 i _ oe - 77 4 ee : — 7 a 7 _—- = Waa Sa 7 a 7 -_ a 7 7 7 - 7 7 ee es iodo ae oo : — 2 y oe: - ie eg : ee ee ours s | 7 a ne 2 7 _ 7 fr. 7 — oe - = ©oG _ er nr, > i a rr ce ee eae : 7 : . es > PE oll pen — See ee ev Se aw oe a ail : — oe haa 7 - - 4 7 im - - : a oe 7 a : ae ae Caen & _ = _ 2 oa _ Derr Cast” 7 Say a a 7 7 7 , r a _ a S | io py = a - = ae = : oor | eo a _ wt a. : a : a ‘ 7 7 ” - - : ne a ee ie aor ey — 7 nk a ae 7 | Qa as 3 ae = | +i ; _ as : : a= : Sy = = ' sm oS ot Te oe a 7 : - - S 7 A of - Ba - Ue | _ oa Pe. an, y= -_ - ee ee ee - : 7 = — 7 i _ 7 = ao = - - a - vs ; i a, - 7 : =~ ins - 7 _ i == 0 7 ‘ Ota. 7 San a7 _ - a " aa : a : 7 Sap _ = az : 7 e 7 a : 7 an a ao eo = ; are a - 7 bre 2 oy ——_— - Sy + ss” = lors _ a —_ ; _ — : ; : Oe ee 2 — | 2 = — 7 - - - = - - " a a - - 7 — SS > ae a — = ant = - 7 : 7 a = 7 7 7 7 7 as : : 7 - - —, ae : : ae. * oa re 7 =e a = ee: a: a 7 7 > atin : : ; el nn ie i a | en oe - - 7 Co = . - = 7 = : - a : : a an ; : - ee ; ; Pees + ae : ; _ : ae a "Ms a a - : 7 = =~ vn a - : - 7 a 7 7 <> 7 a : 7 : 7 aed ae 7 7 a = io 7 = a a - = : a 7 7s Oe : : a a — 7 7 7 _ a 5 - - 7: : - : ; - : 7 : 7 ’ - - 7 = ; - 7 wi on = 0-2 7 - - 7 - i a , _ cS 7 : , Se oe ; : rT & — ows 4 i - _ - 7 i ‘ British Museum (Natural History) Ferns of Jamaica A guide to the Pteridophytes G.R. Proctor This flora records and describes the 579 species and 30 varieties of ferns occurring in Jamaica. The succinct species descriptions include relevant synonymy and incorpor- ate distributional data both within and outside Jamaica. Special emphasis is given to the subtle distinctions between closely related species and all genera are illustrated. Keys to the genera and species facilitate a wider use of the flora in the West Indies and northern South America. The author, one time Senior Botanist in charge of the Herbarium of the Science Museum, Kingston, Jamaica, is an outstanding field botanist and his expertise is reflected in the practicality of the flora and especially in the habitat and ecological information. This volume represents an important addition to our knowledge of the flora of the West Indies. 1985, 631pp, 135 line illustrations, 22 maps. Hardback. 0 565 00895 1 £50.00 Titles to be published in Volume 16 Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) By N. K. B. Robson The lichen genus Ramalina in Australia By G.N. Stevens An annotated list of vascular plants collected in the valleys south of Mt Everest By G. Miehe Title to be announced Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk Printed in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) An annotated list of vascular plants collected in the valleys south of Mt Everest Georg Miehe _ Botany series Vol 16 No 3 27 August 1987 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum’s resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, available separately, and individually priced. Volumes contain about 300 pages and several volumes may appear within a calendar year. Subscriptions may be placed for one or more of the series on either an Annual or Per Volume basis. Prices vary according to the contents of the individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.) © British Museum (Natural History), 1987 The Botany series is edited in the Museum’s Department of Botany Keeper of Botany: Mr J. F. M. Cannon Editor of Bulletin: Mr J. R. Laundon Assistant Editor: Dr A. J. Harrington Editor’s Assistant: Miss M. J. Short eon 8 . ‘ ge 7 ons euellCATIONS ISSUED 27 AUG 1987 ISBN 0 565 08016 4 Botany series ISSN 0068-2292 Vol 16 No 3 pp 225-268 a NEZ British Museum (Natural History)*3~_ Cromwell Road ~ London SW7 5BD Issued 27 August 1987 An annotated list of vascular plants collected in the valleys south of Mt Everest Georg Miehe , Department of Geography, University of Géttingen, Goldschmidtstrasse 5, D-3400 Géttingen, West Germany Contents ST + ype Heed period le gih ng ee eA a gy POURS eT ROME a Phe Be ey uae arts ee MeN een Tre ere Zen REN R NN ois es cases ee tk nhs and Peto sin aieon Ce anaseace secs NT wee cuk ee teRr ay eae a Mana Tae 225 MONT ETIMEREEI STN OOE TAMRs RI css scd fo eset Sos atsa desu cinch vndschees cbnee vast natecenatecreleets 225 RPONETAMS OE CHIE IIE 65s soi sivn'c ods tides viv dhe sacgeie vaste ee kent tau hale ai ncocescan na seen cness 226 PRMEERAMNI GL CTC Clk NE MORMCUOND 60's ca taick tps bce ceniagadee hbeesrsaeesswads vndeseseevcd Sueuckaeveres 228 PEUATES On ee TE CONOR a. cil os Eas hag nasi Rs inc ee ee dig date steed tenabese ys 231 Remarks on the influence of grazing, burning, re ee 1 Pa ARNO IR ni tye IEE PRR OE Der PERE Tas Oe te LMBE on ey ee 232 Alphabetical list of vascular plants, their life forms, associated formations, and records of collection ..................cccceccecceeeeeeeeeeess 233 BY CE DISHES LONE BVIIDONS oc 2isc nies on cacenan ccna can dOd toe cveidedanst nen sisdencaseenetees 233 BG Uy ae TCR INN SYM oes 5c 65 con sing saicase sid Sad eeneinssacrtesesesysasstnuontecdeen 234 PROCS VIMNIOR TST OL CONGCIORS’ NAICS ii. isc: Ji cos disses soades eeetacssasaennentacinba gases deans Za PASM aDEtUCAl TISE OL HOWETINE DIAING i cc5ics cass 9105340 c00- 7s ogee ksdosvs optete 'yachung Kang @ main settlement a N me © summer settlement —)— Cols and passes (se 4 ) Glacier 0 1 28°N \ Mt. Everest : a £8 ee \ ° 8848 Source: ; Khumbu Area Nepal, Mountain Hazards Index Map; Nepal 1:506.880 East Sheet i Chhu SAVeGo THY 86°45’ 87° 87°15" Fig. 1 The valleys south of Mt. Everest. Hinku Drangka (Inugkhu), Hunku Drangka (Hongu), and Iswa Khola run to the south, to join the Arun valley in the east or the Dudh Kosi in the west. Only a few botanists have visited these valleys and their records all belong to the alpine belt and the upper cloud forests. The vast majority of records are from along the route to Mt Everest. In this area, where the Dudh Kosi valley leads down from the main range between Cho Oyu in the west and Lhotse in the east, the southern limit of the records assembled here is drawn at 2300 m, south of Chaunrikharka. The Dudh Kosi gorge south of Namche Bazar is known as Pharak, while the higher catchment of the Dudh Kosi north of 27° 27'N is called Khumbu. The area covered is rather heterogeneous in its floristic pattern. The south-eastern valleys, the Barun Khola and the Iswa Khola, are likely to prove much more humid (like the east Himalayas) than the Khumbu Himal. The latter was classified by Stainton (1972) as belonging to the ‘inner valleys’, as it shows a degree of rain-shadow effect. The other component contributing to the heterogeneity is man-made, because the valleys of the Khumbu Himal have been cleared to a much greater extent than is typical for the upper cloud forest belt of the southern slope of the main Himalayan range. The more arid and continental aspect may be due in part to the presence of ruderals and weeds, thus reflecting the influence of man. Outline of the climate Climatic data in mountain areas are usually inadequate. All meteorological stations are typically situated near settlements, and these are mostly located on valley floors or on less cloudy sites. Thus, for most mountain areas, the data available give the minimum possible rainfall, and are valid only for that small strip on the valley bottom, covering less than 5% of the area. The data available are summarized in the climatic diagram (Figs 2—4) after Gaussen as drawn by Walter & Lieth (1966), which takes the temperature line as an approach to evapo- transpiration. The relation between temperature and rainfall is simplified as 10°C = 20 mm, taken from the mean temperature and the mean monthly rainfall, which is reduced to one-tenth PLANTS SOUTH OF MT EVEREST 227 Chaunrikharka (2743m) 2283,6mm (mean monthly temperatures of Namche Bazar +0,58°C / 100! /1949-1962) °c Namche Bazar (3440m) ca. 6,6° 1048 mm 80 = (1948-1963, 1953?-1976) - 120 bs mm 50 100 50 - | L 100 40 4 L 80 40 - + 80 30 + + 60 30 4 + 60 20 4 L 40 20 4 + 40 > aa 10 4 L 20 10 - ~ N L 20 -7,9° 0 T T T D T T T T T T a7 0 0 al T T T T Ai T T ql a T T 7 0 Ce VE NY Wel OC exe 00 ON Oe i a i Vi VO ee OE Months Source: Dhar & Narayanan 1965, Joshi 1981 Months Source: 0. N. Dhar, J. Narayanan 1965, Fi 2 D P. Josh 1981 'g. Fig. 3 ‘Lhajung’ (4420m) -1,27°C (27°53,6’ N/86°49'7” E.Gr.) 528mm (1974, 1974) ¢ mm ne” 5° - 100 40 + + 80 30 - + 60 20 4 L 49 10 4 L 20 LU N i -113° L : “Mr yp on we vw we vm x x Nox Months ho Source: K. Higuchi (Ed.) 1984, App. 3 Fig. 4 Figs 2-4 Climatic diagrams. After Walter & Lieth (1966). for graphical reasons, if more than 100 mm. Where the temperature line lies above the rainfall line it is relatively arid, whereas in the opposite case it is relatively humid. From the available data (Dhar & Narayanan, 1965) it is probable that the precipitation decreases from the periphery of the mountain area to the centre, roughly by 50% for every 12 km, but this decrease is also combined with a change in altitude. Thus the decrease is nearly uniform every 1000 m between 2400 and 4400 m. Since the temperature is decreasing also, it does not become drier. 228 G. MIEHE Only in the medium cloud forest belt, at Chaunrikharka is there a dry season in winter, although the precipitation is more than double that at Namche Bazar. At Chaunrikharka 86% of the precipitation is concentrated in the monsoon period, whereas Namche Bazar receives only 74% summer rain. At Namche Bazar, situated in the upper cloud forest belt at 3440 m, relatively dry months may still occur independently of the overall annual rainfall (1961: 921 mm, with no dry month; 1953: 1710 mm, with March and April dry). The number of rainy days per month is higher in Namche Bazar, except in June and July, but fewer millimetres per day fall than in Chaunrikharka. The distribution of heavy precipitation, especially snowfall, is of some econ- omic importance. At Namche Bazar, the months of heavy 24 hour to 72 hour precipitation are those of sowing and harvesting, thus especially before the introduction of the potato (Firer- Haimendorf, 1964: 8-11), the harvest, which was then of buckwheat and barley, was always uncertain. Climatic information is also provided by the vegetation, even though we have no exact calibration. Closed plant communities, the presence of epiphytes (especially obligate species that only grow on trees or rocks), and a dense lichen cover can be interpreted as an indicator for humid high-mountain conditions. Temperature data have rarely been published. For Namche Bazar (3440 m), the mean temperature of the coldest month (January) is —0-4°C, while the mean temperature of the hottest month (July) is 12°C (Joshi, 1982: 400). In the alpine belt at 4420 m, the mean temperature in January is —7-7°C and that of July is nearly 10°C, over a four year period (Higuchi, 1984: appendix). Valley winds blow every day, setting in mostly between 10 and 11 a.m. and dropping before midnight. They are not as strong as in the transverse gorges like the Karnali or the Kali Gandaki in west and west-central Nepal, but both areas show the same type of clouds, which cover the upper slopes of both sides of the valleys with strips, while the sky over the valley bottoms remains blue. In monsoon and late monsoon conditions, a closed cloud cover from the south often sweeps into the valleys, nearly as far as closed forests occur, but above 4200 m only strips of cloud on each side of the main valleys advance further up. The valley winds obviously effect the wind-exposed vegetation strongly, as indicated by treeless areas and wind-sculptured trees. Altitudinal belts of vegetation The valleys south of Mt Everest are mostly of the same type as those of the European Alps, being typical of high mountains with adequate rainfall to the north of the tropics. Therefore altitudinal limits should be expected to be governed by temperature, decreasing at the rate of 5-8°C per 1000 m and not caused by seasonal drought. This humid type of high mountain is indicated by the interlacing of the belt of glaciers down into the free solifluction belt and down into the alpine belt. The most striking demonstration of that classification is that the free solifiuction belt is squeezed into a narrow zone between the glaciers and the vegetation cover of the alpine belt. Thus the upper limit of the characteristic semi-arid altitudinal zone of the high mountains is covered with ice, while the potential lower limit of the solifluction process is suppressed by the closed vegetation cover of forests and alpine scrub (Kuhle, 1978). All belts are interlocked both through natural factors and man’s activities. The forest belt is widely cleared and replaced by predominantly alpine associations, and 200 m above the present upper forest line, isolated trees are found growing in alpine scrub. The alpine belt can be roughly divided into a lower zone of moist alpine scrub (dwarf rhododendrons and prostrate junipers) and an upper zone with Kobresia pygmaea mats. Nevertheless, in wind-exposed situations, the Cyperaceae dominated pastures extend down to 3000 m, which is at the maximum about 800 m below the potential upper tree-line, and possibly the result of initial clearance by fires. Dwarf Rhododendron nivale cushions are found near the upper limit of the alpine belt at 5500 m. Hence the altitudinal belts are only differentiated by the dominance and not the exclusiveness of their characteristic features. The lowest altitudinal belt is the forest belt which extends up to roughly 4200 m. It is characterized by epiphytes and should, therefore, be classified as cloud forest. There are both PLANTS SOUTH OF MT EVEREST 229 obligate and facultative epiphytes and the latter can be regarded as indicators of a humid climate: creeping ericaceous shrubs are confined to moist soil and are only found on moss- covered tree trunks when this substratum is almost constantly damp through rain and fog. Some of the cloud forest epiphytes even invade the alpine belt, growing in crevices surrounded by cushion-forming alpines, and certain species, mostly found around flushes (Kobresia schoenoides, Rhodiola spp.), sometimes even become chasmophytic. Mosses are the dominant epiphytes of the forest belt up to 3000 m. They cover trunks and branches and are colonized by small phanerophytes as facultative epiphytes. Epiphytic ferns and beard lichens occur, but are not dominant. The minimum rainfall in this forest belt is unlikely to be less than 1000 mm per year. This medium cloud forest belt is roughly identical with the Tsuga dumosa forest and the upper temperate mixed broad-leaved forest described by Stainton (1972: 1-11, 98-100). Above 3000 m the upper cloud forest of these transverse valleys is somewhat differentiated with regard to dominant tree species depending on the exposure. This is possibly caused by decreasing rainfall, which is here probably less than 1000 mm. Forest of Abies spectabilis still occurs on south-facing slopes (more those to the south-west than the south-east) and on wind exposed ridges on northerly slopes, whereas Betula utilis, which quite often forms an understory in Abies forests, becomes dominant on northerly slopes. Rhododendron thickets form the shrub layer in fir and birch forests and also, above the upper limit of Abies and Betula, dominate to form pure evergreen thickets ranging from four metres down to one metre near the timber-line, representing a transition zone from forest to alpine scrub. In this upper cloud forest belt, branches may still be covered with mosses and ferns, but beard lichens are more frequent than in the medium cloud forest belt. On shady slopes, Usnea longissima is apparently more widespread and longer than on sunny slopes; branches growing out into the valleys are more densely covered with beard lichens. Whereas birch forests and the remnants of fir ferests are concentrated on the shady and sunny slopes respectively, Rhododendron thickets are apparently limited to those slopes which are well protected from the wind, but are not‘confined to any particular exposure to the sun. The alpine belt extends from about 3400 m (owing to forest clearance) up to a maximum altitude of 5500 m and is characterized by an almost closed layer of dark alpine soil. The most widespread plant associations between 4100 and 5500 m are dwarf shrub communi- ties of Rhododendron setosum, R. anthopogon, and R. nivale, with mats of Kobresia pygmaea. Wind-blown and south-facing slopes up to 5000 m are dotted with dwarf prostrate junipers. The lower limit of the alpine belt is relatively clear cut on the shady slopes protected from the wind, where Rhododendron campylocarpum and R. fulgens mark the upper limit of the forest belt at about 4200 m. The sunny slopes are mostly cleared, and between 3400 and 4400 m forest and alpine species are intermixed to form the winter grazing grounds. Thus there is a transition zone of nearly 1000 m which could possibly be divided into village grazing ground and alpine pastures by the presence or absence of prostrate Cotoneaster spp. which are abundant around the villages. : The potential temperature limitation of the alpine belt is thought to be at about 5500 m. For the most part this is limited by other factors such as late ice melting or constant falling of frost debris from upper slopes. In the upper part of the alpine belt, the cliffs and boulders of white granite which have been free from glaciers for long enough are covered with a black carpet of lichens. Clearly the growth of lichens, even though fairly slow, occurs at a faster rate than the creation of fresh rock faces by frost action. It can therefore be assumed that with these rocks at this altitude, the process of frost cracking is negligible, a controversial conclusion because this belt has always been characterized by the dominance of periglacial processes. In this upper alpine belt, three herbaceous communities can be distinguished on the alpine soil. The first occurs roughly between 4600 and 5000 m on sunny slopes exposed to the wind. Kobresia pygmaea provides about 10-30% of the plant cover, but more than 30% is covered by flat cushion-forming species such as Anaphalis cavei, Leontopodium monocephalum, L. brachyactis, and Arenaria bryophylla. In the second community, which occurs above 5000 m, Kobresia pygmaea covers only 5% and the area is dominated by hemispherical cushions such as 230 G. MIEHE Arenaria polytrichoides, by lichens covering the dead remains of plants (one third of which are usually dead), and the dark humid soil. The third community consists of nearly 100% cover of living Kobresia pygmaea mats, and is confined to depressions between the valley slopes and the lateral moraines of the glaciers. These depressions are believed to be younger than the other biotopes on the evidence of the depth of the alpine soil and the diameter of the lichen Rhizocarpon geographicum. The problem for all these communities is that the alpine soil was obviously established by Kobresia pygmaea, but today this member of the Cyperaceae has nearly disappeared except in the depressions along the lateral moraines. Further research is needed to decide whether this development has coincided with a climatic change to drier or colder conditions, or if it is simply a feature of the succession that at a certain age of alpine soil cushion plants become dominant. The lower alpine belt is typically covered by moist alpine scrub (as described by Stainton, 1972: 128). Cushions of Rhododendron anthopogon, R. setosum, R. nivale, and junipers with prostrate growth are characteristic of this zone and possibly represent the climax vegetation, whereas Lonicera spp., Spiraea arcuata, Hippophae tibetana, Potentilla fruticosa var rigida, and shrubby willows are confined to moist ravines or are part of a plant succession on moraines. The dwarf prostrate junipers seem to prefer sunny, wind-exposed slopes, but are more frequent on the shady lee slopes when Rhododendron setosum and R. anthopogon are absent. It thus seems quite reasonable to suggest that the junipers are restricted by competition to the sunny, wind-exposed slopes. The cushion-forming rhododendrons show a change in biotope as is explained by the ecological law of ‘Relative Habitat Constancy and Changing Biotope’ (Walter & Walter, 1953). At 3480 m and probably 900 mm annual precipitation, R. setosum occurs on sunny, wind- exposed slopes; at 4400 m its evergreen cushions are confined to wind-blown but fiat sites, whereas sunny wind-exposed slopes are free from R. setosum and only R. nivale covers the steep parts of the solifluction terraces, while R. setosum is restricted to the shady lee slopes where the precipitation amounts to about 500 mm per year. Obviously these rhododendrons change biotopes into the ecological niche of a wind-sheltered slope, where the extra-zonal advantage compensates for the zonal disadvantages. From eco-physiological experiments carried out by Larcher (1963) with R. ferrugineum, it is fairly certain that the Himalayan dwarf evergreen rhododendrons are highly sensitive to frost-drought, exacerbated by wind. Thus the daily valley winds provide the limiting factor for the distribution of dwarf rhododendrons in the alpine belt, but only do so in association with temperatures below 0°C and in the absence of snow protection. During the moist summer months the rhododendrons are quite exposed to winds and form wind-deformed cushions but in winter they are protected in low ravines and other wind sheltered depressions, hence, in the alpine belt of the Khumbu Himal the distribution of dwarf rho- dodendron cushions, prostrate junipers, and mats of Kobresia pygmaea depends on exposure to the wind, as determined by micro-relief. Thus vegetation patterns in the alpine belt of these subtropical mountains at 28°N are more similar to those found in the Alps at 47°N than would be expected. The free solifiluction belt above the limit of plant cover on alpine soil extends to the maximum altitude of thawing (Kuhle, 1978) which is at about 7100 m, according to a record of a Lecidea sp. on the south wall of Mt Makalu (4th Jugoslavian Makalu Climbing expedition, pers. comm. J. Poelt) which must have been dependent on water. The highest records of flowering plants (species unknown) are from c. 6300 m on Cho Oyu (H. Heuberger, pers. comm.) and at 6350 m in the western Cwm between camps 4 and 5 by Zimmermann and Hofstetter (Zimmermann, 1954—1956) on the south side of Mt Everest. Both records refer to small cushions in crevices. This biotope provides conditions that are not typical for these altitudes, because daily movement of the substratum by thawing and freezing during the vegetative period is the most decisive limiting factor for plant colonization, but these processes are minimal in crevices densely stuffed with cushions. At the highest altitudes that I was able to find plants (5690 m) it was Saussurea simpsoniana that was growing in dense cushions in these crevices. Here competition is often very strong, in contrast to the surrounding free solifluction areas which are almost devoid of plants. The rather excessive movement of the substratum is only tolerated by a few specialized plants, PLANTS SOUTH OF MT EVEREST mot which have shoots that lengthen with the downward creeping debris. Probably the shoot length of one growing season is approximately equivalent to the distance covered by the debris. Plants adapted to this unstable ground might be called ‘solifluction acrobats’ and include Gentiana urnula, Eriophytum wallichii, and Veronica lanuginosa. All other plants are confined to small flat patches protected from the stronger debris movement. There may occur here a nearly closed cover of plants, small cushions of crucifers, loose flat cushions of Stellaria decumbens, mostly with one half dead, and grasses with huge, dense root systems. Typically, their innovation buds are not at the surface, as would be expected for hemicryptophytes, but have sunk into the debris which has already moved on after the end of the growing period and are thus well protected during winter. The flora of the free solifluction belt is quite distinct from the one that occurs in patches of debris, ravines, or on gravel along streams in the alpine belt, and only a small number of species occur on both sides of the main Himalayan range. Remarks on timber-line ecology The striking feature of forest distribution in the valleys to the south of Mt Everest, mainly in the Khumbu Himal, is that forests predominate on the shady slopes, while the sunny slopes are covered with open woodland or grazed open pastures. In the same way, the observation that on flat ground, or on one side of ravines, forests grow higher up than on the other slope, suggests that open sunny slopes may not be favourable for forests and that this might be the result of drought. Yet one can always find solitary trees or groves on the sunny slopes. If these trees or groves do not depend on special tectonic conditions such as fault springs, which can be ruled out if the herb layer does not indicate moist conditions, one would expect woodlands or forests to occur where the topoclimatic conditions are tolerable for a single tree. The conclusion must therefore be that single trees or groves on sunny slopes are forest remnants that have survived clearances. As all these remnants are strongly influenced by grazing we do not know their natural state, but the presence of Abies spectabilis, Juniperus recurva, and Rhododendron spp., partly covered with Usnea longissima, suggests that cloud forests may be the potential vegetation, as on the shady slopes. If these conclusions are admitted, it follows that the highest altitudinal records of trees show distinct differences on different slopes. The highest altitude at which I encountered tree-forming species was over 4400 m. The very highest record was a Juniperus recurva south of the Ngozumpa glacier at 4440 m. It was obviously a young specimen, 60 cm tall and with a trunk diameter of 5 cm; the exposure was SSW. In the same valley and at the same exposure, I found several tree junipers between 4440 and 4420 m. The tallest were 3.5 m high with trunk diameters of 20 cm near the ground; all had been occasionally lopped. The highest grove of trees in the Khumbu Himal is situated around the houses of Pangpoche at 4240 m, with a southerly exposure. The trees are 6 m high, the diameter at breast height is 25 cm at most, and the canopy is closed. It is obviously a holy grove, with prayer flags placed on trees whose branches have not been lopped. If these highest records indicate the minimum altitude of the upper limit of forests which could then be regarded as the climatic limit, it is obvious that the potential natural and climatic upper limit of forests in the valleys south of Mt Everest, especially in the Khumbu Himal, is at least 4400 m. This height for the potential upper limit of forest is clearly applicable only to the sunny slopes; on the shady slopes, it is at least 4200 m, which coincides with the highest records for individual trees. If the tree line is higher on the sunny than the shady slopes, it can be assumed that the more favourable temperature conditions of the sunny slopes determine the difference. Thus in these valleys at the northern borders of the tropics the upper limit of forests is controlled by temperature. The rainfall is obviously not a limiting factor as an annual precipitation of as little as 500 mm, concentrated in the growing period, is enough to make the sunny slopes the more favourable ones for forests. Thus the humid parts of the Himalayas and the European Alps have similar vegetational characteristics. This quasi theoretical upper forest line is overlain by other factors in both mountain regions. In these south-eastern parts of the Himalayas, daily cloud formation in the late morning is characteristic, with the effect that the east-facing slopes get full sun during the first four or five hours of daylight, whereas the westerly slopes are under clouds in the afternoon, when otherwise the sun would warm this exposure. At first glance it seems quite obvious that this determines the 252 G. MIEHE differences in the vegetation pattern of these valleys. However, there are only a few strictly east- and west-facing slopes and most are exposed to the south-east or north-east and, moreover, the south-eastern slopes are more easily accessible and more widely used for grazing and cultivation. Where strictly north-south running valleys occur, they are by contrast wooded on the eastern slopes and widely cleared on the west-facing slopes. Another factor overlying the influences of insolation is the valley wind which occur almost daily, starting up in the late morning. The exact effect of these winds has not yet been determined, but the difference between the vegetation on wind-exposed flanks compared with that found on lee slopes is striking. Slopes mainly covered by birch forests show firs on the exposed ridges, whereas fir-covered slopes have pines on the wind-blown parts. Near the upper limit of forest, where rhododendron woodlands become dominant, exposed slopes are devoid of trees, especially rhododendrons, but dwarf rhododendron cushions (R. setosum and R. antho- pogon) or dwarf prostrate junipers occur. It appears that on the wind-blown slopes, the vegetation is replaced by species that may be expected to be more drought resistant. As has been proved by experiments in the European Alps (Larcher, 1963), and with the knowledge that the growing season is undoubtedly humid while the winter is often dry and cold, one can assume that full solar radiation in winter, when the soil is frozen and water is not available, leads to transpiration rates that are significantly increased when the winds blow. This effect of seasonal drought due to freezing and enhanced by wind, probably leads to a tree-line which is naturally lower on the sunny and exposed slopes than on the shady lee slopes. As the leeward slope of a ravine in a north-south running valley is also the shady one, one can erroneously assume that the shady slope is more favourable for forests than the sunny slope. Remarks on the influence of grazing, burning, and wood-cutting In the area discussed there is great variation resulting from human influences. In the Barun Khola, the Iswa Khola, the Hunku Drangka, the Hinku Drangka, and the Lumding Khola, the cloud forests are in a condition which is fairly typical for the southern side of the main range. The upper cloud forest belt is cleared near the timberline for the seasonal settlements of summer grazing places. Burning is the common initial process in making clearings in order to improve the grazing conditions by the removal of the woody competitors of herbs. Usually the forests are penetrated by these clearings which are surrounded by areas of open woodland which have been gradually cleared by extensive uncontrolled grazing. In the alpine belt there are usually summer grazing places of transhumant sheep and goats. Here the herdsmen partly burn the moist alpine scrub consisting mainly of Rhododendron setosum. In the upper catchment of the Dudh Kosi, in the Khumbu Himal, the natural vegetation has been more extensively removed than is usual for this side of the Himalayas. Instead of seasonal settlements, the Sherpas have cleared the forests for the establishment of permanent settle- ments. Thus the influence of grazing is continuous throughout the year, in winter near the villages, in summer mainly in the lower alpine belt up to 5000 m. As the Khumbu Himal receives monsoon rains, there is relatively little snow in winter (although heavy falls may occur). For this reason, only the calves are kept in the houses during the winter and most of the cattle graze around the villages. After snowfall, the cattle at first remain in the fields near the houses, but then concentrate on grazing the slopes where the snow melts first. The influence of grazing is strongest on the south-facing 30° slopes. Steepness of slope is not a major factor governing the amount of grazing and a 10° slope close to a village is usually less grazed than a 20° slope at some distance from a village. On these heavily grazed slopes, most of the palatable herbs other than grasses are displaced by inedible species like Iris spp, Euphorbia spp, species of Labiatae, Campanulaceae, Gentiana spp, orchids, and dense creeping Cotoneaster spp which stabilize the slopes. The duration of snow cover is a deciding factor in governing the type of vegetation. This explains partly why some forests occur on shady slopes close to the villages. Furthermore, these are to a certain extent protected by the indigenous rules of the Sherpas, who elect a forest warden whose duty is PLANTS SOUTH OF MT EVEREST 233 to control wood-cutting, litter collecting, and even grazing in the forest (Firer-Haimendorf, 1964). Thanks to these laws the Sherpas have as yet no shortage of wood for fuel. The fact that in the Khumbu Himal the natural vegetation is so largely replaced by the secondary associations is partly due to the relief. The south side of the main range is directly in contact with the Tibetan north slope and can be entered via the Nangpa-La, a pass which is only 5716 m high and is passable even for cattle. The Sherpas, who originated on the eastern slope of the Tibetan Plateau, crossed the Nangpa-La in the first half of the sixteenth century (Oppitz, 1968). It is quite probable that the original 20 to 50 settlers first practised shifting cultivation and cleared the sunny slopes by fire, while the cattle, preferring particular species such as Abies spectabilis, transformed the forests into open woodlands. Owing to the continually uncertain harvests the population remained small but, nevertheless, their influence on the vegetation was great because the burnt areas were invaded by Juniperus recurva. This species is very typical of the clearings (Stainton, 1972: 107) and, because it is not grazed by cattle, forms a comparatively stable secondary forest. Evidently there were larger forests of junipers in the Khumbu Himal in the nineteenth century. The potato was introduced into the Khumbu area about 1850 (Fiirer- Haimendorf, 1964: 8). A drastic population increase followed and the juniper forests have been increasingly used for fuel, so that these secondary forests have, in turn, become open woodlands. Alphabetical list of vascular plants, their life forms, associated formations, and records of collection Key to plant life form symbols The present information regarding plant life forms is based on the revised list of the Raunkiaer System of plant life forms (Ellenberg, 1956: 26) as completed by Ellenberg & Mueller-Dombois (1967). Phanerophytes (P) Mes Pscap = large trees (S—50 m). Mi Pscap = small trees (2-5 m). N Pscap = dwarf trees (< 2m). Mi Pcaesp = tall shrubs (2-5 m). N_ Pcaesp = normal-sized shrubs (< 2 m). N Pgram = bamboos. Chamaephytes (Ch) Ch frut = woody dwarf-shrubs up to 50 cm, woodiness completed into branch-tips. Ch suff = semi-woody dwarf-shrubs. Woodiness restricted to the base of the shoot system. Ch herb = herbaceous chamaephytes: all non-woody perennial forbs, grasses, and ferns. Most of the non-rhizoid evergreen ferns of the Khumbu area should be included here. a = cold-deciduous (summergreen). e = evergreen. S = sclerophyllous. s/e,s/a = facultatively deciduous depending on altitude, shelter from frost. p = aphyllous (Ephedra). b = needle-leaved. rept = reptant. pulv = pulvinate. fpulv = flat cushions (Silene acaulis, Gentiana acaulis). gpulv = globose cushions (Androsace helvetica). lsucc = leaf succulents dying back to shoots, some are above the ground, mostly globose-cushions. Hemikryptophytes (H) Remnant shoots flat on the ground, often protected by dead shoot remains. Active shoots during growing season raised above the perennial ground-shoots. Typically herbaceous, the maturing stem may be lignificated. H caesp = branched or circular shoot system. 234 G. MIEHE Hrept = creeping and/or matting. Hscap = scapose without rosette. Hros = rosette. Hsem = rosette, but with leaves on the stalk. Hscap = without rosette, mainly tall forbs. Geophytes (G) Survival organs protected in the soil. Grad = root-budding geophytes. G bulb = bulbous geophytes. G rhiz = rhizome geophytes. With some species information concerning the supraterrestrial growth habit is given (caesp, scap, sem, Teptaces): Therophytes (T) Live less than a year. Growth habit see hemikryptophytes. Lianas (L) PL = phanerophytic and chamaephytic lianas including all climbing plants that do not die back periodically. There are woody (frut), semi-woody (suff), and herbaceous (herb) life forms. rPL = root climbers. stPL = winding climbers. elPL = tendril climbers. dPL = spread climbers. HL = hemikryptophytic climbers, dying periodically to a remnant shoot system near the ground. GL = geophytic lianas, dying periodically to subterranean storage organs. TL = therophytic lianas. Epiphytes (E) fac E = some phanerophytic, mainly chamaephytic woody or semi-woody facultative epiphytes, creeping on moss-covered trunks or boulders in the cloud forest belt. obl Ch E succ = obligative epiphytes with strongly modified root system and succulent leaves. GE bulb = bulbophytic, in moss of tree trunks. Further morphological characteristics rept/sem : facultatively straggling, creeping. G rad/H sem : with storage root, sometimes shoot under the surface, sometimes near the surface. Par : parasitic. hyd : water plant. fac hyd : facultatively in water (flushes, wet rocks). In the free solifluction belt changing life forms are characteristic: without the daily moving solifluction scree most of the high altitude phanerogams would be of the hemikryptophytic or flat cushion habit. So, in flat sites, where the scree movement is little, the genetically fixed life form is typical, whereas on solifluction slopes the shoots are deformed by adapting to this movement and the plants become rhizoid. * Drawing or colour photograph in Polumin & Stainton (1984). + Not included in Hara, Stearn & Williams (1978), Hara & Williams (1979), and Hara, Chater & Williams (1982). Key to plant formation symbols Cloud forest belt I cloud forest (medium belt) between 2400 m/8000 ft and 3200 m/10500 ft which is altitudinally characterized by the huge Tsuga dumosa trees; tree-trunks and branches are moss-covered. II open, mostly south-facing, wind-exposed grazed pine forests (Pinus wallichiana) up to 3500 m/11500 ft III cloud forest (upper belt) between 3200 m/10500 ft and the upper tree line (4200 m/13 800 ft) with Abies spectabilis, Betula utilis, and Rhododendron thickets. Beard lichens (Usnea longissima) are characteristic. PLANTS SOUTH OF MT EVEREST px IV wet, moss-covered rocks in the forest belt with species of mainly subtropical distribution (Corallodis- cus, Spathoglottis) and alpine cushions (Rhodiola, Saxifraga). V grazing grounds around the villages, with the plants originating from forest clearings and forest undergrowth, and alpine plants spreading down to the potential forest belt after deforestation. Alpine belt VI mostly south-facing, wind-blown dwarf juniper scrublands between 4000 m/13200 ft and 5000 m/16 500 ft. VII mostly north-facing, wind-sheltered moist alpine scrub (Rhododendron anthopogon, R. setosum, R. nivale) between 3800 m/12 500 ft and 5200 m/17 000 ft. VIII flushes. IX Kobresia pygmaea — mats between 5000 m/16500 ft and 5500 m/18 000 ft. Free solifluction belt X phanerogams adapted to the solifluction moved scree up to the highest open slopes at about 6000 m/19700 ft. Abbreviation list of collectors’ names The cited specimens are mainly those of the BM. The plants collected by A. Byers, J. F. Dobremez, J. Poelt, G. F. Smith, L. W. Swan, T. Wraber, and A. Zimmermann are partly cited according to duplicates in BM or determination lists which I was able to consult by the kind support of Susan Sutton (BM), Annelies Schreiber (M), Alton Byers (COLO), and George F. Smith (Stockport). Collections made after 1979 are mentioned by the date and partly by the number of collected plants and the herbarium. For further information and the collections before 1979 see Sutton (1978) and Hara, Chater & Williams (1982: 8). B = M. L. Banerjee. Cited specimens in BM. Beer = L. Beer. Cited specimens in BM. BL = S. Bowes Lyon. Cited specimens in BM. BLM =L.Beer,C. R. Lancaster, D. Morris. By = A. Byers. Between March and November 1984, c. 2500 specimens. Cited according to the collector’s list. The collection will be given to COLO. D = J. F. Dobremez. Cited specimens in BM and according to the collector’s list (BM). Einarss = S. Einarssen. Cited specimen in BM. dH. = J. de Haas. Between August and September 1974. Cited specimens in BM. H = A. Horsfall. Cited specimens in BM. Heub. = H. Heuberger, pers. comm. Record of Arenaria bryophylla between 6000 and 6100 m, Cho Oyu NE ridge, 1954. L = Lall Dhwoj. Cited specimens in BM. Mc = D.J. McCosh. Cited specimens in BM. M = G. Miehe. Between August and November 1982, 318 specimens, in the BM. N = D. H. Nicholson. Cited specimens in BM. r = J. Poelt. Cited specimens in BM and according to the collection list compiled by A. Schreiber (M). S = J.D. A. Stainton. Cited specimens in BM. Sch = A. T. Schilling. Cited specimens in BM and according to collector’s list (BM). Sh = Puspa Ratna Shakya. Cited specimens in BM. Sharma = Khadanand Sharma. Cited specimen in BM. Sm = G. F. Smith. Between July and August 1984, above 4500 m. Cited specimens according to ‘provisional list of plants . . .’, typescript, Stockport, Gottingen. Sw = L. W. Swan. Cited specimens in BM and according to collector’s list (BM). Tilman = H. W. Tilman. Cited specimen in BM. WwW = T. Wraber. Cited specimens in BM and according to determination list (BM). Z = A. Zimmermann. Cited specimens according to duplicates in BM and according to Candollea 15-17. Alphabetical list of flowering plants The plant names and their authors follow An enumeration of the flowering plants of Nepal (Hara, Stearn & Williams, 1978; Hara & Williams, 1979; Hara, Chater & Williams, 1982). 236 Abies spectabilis (D. Don) Mirb.* Acanthocalyx nepalensis (D. Don) M. J. Cannon * Acanthopanax cissifolius (Griff. ex Seem.) Harms * Acer caudatum Wall. * A. pectinatum Wall. ex Pax * A. stachyophyllum Hiern Aconitum balfourii Stapf A. ferox Wall. ex Seringe * A. heterophyllum Wall. ex Royle A. hookeri Stapf * A. laciniatum (Brihl) Stapf var. laciniatum A. rotundifolium Kar. & Kir. A. spicatum (Brihl) Stapf * Aconogonum campanulatum (Hook.f.) Hara var. campanulatum * A. campanulatum var. oblongum (Meisn.) Hara * A. molle (D. Don) Hara A. rumicifolium (Royle ex Bab.) Hara * A. tortuosum (D. Don) Hara * Acronema cf. hookeri (C. B. Clarke) H. Wolff A. johrianum Babu + A. tenerum (DC.) Edgew. Aechmanthera gossypina (Wall.) Nees * Aeschynanthus hookeri C. B. Clarke Agrostis munroana Aitch. & Hemsley A. nervosa Nees ex Trin. A. pilosula var. wallichiana (Hook.f.) Bor A. triaristata (Hook.f.) Bor. Ainsliaea aptera DC. * Ajuga bracteosa Wall. ex Benth. A. lobata D. Don * Aletris pauciflora (Klotzsch) Hand.-Mazz. * Allium wallichii Kunth * Alnus nepalensis D. Don * Anaphalis busua (Buch.-Ham. ex D. Don) DC. A. cavei Chatterjee A. contorta (D. Don) Hook.f. G. MIEHE b Mes P scap aHsem a NP caesp a Mi P scap a Mi P scap a Mi P scap a G bulb a G bulb a G bulb a G bulb a G bulb a G bulb a G bulb a Grad a Grad a Grad aG rhiz a Grad a Grad a Grad a Ch suff sobl.d EPL a H caesp a H caesp a H caesp a H caesp a H ros a H caesp aH rept a Grad aGrad a/e Mes P scap a H scap/Ch suff af Ch herb pulv aHsem Lagi Clinopodium umbrosum (M. Bieb.) C. Koch Clintonia udensis var. alpina (Kunth ex Baker) Hara * Coccinea grandis (L.) Voigt Codonopsis ovata Benth. * C. thalictrifolia Wall. * Coleus barbatus (Andrews) Benth. * Colutea multiflora Shap. * Commelina paludosa Blume * Corallodiscus lanuginosus (Wall. ex DC.) Burtt * Cortia depressa (D. Don) C. Norman * G. MIEHE a H scap s Ch frut s Mes P scap ad LP frut aH rept a Grad T scap a f Ch frut pulv e g Ch herb pulv H caesp/scap H caesp a G rhiz/scap a G rhiz/sem G rhiz/scap a Grad/H scap a Grad/H scap a Grad a G rad/a H scap T rept/scap T rept/scap T rept/scap a G rad/sem a G rad/scap ael PL frut a H caesp a G rhiz owe a H scap a G rad/rept a H scap/Ch suff a NP caesp aH scap e Hros aGrad II W.397 By.469, 687, D.379, W.248, Z.612, 634, 702, 1559, 1756, Sm.s.n. f.M. S.4484 By.498 M.936 By.724, P.s.n. By.1124 BL.2112, By.41, M.1035, Mc.329, W.249, Sm.s.n. Sw.224, 225 Z.805 By.38, S.7117 M.983 Z.458 BL.2035, Mc.53 dH.2950a, P.s.n., Z.1791 P:s.n. By.586, 712, P.s.n. dH.2941, M.128 W.317 P:7 S.5750 W.377, Z.1670 By.601 By.654, dH.2975, Sch.2186, Z.502 P.s.n., W.307, Z.1670 By.26, Z.641 Psa. By.362 By.342, 430, 453, M.1058, Sm.s.n. Z.1833, 1865 dH.2951, Sch.2165 By.465 D.321, Mc.394, M.811, Z.422, 1225 BL.2179, D.368, P.s.n., G.259, Sm.s.n. Cortiella hookeri (C. B. Clarke) C. Norman * Corydalis cashmeriana Royle * C. casimiriana Prain C. cavei D. G. Long C. chaerophylla DC. C. chasmophila Ludlow C. flabellata Edgew. C. flaccida Hook. f.& Thoms. C. govaniana Wall. * C. hookeri Prain C. juncea Wall. * C. leptocarpa Hook.f. & Thoms. C. longipes DC. C. meifolia Wall. var. meifolia * C. polygalina Hook.f. & Thoms. C. stracheyi Duthie ex Prain + Corylus ferox Wall. * Cotoneaster acuminatus Lindl. C. adpressus Bois C. frigidus Wall. ex Lindl. * C. integrifolius (Roxb.) Klotz C. meuselii Klotz C. microphyllus Wall. ex Lindl. * C. nitidus Jacques C. sanguineus Yi Cremanthodium decaisnei C. B. Clarke * C. ellisii (Hook.f.) Kitam. * C. nepalense Kitam. * C. oblongatum C. B. Clarke * C. reniforme (DC.) Benth. * C. retusum (Wall. ex Hook.f.) R. Good * Crepis tibetica Babcock Cryptothladia polyphylla (Wall. ex DC.) M. J. Cannon * Cuscuta reflexa Roxb. Cyananthus hookeri C. B. Clarke C. incanus Hook.f. & Thoms. * C. inflatus Hook.f. & Thoms. C. lobatus Wall. ex Benth. * C. microphyllus Edgew. * C. pedunculatus C. B. Clarke C. spathulifolius Nannfeldt Cyanotis vaga (Lour.) J. A. & J. H. Schultes * Cymbidium hookerianum Reichenb. f. * PLANTS SOUTH OF MT EVEREST aGrad IX, X a G bulb V, VII Vv Vv I a H caesp VII t Ill a H caesp III, V VII a G bulb/scap LEY Ill a G rad/caesp Ill a H caesp Vil a G bulb vil a H caesp VI a Mes P scap I a N P caesp He a Ch frut rept II, V a Mi P caesp II s Ch frut rept iV a N P caesp II s Ch frut rept Me s/a N P caesp II, V a N P caesp II a Hros VIII, X aHros VIII, X aH ros VII aHros VII, X aHsem vil aH sem VII a H ros/sem ? aGrad Vv ELT Par II aH rept V, VI a Grad/H rept Vi¥i a H rept/scap Vv a G rad/H rept Vv a G rad/H rept Vv aH rept V, VI a Grad/rept VI, VII aH II fac. e Ch E succ I 241 M.1001, W.164, 233, Sm.s.n. D.431, Mc.283, P.s.n. H.10 (f. Lidén) H.10 (f. Lidén) Mc.266 BL.2163, $.628 By.193 Mc.295, S.8360 M.879, Z.809, Sm.s.n. S.7153 BL.2164, Mc.341, $.4538 S.4536 By.525, 1167, dH.2935, M.850, W.386 M.1083, Beer 9555, Sm.s.n. Mc.327 1:357 S.4606 dH.2930, M.854, Mc.273, S.4614, Sch.2030 Sch.2031, 2263 Sch.2442A 9.7125 Sch.2355 By.1089, P.s.n., Sch.2032 M.932 Mc.304, BL.2168 BL.2121, Mc.371, Sm.s.n. M.1108, W.165, Z.1456B, Sm.s.n. By.592, H.46, Sm.s.n. By.545, M.901, 966, Z.1471 By.506, 639, M.895, Z.1451 Z.1451 W.299 Mc.381 P.s:n; H.A7,.33,,M:831, 915; W.206 By.76, 299, 328, 678, 1082, dH.2943, W.238, Z.1648, 1714, Sm.s.n. By.492, H.9, M.914, P.85, 88 P.89 H.5, 55, P.94 By.587, 795, M.873, P.s.n. Mc.308 By.466 S.4476 242 Cynanchum auriculatum Wight Cynoglossum glochidiatum Wall. ex Benth. * C. zeylanicum (Vahl) Thunb. ex Lehm. Cypripedium himalaicum Rolfe apud Hemsl. * Dactylorhiza hatagirea (D. Don) So6 * Danthonia cachemyriana Jaub. & Spach D. cumminsii Hook.f. Daphne bholua Buch.-Ham. ex D. Don D. retusa Hemsl. * Datura stramonium L. * Delphinium caeruleum Jacquem. ex Cambess. D. drepanocentrum (Brihl) Munz * D. glaciale Hook.f. & Thoms. D. cf. nepalense Kitam. & Tamura D. viscosum Hook.f. & Thoms. * Desideria nepalensis Hara Desmodium elegans DC. * D. heterophyllum (Willd.) DC. Deutzia bhutanensis Zaikonn. * D. compacta Craib * D. staminea R. Br. ex Wall. * Deyeuxia pulchella (Griseb.) Hook.f. Dicentra scandens (D. Don) Walp. Dichrocephala benthamii C. B. Clarke D. integrifolia (L.f.) O. Kuntze Didymocarpus aromaticus Wall. ex D. Don D. oblongus Wall. ex D. Don D. primulifolius, D. Don * Diphylax urceolata (C. B. Clarke) Hook.f. Diplarche multiflora Hook.f. & Thoms. Dipsacus inermis var. mitis (D. Don) Y. Nasir Disporum cantoniense (Lout.) Merr. * D. cantoniense var. parviflorum (Wall.) Hara Dodecadenia grandiflora Nees * Draba altaica (C. A. Meyer) Bunge * D. amoena O. E. Schulz * D. elata Hook.f. & Thoms. G. MIEHE ad PL frut a H scap a H scap/caesp a G rhiz a Grad a H caesp a H caesp s NP caesp s NP caesp a H scap aH sem aHsem aHsem aHsem aH sem a NP caesp a NP caesp a NP caesp a NP caesp a NP caesp a G rhiz ad HL aHsem aHsem a G bulb/scap a G bulb/scap a G bulb/scap a Grad s Ch frut/rept a H sem/scap a G rad/scap a Grad s Mes P scap a H ros/a f Ch herb pulv aHsem aHsem — a II, V-VIII II I, VI Mc.395 B.5659, Mc.275, P.193 By.397, 872 BL.2199, By.120, D.395, Sm.s.n. By.142 M.892c, P.s.n. M.877, 851b, W.349 S.4480 W.328 f.M. P:s.n; By.423, 537, 621, dH.2969, L.0191, M.882 M.1060, P.s.n., W.172, 219, Sm.s.n. M.1033 Beer 25387, W.372 Sw.71-72 By.364 Pes hi: Mc.38, $.4621 BL.2157, By.499, S.7123 Sch.2087, Z.472, 1860, s.n. M.817, 817a, 1021b, 1028, W.247 P.s.n., Z.1883 Z.749B Z.749 Z.746 S.7163 S.7164 L.0529 Beer 25361, Sw.441 Z.1671 Z.765 Sch.2366A S.6559 Mc.325, W.167 BL.2178, D.424. S.7144 BL.2114, D.440, Mc.287, Z.521, 581 D. ellipsoidea Hook.f. & Thoms. D. gracillima Hook.f. & Thoms. D. lasiophylla Royle D. lasiophylla var. leiocarpa (Pamp.) O. E. Schulz D. cf. oariocarpa O. E. Schulz D. oreades Schrenk * Drosera peltata Smith Drymaria cordata (L.) Willd. ex Roem. & Schult. Dubyaea hispida DC. * Elaeagnus infundibularis Momiyama Elsholtzia concinna Vautier E. densa Benth. E. eriostachya (Benth.) Benth. * E. eriostachya var. pusilla (Benth.) Hook.f. E. fruticosa (D. Don) Rehder * E. strobilifera (Benth.) Benth. Elymus canaliculatus (Nevski) Tzvel. E. nutans Griseb. Enkianthus deflexus (Griff. ) Schneid. * Ephedra gerardiana var. sikkimensis Stapf * Epilobium conspersum Hausskn. E. leiospermum Hausskn. E. cf. wallichianum Hausskn. * E. williamsii Raven Erigeron bellidioides (Buch.-Ham. ex D. Don) Benth. ex C. B. Clarke * E. kumaunensis (Vieth.) Wendelbo E. multiradiatus (Lindl. ex DC.) C. B. Clarke * Eriophyton wallichii Benth. * Erysimum hieraciifolium L. Euonymus amygdalifolius Franch. E. frigidus Wall. E. hamiltonianus Wall. * E. pendulus Wall. E. porphyreus Loesener + PLANTS SOUTH OF MT EVEREST T caesp Vv T sem V, VII af Ch herb pulv/a IX, X H ros a f Ch herb pulv VI, VII a Hros/a f Ch herb VII, IX pulv aHros/afChherb X pulv/G rhiz aH ros III, V aHrept/afChherb IX,X pulv a H scap/G rad V, VI a/e Mes/Mi P I scap/caesp T scap/caesp TV. a H scap Vv T scap | | "3 T scap 8 hf aNP caesp/Ch suff iI T scap II, Ill a H caesp ViVi aH caesp VI s/a Mi P caesp if p Ch frut LEV VIS aH scap V, VI a H scap ? a H scap IV, VII ? aHsem II aHsem I, V aHsem V, VI a G rhiz/scap X T scap II, V s NP caesp I s NP caesp I a Mi P scap | s Mes P scap I a Mi P scap I 243 W.204 Z.807 BL.2075, M.1102, Sm.s.n., S.7158, W.175, 184 D.445 P.G.206 BL.2190, M.1098, Z.569 By.365, D.333, Z.1694 BL.2119 By.583, 590, 591, dH.2962, M.880, P.s.n., W.330, Z.8:n. P.s.n., Sch.2442A, S.6574, Z.1850 Z.1778 Z.683 By.205, 675, D.339, L.0180, M.889 Z.643 By.509, 611, 882, 1123, M.822, P.s.n., Z.1654, 1828 M.889a, P.s.n., W.302, £31120 M.851la P:s-n: BL.2013, Sch.2336, Z.426 By.736, BL.2069, D.439, M.993, P.R.333, G.293, Sch.2102, W.258, Sm.s.n. By.310, P.R.336 W.384, 392 M.844 W.300 D.325, Z.447, 451, 727 Z.1864 By.294, M.876, Z.451, 1565, 1622 By.1173, M.1007, W.236, 2.1510; Sm.s.n; Z.450 S.6573 P.8.7: By.988 $.6572 Mc.43, $.5949 244 E. tingens Wall. * Euphorbia himalayensis Klotzsch E. longifolia D. Don E. pseudosikkimensis (Hurusawa & Ya. Tanaka) Radcliffe-Smith E. sikkimensis Boiss. E. stracheyi Boiss. E. wallichii Hook.f. * Euphrasia himalayica Wettst. * E. platyphylla Pennell E. secundiflora Pennell Eurya acuminata DC. * Fagopyrum dibotrys (D. Don) Hara F. esculentum Moench Festuca ovina L. F. polycolea Stapf Festuca tibetica (Stapf) Alexeev + Ficus sarmentosa Buch.-Ham. ex Sm. Fragaria nubicola Lindl. ex Lacaita * Fritillaria cirrhosa D. Don Galearis stracheyi (Hook.f.) P. F. Hunt * Galinsoga parviflora Cav. Galium aparine L. var. aparine G. asperifolium Wall. G. asperifolium var. sikkimense (Gandoger) Cufodontis G. asperuloides subsp. hoffmeisterii (Klotzsch) Hara Gaultheria fragrantissima Wall. * G. hookeri C. B. Clarke G. nummularioides D. Don G. pyroloides Hook.f. & Thoms. ex Mia. * G. trichophylla Royle * Gentiana algida Pall. * G. algida var. nubigena (Edgew.) Kusn. G. algida var. parviflora (C. B. Clarke) Kusn. G. algida subsp. przewalskii (Maxim.) Kusn. G. argentea (D. Don) C. B. Clarke G. capitata Buch.-Ham. ex D. Don * G. crassuloides Bur. & Franch. G. depressa D. Don * G. MIEHE s Mi P scap a G rad/scap a G rad/scap a G rad/scap a G rad/scap a Grad/rept a G rad/scap T scap T scap T scap s Mes P scap aHsem a H scap a H caesp a H caesp a H caesp sr PL frut a H caesp a G bulb a G rhiz T scap T rept T rept T rept s NP caesp s NP caesp s Ch frut rept s Ch frut rept s Ch frut rept e H caesp/rept e H rept/sem e Hrept e H rept/sem e H scap e H ros e H caesp e f Ch herb pulv TV val VI, VI Vv Ill, V Ill IV Ill, V S.4607 Sch.2086, 2291, 2324 Sch.2069 dH.2959 Sch.2060 B.5698, BL.2131, M.954, Z.490, 410, 537, Sm.s.n. By.101, 1127, M.826, Z.378 By.415, Sm.s.n. S.656 P.s.n. S.4482 Z.766, 1862 By.304, 701 Mc.289, W.183, 242, Sm.s.n. D.373, $.550 M.1072 f.M. By.1129 Z.690, Sm.s.n. By.131, D.427 Z.1765 By.636, P.s.n. W.426 P-R275 W.306 Sch.2050, Z.444, 1856 BL.2022, S.4612 M.820 S.4616, Z.792 By.109, M.970, S.4541, Z.623, 1785 $.6591 M.1075, P.G.203, Sch.2313, Sm.s.n. Sm.s.n. H.49 E. Onyon 5 Sw.525 Sm.s.n. By.362, H.56, M.1117, P.147, 149, Sch.2316, Tilman s.n. G. elwesii C. B. Clarke G. infelix C. B. Clarke G. ludlowii Marq. G. micans C. B. Clarke G. ornata (G. Don) Griseb. * G. ornata forma alba G. pedicellata (D. Don) Griseb. * G. phyllocalyx C. B. Clarke * G. prolata Balf.f. G. sikkimensis C. B. Clarke G. speciosa (Wall.) Marq. G. squarrosa Ledeb. + G. stellata Turrill + G. strobilacea H. Sm. G. tubiflora (G. Don) Griseb. * G. urnula H. Sm. * G. vernayi Marq. Gentiana sp. Gentianella falcata (Turcz. ex Kar. & Kir.) H. Sm. G. glanduligera Airy Shaw G. pedunculata (D. Don) H. Sm. G. stellariifolia (Franch. ex Hemsl.) H. Sm. Geranium donianum Sweet * G. nakaoanum Hara * G. polyanthes Edgew. & Hook.f. * G. refractum Edgew. & Hook.f. * Gerbera maxima (D. Don) Beauverd var. maxima G. nivea (DC.) Sch. Bip. Girardinia diversifolia (Link) Friis * Goodyera fusca (Lindl.) Hook.f. G. repens (L.) R. Br. * Gueldenstaedtia himalaica Baker * Gymnadenia orchidis Lindl. Gypsophila cerastioides D. Don * Habenaria arietina Hook.f. Hackelia uncinata (Royle ex Benth.) C. E. C. Fischer Halenia ellipticaD. Don * Hedera nepalensis K. Koch Hedychium densiflorum Wall. H. spicatum Smith * Hedysarum sikkimense Benth. ex Baker * PLANTS SOUTH OF MT EVEREST aH sem a H caesp e H sem/caesp e H sem/caesp e H caesp/sem/e f Ch herb pulv e f Ch herb pulv H caesp a H sem/G rhiz e H caesp/rept a H scap aH rept a H sem/caesp e Hros e Hros e H ros/G rhiz T caesp aHros T ros/sem a H ros/sem T sem/scap T caesp aHsem aHros aHsem aHsem aH sem/G rad a G rad/H ros a H scap a G rad/rhiz a G rhiz a G rad/H caesp a Grad a H rept/a f Ch herb pulv a G bulb aH scap Tsem sr PL frut a G bulb aG bulb aH rept VII VIII V, VI, VII Vv V, VI VII IX, X X V, VII, 1X X Visi ? IX VI, VII VII Ill, V 245 Beer 25368 BLM 9548 H.2, 25, M.940, P.R.291 dH.2963, Sm.s.n. By.597, H.35, M872, P.R.290, G.212, S.1673, 6587, Sch.2314, Sharma E.465, W.220, Sm.s.n. M.1082 Sw.428 S.589 Sch.2184 W.158 P.B.9, 153, 154, Sch.2081 D.2996, P.s.n. P.G.287, S.1674, 1694, W.239 Sw.314 BLM 9549 M.1080, Sch.2314, $.6595, Sw.21-23, Sm.s.n. H.24, M.1088, W.205 M.1088 M.999, P.R.334, G.253, G.302, Sm.s.n. W.427 M.1110, P.s.n., W.221 S;1677 By.14, 152, D.335, 401, M.947, S.7150, Sm.s.n. By.1146, D.360, Sm.s.n. By.275, P.G.329 By.192, 234, 463, BL.2096, D.420, H.37, Mc.347, Sm.s.n. Z.1872 By.670, 787, P.s.n., W.353 Z.1874 M.925, W.316 By.642 By.268, BL.2065, D.345, H.32, Mc.321, Sm.s.n. L.0193 By.316, BL.2165, M.979, Mc.344 H.70 BL.274 By.239, 431, M.918, W.294 Z.1838, 1840 Sch.2042 Sch.2043 BL.2079, By.43, 438, M.1024, Mc.310, $.7101, Sm.s.n. 246 Helictotrichon sp. Hemiphragma heterophyllum Wall. * Heracleum brunonis (DC.) C. B. Clarke H. candicans Wall. ex DC. H. nepalense D. Don H. wallichii DC. Herminium josephii Reichenb.f. H. lanceum (Thunb.) Vuijk * H. macrophyllum (D. Don) Dandy H. pugioniforme Lindl. ex Hook.f. Hierochloe flexuosa Hook.f. Hippophae tibetana Schlecht. * Holboellia latifolia Wall. H. latifolia var. angustifolia (Wall.) Hook.f. & Thoms. * Hydrangea heteromalla D. Don Hypecoum leptocarpum Hook.f. & Thoms. Hypericum choisianum Wall. ex N. Robson * H. himalaicum N. Robson Ilex dipyrena Wall. * I. intricata Hook.f. I. odorata Buch.-Ham. ex D. Don I. sikkimensis Kurz Impatiens falcifer Hook.f. I. glandulifera Royle * I. puberula DC. I. pulchra Hook.f. & Thoms. I. sulcata Wall. * I. urticifolia Wall. * Indigofera bracteata Grah. ex Baker I. heterantha Wall. ex Brandis * Inula cappa (Buch.-Ham. ex D. Don) DC. * I. hookeri C. B. Clarke * I. nervosa Wall. ex DC. Tris kemaonensis D. Don ex Royle * Jasminum humile L. * J. humile forma pubigerum (D. Don) Grohmann J. humile var. revolutum (Sims) Stokes Juncus benghalensis Kunth J. clarkei Buchenau + J. concinnus D. Don J. grisebachii Buchenau J. himalensis Klotzsch * J. leschenaultii Gay G. MIEHE aH caesp aH rept a Grad a Grad a Grad aGrad aG bulb aG bulb aG bulb a G bulb a H caesp a/e Ch frut/pulv ael PL frut ael PL frut a Ch frut/r PL frut a H caesp/rept a H caesp a H scap s Mi P scap s NP caesp s MiP scap a MiP scap a H caesp T scap a H scap aH scap aH scap aH scap a H caesp a NP caesp a/e Ch suff aH scap a/e G rhiz/H scap a G rhiz sd PL frut sd PL frut sd PL frut a G rhiz a G rhiz a G rhiz a G rhiz a G rhiz a H caesp M.892 dH.2978, Z.1819 By.322, 345 BL.2150, By.850, S.6598 M.883b By.809, M.883c D.393 H.65a BL.2174, M.945, Mc.311 Mc.355 By.520 By.428, dH.2989, f.M., P.G.305, f.Sch.(2268), 2179 $.4490 Z.432, 473 Mc.269 By.768, Sm.s.n. Mc.399, S.7168 Mc.404 Z.1675 P.73, S.4611 S.4610 BL.2019 By.246, 376 L.067 By.300, 573-575, W.341 f.M. dH.2916, P.s.n., $.6570, Z.1693, 1845A dH.2919 By.81 dH.2913, Z.666 By.860 P.G.203 P.R.327 W.322 M.868, P.B.55, 171 W.376 M.1029a, $.567 Z.1724 J. leucanthus Royle ex D. Don * J. leucomelas Royle ex D. Don * J. pseudocastaneus (Lingelsh.) Samuelsson J. sphacelatus Decaisne J. thomsonii Buchenau Juniperus indica Bertol. * J. recurva Buch.-Ham. ex D. Don - J. squamata Buch.-Ham. ex D. Don * Kobresia caricina Willd. K. curvata Kiikenthal K. deasyi C. B. Clarke K. fissiglumis C. B. Clarke + K. macrantha Boeck. K. nepalensis (Nees) Kiikenthal K. pygmaea (C. B. Clarke) C. B. Clarke K. schoenoides (C. A. Meyer) Steudel K. seticulmis Boeck. K. trinervis Boeck. Koenigia delicatula (Meisn.) Hara K. islandica L. K. nepalensis D. Don Lactuca lessertiana (DC.) C. B. Clarke * Lagotis kunawurensis (Royle ex Benth.) Rupr. * Lamium nepalense Hedge Lancea tibetica Hook.f. & Thoms. * Lasiocaryum densiflorum (Duthie) J. M. Johnston L. diffusum (A. Brand) J. M. Johnston L. munroi (C. B. Clarke) J. M. Johnston Leibnitzia nepalensis (Kunze) Kitam. L. ruficoma (Franch.) Kitam. Leontopodium brachyactis Gand. L. jacotianum Beauverd * L. monocephalum Edgew. PLANTS SOUTH OF MT EVEREST a G rhiz a H caesp a G rhiz a G rhiz a G rhiz b N/Mi P scap/Ch frut b N/Mi P scap b Ch frut/rept a H caesp a H caesp a H caesp a H caesp a H caesp a H caesp a H caesp a H caesp a H caesp T rept T rept T rept a H sem/G rad aHros aH scap aH ros T caesp T caesp aHsem aHros aH ros/G rad aH sem/f Ch herb pulv G rhiz/ros/sem af Ch herb pulv IV VIII ? VIII, X Ill, IV Ill, V, VI Ill, V VI, VIII, 1X, Xx IX VIII, X IV, VI, VII, X VI IV VIII VIII Ill, VI, VI VIII, X II VI VI, IX VI, IX ~ 247 By.285, D.409, M.956, W.188a, 252, Sm.s.n. D.409 By.546 D.377, M.1093, P.R.326 By.222, P.G.264, S.7096, W.188, 190, Z.665, 1524, 1540 By.1134, D.459, P.G.300, 318, $.4513, 4515, 4528, 7116 By.80, P.3, 4, $.4529, 4514, W.410 By.1168, Z.1699 BL.2127 M.867, P.132 P.O .267 W.269 BL.2188 M.s.n., 959, 988, 1071a M-95751071b, P.R.332, W.237 M.1053, 1104, 1111 M.866, 1087, Mc.2032, P.133, 134, R226 BL.2197 M.916, P.s.n., W.382 W.245, Z.651 M.981, Mc.312, W.381 M.982, 1023, W.251, Z1529-Smisay BL.2105, By.392, D.407, M.1069, Sm.s.n. L.0139 BL.2117, By.13, D.442, Sm.s.n. By. 668 W.208 Beer 25543 Z.403 dH.2945, Z.1754 M.881, 942, W.403 D.356, Z.1607, 1697, Sm.s.n. BL.2139, D.428, H.48, M.1048, 1068, 1092, PG.207,G.207,.8:6582; W.166, Z.1591, Sm.s.n. 248 L. nanum (Hook.f. & Thoms. ex C. B. Clarke) Hand.-Mazz. L. stracheyi (Hook.f.) C. B. Clarke ex Hemsley Lepidium capitatum Hook.f. & Thoms. Leptodermis kumanoensis Parker * L. stapfiana H. Winkler Leucosceptrum canum Sm. * Leycesteria formosa Wall. * Lignariella hobsonii (Pearson) Baehni * L. hobsonii subsp. serpens (W. W. Smith) Hara Ligularia amplexicaulis DC. * L. fischerii (Ledeb.) Turcz. * Lilium nanum Klotzsch * L. nepalense D. Don * Lindera heterophylla Meisn. L. neesiana (Wall. ex Nees) Kurz L. pulcherrima (Nees) Benth. ex Hook.f. * Listera nepalensis Balakrishnan Litsea cubeba (Lour.) Pers. L. sericea (Wall. ex Nees) Hook.f. Lloydia flavonutans Hara L. longiscapa Hook * L. serotina (L.) Reichenb. * L. serotina var. parva (Marquand & Shaw) Hara Lobelia seguinii H. Lév. & Van. Lomatogonium brachyantherum (C. B. Clarke) Fernald L. carinthiacum (Wulf.) Reichenb. * L. chumbicum (Burkhill) H. Sm. L. graciliflorum H. Sm. L. sikkimense (Burkhill) H. Sm. Lonicera acuminata Wall. L. angustifolia Wall. ex DC. L. cyanocarpa var. porphyrantha Marquand & Shaw * L. hispida Pall. ex Willd. * L. hispida var. setosa Hook.f. & Thoms. L. lanceolata Wall. L. litangensis Batalin L. myrtillus Hook.f. & Thoms. * L. obovata Royle ex Hook.f. & Thoms. * L. rupicola Hook.f. & Thoms. * Luzula sp. G. MIEHE af Ch herb pulv a H caesp T rept a NP caesp a NP caesp a NP caesp a NP caesp T rept T rept aH sem aH scap a G bulb aG bulb s Mes P scap a/s Mes P scap a/s Mes P scap aGrad a Mes P scap s Mes P scap aG bulb a G bulb aG bulb aG bulb a H scap a H caesp aHsem aH sem aHsem aHsem a Ch frut a Mi P caesp a Ch frut a NP caesp a Ch frut a NP caesp a Ch frut a Ch frut a Ch frut a Ch/NP caesp H caesp IX Vv Vv VI, VII VI II VIII, 1X VI-IX BL.2125, M.1041 By.389, dH.2953, M.833, P.111, 112, Sch.2176, Z.1651, 1821 M.1116 S.7160 S.4618 Z.1858 P.96, Z.743 W.331 BL.2207 By.171 By.286, L.0182 BL.2056, 2074, By.150, S.4775 By.176, 179, Sch.2248 S.6566 S.6571, 6599 Z..260 W.339 S.6567 $.5942 Mc.279 Beer 25375, Sm.s.n. BL.2962, By.308, D.375 BL.2170, M.944, Mc.326 Z.1863 H.34, P.R.347 By.1174, M.1056, P.G.223, R.329, $.6593, W.250 M.946, W.402 W.292, 400 W.420 Mc.393 BL.2083, 2152, By.68, 873, M.935, Z.429 Mc.305 D.434, Sm.s.n. BL.2104, M.1050, Mc.350 BL.2023, By.51, 282, 557, 1002, D.452, M.934 Z.520 By.425, D.418 BL.2185, By.903, P.R.357, Sch.2078, Z.614, Sm.s.n. By.37, M.985, P.R.358, Z.509, 1530, Sm.s.n. P.R.341 Lyonia ovalifolia (Wall.) Drude * L. villosa (Hook.f.) Hand.-Mazz. * Lysimachia prolifera Klatt Malaxis cylindrostachya (Lindl.) Kuntze M. muscifera (Lindl.) Kuntze Mandragora caulescens subsp. flavida Grierson & Long * Meconopsis bella Prain * M. horridula Hook.f. & Thoms. * M. paniculata Prain * M. simplicifolia (D. Don) Walp * M. sinuata Prain * Megacodon stylophorus (C. B. Clarke) H. Sm. * Meliosma dilleniifolia (Wall. ex Wight & Arn.) Walp. Michelia doltsopa Buch.-Ham. ex DC: Microcaryum pygmaeum (C. B. Clarke) J. M. Johnston Micromeria biflora (Buch.-Ham. ex D. Don) Benth. * Microsisymbrium axillare (Hook.f. & Thoms.) O. E. Schulz Microula pustulosa (C. B. Clarke) Duthie M. sikkimensis (C. B. Clarke) Hemsl. * Miscanthus nepalensis (Trin.) Hackel Myricaria davurica (Willd.) Ehrenb. M. rosea W. W. Sm. * Nannoglottis hookeri (C. B. Clarke ex Hook.f.) Kitam. * Nardostachys grandiflora DC. * Neilla rubifloraD. Don Neolitsea pallens (D. Don) Momiyama & Hara ex Hara * Neottianthe calcicola (W. W. Smith) Schlechter N. secundiflora (Hook.f.) Schlechter Nepeta coerulescens Maxim. N. lamiopsis Benth. ex Hook.f. Notholirion macrophyllum (D. Don) Boiss. * a/s N/Mi P scap a/s N/Mi P scap a H caesp/rept a G bulb aG bulb a Grad/ros a H scap aHsem aH sem a H scap a H scap a H scap a Mi P scap s Mes P scap T ros a H caesp aHsem a H caesp/scap a H scap a H caesp s Ch frut s Ch frut rept a Hscap/G rad aH sem a NP caesp s MiP scap a G bulb a G bulb a H scap a H rept/scap a Grad PLANTS SOUTH OF MT EVEREST III Ill I, Il 4 ef V>VEIX II i, i; V, Vil TV Vv Vv Ill, V—-VII 249 f.M. M.858, P.G.358, Z.1852 Z.781 Mc.397 By.332, H.15 Z.688 Beer 25425, BL.2161 By:451,54:51,.P:s-n.. Sch.2177, 2351, W.280, Sm.s.n. S.649, W.395 By.28, Mc.348, Z.594 Beer 25445 BL.2153; By.271, L.0183, Z.1703 S.4620 S.4478 M.s.n. Z.400 Z.258 BL.2156 BL.2183, D.367, M.887, Sm.s.n. Sch.2063, Z.1859 Z.510 By.350, dH.2971, M.969, P.s.n., Sm.s.n., W.432 By.220, Einarsson 9, L.0178, M.870, Mc.444, P YSN). cs. 3000, 1150. Z.635, 687 L.0212 BL.2081 S.4482 H.26, M.963, P.16 H.61, P.16 B.5778 B.5636, By.52, D.385, H.16, L.0170, M.905, Mc.359, P.s.n., Z.1766 By.217, D.331, Sch.2105 250 Omphalodes trichocarpa Maxim. + Orchis diantha Schlechter + Oreorchis foliosa (Lindl.) Lindl. * O. micrantha Lindl. Oreosolen wattii Hook.f. * Oxalis corniculata L. * O. corniculata var. villosa (Bieb.) Hohen + Oxyria digyna (L.) Hill * Oxytropis arenae-ripariae Vass. O. nepalensis Vass. O. williamsii Vass. * Paris polyphylla Smith * Parnassia chinensis Franch. P. kumaonica Nekrassova P. nubicola Wall. ex Royle * P. pusilla Wall. ex Arn. Parochetus communis Buch.-Ham ex D. Don * Pedicularis bifida (Buch.-Ham. ex D. Don) Pennell P. brevifolia D. Don P. clarkei Hook.f. P. confertiflora Prain P. elwesii Hook.f. P. flexuosa Hook.f. P. furfuracea Wall. ex Benth. P. gracilis Wall. ex Benth. P. longiflora var. tubiformis (Klotzsch) Tsoong * P. megalantha D. Don P. microcalyx Hook.f. P. mollis Wall. ex Benth. P. muscoides L. P. nana C. E. C. Fischer P. oederi subsp. oederi var. heteroglossa Prain * P. pseudoregeliana Tsoong P. punctata Decr. + P. rhinanthoides Schrenk. P. roylei Maxim. P. scullyana Prain ex Maxim. * P. siphonanta D. Don P. trichoglossa Hook.f. * P. wallichii Bunge Pegaeophyton minutum Hara P. scapiflorum (Hook.f. & Thoms.) Marg. & Shaw * Pentapanax leschenaultii (DC.) Seem. * G. MIEHE a G rhiz aG bulb aG bulb aHros aH rept aH rept aGrad a H caesp a H caesp/rept a H caesp a G rhiz a Hros aH ros aHsem aHros T rept a H caesp/scap aHsem a G rad/H scap T sem H caesp/G rad aHsem aHsem a H scap a H ros/G rad a H scap/G rad G rad/scap a H caesp/scap aH ros/G rad a H ros/sem G rad/H sem a Hros a H caesp/scap a H caesp a H caesp a Hscap a H ros/caesp/G rad aH scap aHsem a H caesp/a f Ch herb pulv a G rhiz/a f Ch herb pulv a Mi P caesp VI Vv VIII VII, VIII Ill Ill Vv VIII Ill, V, VII VII |b ag VIII VI, VII VIII By.41, 213, 374, 892 Mc.294, BL.2048 BL.2026, Mc.292 BL.2025 BL.2124, M.1095, W.230, L.955, ons. Z.1842 Onyon 9 By.287, M.1004, P.s.n., S.7598, Z.599, Sm.s.n. S.7109 BL.2107 By.758 Z.449 M.s.n., Sm.s.n. By.346, M.864, Mc.328 By.518, L.0128, M.900 BL.2184, H.11, Mc.352, Sm.s.n. By.s.n., W.423 H.29 BL.2093, By.258, D.351, W.286 Beer 25339 BL.2030, Mc.268 B.5647, By.371, M.896 H.31, BL.2181, By.277 By.289, 641, M.846 BL.2137 B.5697, By.247, H.1, P.55 BL.2129, Mc.366 D.433 D.411 BL.2086, 2201, S.4603 By.422, Shresta 5661, Sm.s.n. BL.2180, By.434, D.381 By.368, D.350, L.0185, M.965, P.G.234 Mc.285 BL.2088, Mc.388 M.1099, Mc.335, 373 BL.2158 Peracarpa carnosa (Wall.) Hook.f. & Thoms. Peristylus elisabethae (Duthie) Gupta Peristylus fallax Lindl. Persicaria glacialis (Meisn.) Hara P. microcephala var. sphaerocephala (Wall. ex Meisn.) Hara P. polystachya (Wall. ex Meisn.) H. Gross P. runcinata (Buch.-Ham ex D. Don) H. Gross Philadelphus tomentosus Wall. ex G. Don * Phleum alpinum L. Phlomis macrophylla Wall. ex Benth. P. rotata Benth. ex Hook.f. * Phyllanthus clarkei Hook.f. Physospermopsis obtusiuscula (DC.) C. Norman Picrorhiza scrophulariiflora Pennell Pieris formosa (Wall.) D. Don * Pilea racemosa (Royle) Tuyama P. symmeria Wedd. Pinguicola alpina L. * Pinus wallichiana A. B. Jackson Piptanthus nepalensis (Hook.) D. Don * Platanthera clavigera Lindl. P. latilabris Lindl. Pleione hookeriana (Lind1.) J. Moore Pleurospermum angelicoides (DC.) C. B. Clarke P. apiolens C. B. Clarke * P. benthamii (DC.) C. B. Clarke P. brunonis (DC.) C. B. Clarke P. aff. corydalifolium Aitch. & Hemsl. + P. dentatum (DC.) C. B. Clarke P. hookeri C. B. Clarke P. rotundatum C. B. Clarke Poa annua L. P. hirtiglumis Hook.f. P. pagophila Bor P. polycolea Stapf Podophyllum hexandrum Royle Polygonatum cf. cirrhifolium (Wall.) Royle * P. hookeri Baker * * T rept/scap aG bulb aG bulb T rept aH rept a G rhiz a NP caesp a H caesp aH scap aHros a Ch frut/a NP caesp aGrad a G rhiz/ros s NP scap a G bulb a G rhiz aH ros b Mes P scap a Mi P scap aG bulb aG bulb fac a GE bulb a Grad a Grad a Grad a Grad a Grad a Grad a Grad a Grad T caesp a H caesp a H caesp a H caesp a G bulb a G rhiz a G rhiz PLANTS SOUTH OF MT EVEREST I ng. 251 Z.778 H.60 Mc.396 M.917, W.244 Zit3t P.s.n., Sch.2036, 2039 P-s.n: L.2089, S.4630, Z.733, 1829 P.G.311, W.428 By.585, P.R.310, Sch.2190, 2280, Z.776 BL.2130, M.1055, P3G:247., 2.50), 375, Sm.s.n. M.819 M.1036 BL.2051, Mc.280 By.911, M.857, P.s.n., ZAs19 51937 W.333 BL.2014 Beer 25369, S.660 M.s.n. By.1175, P.s.n., Sch.2082 H.65 H.66, L.0171 M.s.n. (Gottingen, Bot. Garden) F.8-0; M.910a, 937, 987, 1016, Psen: By.644, L.7, M.921 M.847 By.656 By.663, 831 By.449, M.1085, Sm.s.n. By.663, 831, M.910b Sm.s.n. W.243 M.1103, Sm.s.n. P.G.306 Z.618 §.7124 BL.2066, By.880, M.941, Sm.s.n. 252 P. cf. kingianum Colett & Hemsl. 4+ P. verticillatum (L.) All. Ponerorchis chusua (D. Don) S06 * Potentilla argyrophylla Wall. ex Lehm. * P. argyrophylla Wall. ex Lehm. var. argyrophylla P. argyrophylla var. atrosanguinea (Lodd.) Hook.f. * P. caliginosa Sojak P. coriandrifolia D. Don * P. cuneata Wall. ex Lehm. * P. eriocarpa Wall. ex Lehm. * P. fruticosa var. rigida (Wall. ex Lehm.) Wolf * P. fulgens Wall. ex Hook. P. griffithii Hook.f. P. cf. kleiniana Wight P. leschenaultiana Seringe P. leuconota D. Don P. microphylla D. Don * P. microphylla var. achilleifolia Hook.f. P. microphylla D. Don var. microphylla P. monanthes Wall. ex Lehm. P. peduncularis D. Don * P. polyphylla Wall. ex Lehm. P. saundersiana Royle Primula alpicola var. luna (Stapf) W. W. Sm. & Hetcher + P. atrodentata W. W. Sm. * P. buryana Balf.f. P. calderiana subsp. strumosa (Balf.f. & Cooper) A. J. Richards * P. capitata Hook.f. subsp. capitata P. capitata subsp. crispata (Balf.f. & W. W. Sm.) W. W. Sm. & Forrest P. caveana W. W. Smith * P. concinna Watt P. denticulata Sm. * P. deuteronana Craib G. MIEHE a G rhiz a G rhiz a G bulb aHsem aHsem aHsem aHsem aHsem aH rept/f Ch herb pulv a G rhiz a Ch frut aHsem a H caesp aHsem aHsem aHsem af Ch herb pulv af Ch herb pulv af Ch herb pulv aHsem aHsem aH sem/G rhiz aHsem aHros aHros aHros aHros aHros aHros aHros aHros aHros aHros SY IX VI, IX HV, Vil VII, VIII, X VII, X av, Vo VII IX VIII, IX LO. IX P.s.n. By.878, M.926, Mc.303, Z..669 By.226 D.403, S.4774 P.G.255, Z.586 f.M. Sw.258 D.363, Mc.383, P.35 By.196, D.329, M.874, P.33, 34, Sch.2035, W.439, Z.468, 613 D.425, W.438, Sm.s.n. BL .2159; 2533, 7.21, D.329, M.852, 1079, P.G.294, G.724, S.4752, Sch.2033, 2034, Z.613 M.821, 878, Z.1659 P.G.338 M.888 By.942, D.348 BL.2053, M.974, W.418 By.1097, Mc.251 D.349, M.1038, Z.619, 665a M.967, 1073 BL.2108, Z.580a D.364, Z.656 By.233, $.4689, 4787 By.45, 208, M.1025, Z.500, 628, 685, 1610a Sch.2015 By.417, S.4817, Z.269, 340, 512, 563, Sm.s.n. BL.2054, Mc.378 Beer 25354, Mc.281 Sch.2012 Beer 25408 BL.2123A, P.G.283, S.6596, W.261, Z.1508, Sm.s.n. M.1059, 1109, Z.1505, Sm.s.n. By.129, P.s.n., Sch.2198, 2292, Z.370, 402, 505 §.4530, Z.815c P. dickieana Watt P. gambeliana Watt P. geraniifolia Hook.f. * P. glabra Klatt P. glomerata Pax. * P. gracilipes Craib P. hookeri Watt P. listeri King ex Hook.f. P. macrophylla D. Don * P. macrophylla var. moorcroftiana (Wall. ex Klatt) W. W. Sm. & Fletcher P. megalocarpa Hara P. muscoides Hook.f. ex Watt P. obliqua W. W. Sm. * P. primulina (Spreng.) Hara * P. pulchra Watt P. reticulata Wall. P. rotundifolia Wall. * P. scapigera (Hook.f.) Craib P. sikkimensis Hook.f. * P. stuartii Wall. * P. tanneri subsp. nepalensis (W. W. Sm.) A. J. Richards P. tenuiloba (Watt) Pax P. uniflora Klatt P. walshii Craib * P. wollastonii Balf.f. * Prunella vulgaris L. Prunus rufa Hook.f. * Pseudostellaria heterantha (Maxim.) Pax P. heterantha var. himalaica Ohwi Pueraria peduncularis (Benth. ) Grah. Pyrola sikkimensis Kiisa Quercus glauca Thunb. var. glauca * Q. lanata Sm. * Q. semecarpifolia Sm. Rabdosia lophanthoides var. gerardiana (Benth.) Hara PLANTS SOUTH OF MT EVEREST aHros aHros aHros a Hros a Hros a Hros aH ros aHros aHros aHros aHros af Ch herb pulv aHros aHros a Hros aHros aHros aHros a Hros aHros aHros T ros aHros aHros aHros T scap a Mi P scap a G bulb aG bulb a/e d PL suff aHsem s Mes P scap s Mes P scap s Mes P scap aH scap V, Vil Ill, V I, Ill Vv TV. Vil Il VIII VIII, 1X VI, IX Ill, 1V, VIII 253 Beer 25352 By.390, BL.2037, Z.815 BL.2173, Mc.390, Z.726, 1747 Z.814 By.352, 649, dH.2983A, H.57, M.894, 991, Z.816, Sm.s.n. S.4526 Beer 25374 S.6975, Z.779 BL.2097, W.278 Z.1463 Beer 25373, BL.2192, dH.2986, M.984, P..G.284, 2.1565, Sm.s.n. Mc.389 Beer 25353, Mc.382 By.442, BL.2094, D.383, 432, M.972, W.259, 2.5.1: S022, 001-50: BL.2039, $.7090 Mc.406 By.436, D.361, M.909, Mc.278, L.0187, Z.506, 636, 1700, Sm.s.n. Z.800 By.547, 592, 605, 636, 649, 1500, D.369, Sm.s.n. Mc.386 Z.799 BL.2191 Beer 25360, S.1649 P18:0., Ze282,00l,02e, Sm.s.n. BL.2101, 2136, By.274, D.402, H.18, M.842, P.R.312, 2.1626, 1807, Sm.s.n. By.585, W.414 Sch.2256 By.828 D.344 dH.2911A Z.483 D.320 S.4539 f.M. Z.1879 254 Ranunculus adoxifolius Hand.-Mazz. * R. brotherusii Freyn var. brotherusii * R. brotherusii var. tanguticus (Maxim.) Tamura R. diffusus DC. R. hirtellus Royle ex D. Don * R. hyperboreus Rottb. R. oreionannos Marq. & Shaw R. pedatifidus Sm. + R. pegaeus Hand.-Mazz. R. pulchellus C. A. Meyer * R. pulchellus var. ellipticus Tamura + R. pulchellus var. sericeus Hook.f. & Thoms. R. pulchellus var. stracheyanus (Maxim.) Hand.-Mazz. Ranunculus sp. Rheum acuminatum Hook.f. & Thoms. ex Hook. Rhodiola amabilis (H. Ohba) H. Ohba R. bupleuroides (Wall. ex Hook.f. & Thoms.) Fu * R. crenulata (Hook.f. & Thoms.) H. Ohba R. cretinii (R.-Hamet) H. Ohba R. discolor (Franch.) Fu R. fastigiata (Hook.f. & Thoms.) Fu * R. himalensis (D. Don) Fu * R. humilis (Hook.f. & Thoms. emend. R.-Hamet) Fu R. ovatisepala (R.-Hamet) Fu R. prainii (R.-Hamet) H. Ohba R. quadrifida (Pallas) Fischer & Meyer * R. sinuata (Royle ex Edgew.) Fu R. wallichiana (Hook.) Fu * Rhododendron anthopogon * D. Don var. anthopogon R. anthopogon var. hypenanthum (Balf.f.) Hara R. arboreum Smith * R. arboreum var. cinnamomeum (Wall. ex G. Don) Lindl. R. barbatum Wall. ex G. Don * a Hros a Hros aH ros aH ros aH ros aHros a Hros aHsem G. MIEHE a H ros, fac hyd aH ros aH ros aH ros aH ros aHros aH sem/G rad a Ch 1 succ a Ch1 succ/G rad a Ch1 succ a Ch1 succ/G rhiz/caesp a Ch | succ/caesp a Ch | succ/a f Ch herb pulv a Ch] succ a Ch 1 succ e Ch! succ a Ch 1 succ a Ch1succ a Ch 1 succ a Ch1succ s NP caesp/s Ch frut pulv s NP caesp/s Ch frut pulv s Mes P scap s Mes P scap s Mi P caesp Ill VII, VIII VIII, 1X I, i VII, Vill VIII VIII Vv VIII VIII Ill VIII IX IV IV VII, VIII X Ill, 1V IV, VIII VII Vil I-III I-III I, Il S.7091 D.338, H.38, Sm.s.n. W.202 Z.745 W.458, Z.582, Sm.s.n. Sm.s.n. BL.2080, 2189, Sm.s.n. P.R.284 BL.2146, W.197 By.230, BL.2076, M.1113, 1115, P.s.n. $.7092 Sm.s.n. BL.2106, D.406, S.7110, S.7134 By.780 Beer 25514, BL.2154, By.180, M.1017, 1044, P.G.233, Sm.s.n. By.779, M.1063, W.168, Sm.s.n. BL.2077 L.0211, Mc.349, S.4615 BL.2084, 2085, 2095, By.49, M.1105, Mc.339, Z.1514 By.408, D.378, Sch.2180, Z.701, Sm.s.n. M.863 H.13, 68, M.841 D.396, 397 M.1062, 1096 By.702, 793, 794, M.838, Mc.270, P.103, Z.1628 H.27, 58, M.923, P.R.301, s.n. M.951 Z.562, 1552, 1812, Sm.s.n. By.30, 31, 972, M.823, Sch.2297, Z..498 Mc.46, Sch.2187, Z.271, 371, 497, 1784 By.554, 605, 606, S.4525, Z..1848 R. campanulatum var. wallichii (Hook.f.) Hook.f. * R. campylocarpum Hook.f. R. ciliatum Hook.f. * R. cinnabarinum Hook.f. * R. dalhousiae Hook.f. * R. fulgens Hook.f. R. hodgsonii Hook.f. * R. lepidotum Wall. ex G. Don * R. lepidotum var. eleagnoides (Hook.f.) Hook.f. R. lindleyiT. Moore * R. nivale Hook.f. * R. setosum D. Don * R. triflorum Hook.f. * R. wightii Hook.f. * Ribes glaciale Wall. R. griffithii Hook.f. & Thoms. * R. himalense Royle ex Decne. R. luridum Hook.f. & Thoms. R. orientale Desf. * R. takare D. Don * R. takare forma desmocarpum Hara R. vilmorinii Janczew. Rodgersia nepalensis T. A. Cope ex Cullen * Rosa macrophylla Lindl. * R. sericea Lindl. * Roscoea alpina Royle * Rubus biflorus Buch.-Ham. ex Smith R. fockeanus Kurz R. hypargyrus var. niveus (Wall. ex G. Don) Hara R. nepalensis (Hook.f.) Kuntze * R. paniculatus Smith * Rumex nepalensis Spreng. * Sabia campanulata Wall. ex Roxb. * PLANTS SOUTH OF MT EVEREST s Mi P caesp I, Ill s Mi P caesp III s NP caesp/sChfrut I, II s MiP caesp I s NP caesp/fac E I s NP caesp/s Mi P Il caesp s Mi P caesp Ill s/a NP caesp Ill, V s/a NP caesp V, VII s NP caesp/fac E I s Ch frut/s Ch frut Vit be pulv sNPcaesp/sChfrut VII pulv sNPcaesp/sChfrut I, I s Mi P caesp Ill a NP caesp Div a NP caesp Il a NP caesp I a NP caesp Vil a NP caesp I a NP caesp III a NP caesp Il a NP caesp Vv e Grad II a NP caesp III, V a NP caesp Ill, V a Grad I-III, V a NP caesp ? a Ch frut rept Vv a NP caesp/a PL Vv frut a Ch frut rept I, I ad PL frut I,IV a Grad i, Vx ad PL frut I as Mc.47, S.4527, Sch.2252 By.100a, s.n., M.856, 931, Mc.49, Sch.2188, $.2252, 4602, Z.373, 412, 499 Mc.57, $.4613, Z.793 By.186, BL.2031, Mc.277, S.4532 Z.463 M.930, Mc.288, S.4524 By.685, 686, 968, Mc.48, $.4521, Sch.2343, W.411, 2.645 By.184, 296, 385, 659, 1008, 1109, D.347, M.853, 860, Sch.2281, Z.673, 1669, 1782, 1783B Sm.s.n. Mc.39, $.4531 By.915, D.430, M.996, Sch.2269, Z.523A, 538, 1484, 1811, Sm.s.n. By.688, M.950, Sch.2048, P.s.n., W.442, Z.504, 590, 1810, Sm.s.n. Mc.40, Sch.2048, S.4520, Z.431 Z.1784A M.920 S.4534 Z.414, 462 D.355 Z.616 L.0152, $.4619 By.5, dH.2976, L.0168, P.37 By.17, dH.2938, M.855, P.38, 39, R.278, 305 f.M., By.18, 327, 372 By.579 Z.692 dH.2990, M.904, P.40 dH.2931 dH.2931 By.215239),.0 8-0; Z.1720, Sm.s.n. BL.2018, Z.481, 489 256 Salix calyculata Hook.f. ex Anderss. * . daltoniana Anderss. . disperma Roxb. ex D. Don . eriostachya Wall. ex Anderss. . hylematica Schneid. . ef. insignis Anderss. + . lindleyana Wall. ex Anderss. . longifora Anderss. . luctuosa H. Lév. S. serpyllum Anderss. S. sikkimensis Anderss. * S. tetrasperma Roxb. Salix sp. Salix sp. Salvia campanulata Wall. ex Benth. * S. castanea Diels AN AN AnNN i) S. hians Royle ex Benth. * Sanguisorba diandra (Hook.f.) Nordborg Sanicula elata Buch.-Ham. ex D. Don Sarocococca hookeriana Baill. Satyrium ciliatum Lindl. S. nepalense D. Don * Saussurea auriculata (DC.) Sch. Bip. * S. candolleana (DC.) Sch. Bip. S. deltoidea (DC.) Sch. Bip. * S. gnaphalodes (Royle ex DC.) Sch. Bip. * S. gossypiphora D. Don * . graminifolia Wall. ex DC. * . hieracioides Hook.f. . leontodontoides (DC.) Sch. Bip. . leontodontoides var. filicifolia (Hook.f.) Hand.-Mazz. . nepalensis Spreng. * . simpsoniana (Field. & Gardn.) Lipschitz * . tridactyla Sch. Bip. ex Hook.f. . uniflora Wall. ex Sch. Bip. . wernerioides Sch. Bip. ex Hook.f. S. yakla C. B. Clarke Saussurea sp. Saxifraga andersonii Engl. * HAN AN ~ ~H H i) G. MIEHE a Ch frut rept a Ch frut/a NP caesp a Mi P scap a Ch frut rept a Ch frut rept a NP caesp a Ch frut rept/f pulv a Mi P scap a NP caesp a Ch frut rept a NP caesp a NP caesp a NP caesp a Mi P scap aHsem aH sem aH sem aH scap aHsem s NP caesp a G bulb a G bulb aHsem a H scap a H scap a G rad/H ros a H scap/G rad a H scap/g Ch herb pulv/G rad a G rad/H ros aH ros/G rad/f Ch herb pulv aHros a H ros a G rad/f Ch herb pulv/H scap a H scap/G rad a H scap/G rad af Ch herb pulv H ros/G rad a Grad/ros e g Ch herb pulv VII Ill, VII I VII VII IX Ill, Vil VI, IX, X IX, X B.5678, BL.2134, By.550, D.387, 388, M.924, 952, Sm.s.n. By.6, 446, D.457, M.1003, W.343, Z.376 Z.471 By.189, 531, W.311 By.543, D.354, 437, W.431, Z.1518 P.R.344 D.436, M.962, 990, 1005, Sm.s.n. Z.387 Z.388, 491 Z.1518 BL.2118, W.419 Z.479 dH.2987 M.922 By.204, S.7141, Z.371A, 374, 407, 654, 1813 By.973, P.s.n., Sch.2055, 2057 BL.2151, By.194, D.447, M.825 M.871, P.42 D.324 f.M., Sch.2083 By.646, H.64 dH.2881, H.69, P.26 By.622 P.R.352 Pesin: P.G.236, 256, 296 By.145, M.1000, 1046, 1084, P.G.218, 279, W.215, Z.1452, Sm.s.n. M.1061, P.s.n., Z.1450, Sm.s.n. P.R.335, W.212, Z.1476, Sm.s.n. By.558, 1135, M.953, 995, 1081, W.226, Sm.s.n. P.s.n. (f.Enum) Pan: M.1046A, S.6583, W.214, Z.1516, Sm.s.n. M.1047, Z.576, Sm.s.n. Beer 25390 M.1049, P.G.235, W.217, Sm.s.n. P.R.303, W.385 M.919 D.419, 443, Z.566, Sm.s.n. S. aristulata Hook.f. & Thoms. S. brachypoda D. Don * S. brunonis Wall. ex Seringe * S. cordigera Hook.f. & Thoms. S. diversifolia Wall. ex Seringe S. engleriana H. Smith * S. filicaulis Wall. ex Seringe S. gageana W. W. Smith S. georgei Anthony S. granulifera H. Smith S. hirculoides Decne. S. hispidula D. Don S. hookeri Engl. & Irmsch. S. jacquemontiana Decne. * S. lowndesii H. Smith S. lychnitis Hook.f. & Thoms. * S. cf. melanocentra Franch. * S. moorcroftiana (Seringe) Wall. ex Sternb. S. mucronulata Royle * S. mucronulata subsp. sikkimensis (Hulten) Hara S. nutans Hook.f. & Thoms. S. pallida Wall. ex Seringe S. palpebrata Hook.f. & Thoms. S. parnassifolia D. Don * S. perpusilla Hook.f. & Thoms. S. pilifera Hook.f. & Thoms. S. pseudo-pallida Engl. & Irmsch. * S. punctulata Engl. S. roylei H. Sm. * S. saginoides Hook.f. & Thoms. * S. serrula H. Smith S. sibirica L. * S. sikkimensis Engl. PLANTS SOUTH OF MT EVEREST g Hherb pulv e H scap e Hros H sem aHsem e Hros e H caesp/scap/f Ch herb pulv e Hros e g Chherb pulv a Hros a H ros/f Ch herb pulv aH caesp a H ros e H ros/g Ch herb pulv e H ros/rept/g Ch herb pulv/G rhiz af Ch herb pulv a Hros a g Ch herb pulv/H TOs e Hros e H ros/rept aH ros/Ch herb pulv aHsem aHsem aHsem e Hros/f Ch herb pulv e H ros/rept a Hros e Hros e f Ch herb pulv e g Chherb pulv e H rept/caesp a H ros/sem a H scap IV, VI IV VI, IX IV, IX, X Vv X IV, VII IV, VII, [X X Ill, V IX, X IV, VI, Ix IX, X VIII-x 21 By.673, H.27, 52, M.843, 907, P.G.214, W.211, Z.1479, 1526, 1615B, Sm.s.n. Beer 25412, By.489, 666, H.7, M.843, Sm.s.n., P.s.n., W.287, Z.1876 By.487, M.1018, P.R.350, W.338, Sm.s.n. Beer 10003 P.48, Z.1292, 1805 Mc.332, S.7106, Sm.s.n. By.604, M.865, Z.s.n. Beer 9581, M.1020 Sw.500 dH.2975, M.898, P.R.340, W.284, Z.1506, Sm.s.n. By.542, P.G.s.n., Sm.s.n. M.908, 1020, P.G.331, W.260 P.47, S.628, W.374 M.1086 Sw.421 M.1089, Mc.329, Sm.s.n. By.450, Sm.s.n. M.1026, Z.1549 By.443, D.417, M.1013, Z.1615a, Sm.s.n. BL.2115, Mc.356 H.30, W.207, 287, Z.1526, 1623, Sm.s.n. BL 2156, By.515,,D.352, 440, L.2202, Mc.297, S.4780 Beer 9554, D.461 By.391, 781, M.810, Z.1805 M.1064, 1106, Mc.340, P°G:237, W232 BL.2143, Sm.s.n. BL.2095, D.398, 438, M.1012, Mc.333, Sm.s.n. BL.2145, M.1078, P.G.274, W.180, Sm.s.n. P.s.n. BL.2141, By.778, D.413, H.42, M.1040a, Mc.331, P.G.238, W.161, 223, 277, Sm.s.n. P.50 B.5777, By.512 Beer 25512, S.1644 258 S. stella-aurea Hook.f. & Thoms. S. stenophylla subsp. hoffmeisteri (Klotzsch) Hara * S. strigosa Wall. ex Seringe * Schisandra grandiflora (Wall.) Hook.f. & Thoms. * Scopolia stramonifolia (Wall.) Shrestha * Scrophularia pauciflora Benth. Sedum chauveaudii R.-Hamet S. gagei R.-Hamet S. oreades (Decne.) R.-Hamet * S. triactina Berger S. trullipetalum Hook.f. & Thoms. Selinum candollii DC. S. cortioides C. Norman S. tenuifolium Wall. ex C. B. Clarke * Senecio alatus Wall. ex DC. S. albopurpureus Kitam. S. cappa Buch.-Ham. ex D. Don S. chrysanthemoides DC. * S. diversifolius Wall. ex DC. S. graciliflorus DC. * S. lancifera J. R. Drummond S. rufinervis DC. S. scandens Buch.-Ham. ex D. Don * S. triligulatus Buch.-Ham. ex D. Don S. wallichii DC. Sibbaldia cuneata Hornem. ex Kuntze * S. micropetala (D. Don) Hand.-Mazz. S. perpusilloides (W. W. Smith) Hand.-Mazz. S. purpurea Royle * Silene aff. gonosperma (Ruprt.) Bocquet S. gonosperma subsp. himalayensis (Rohrb.) Bocquet var. himalayensis S. nigrescens (Edgew.) Majumdar S. setisperma Majumdar S. thomsonii Majumdar Sinocarum sp. Skimmia laureola (DC.) Sieb. & Zucc. ex Walp. * Smilacina oleracea (Baker) Hook.f. * S. purpurea Wall. * Smilax menispermoidea DC. Sorbus cuspidata (Spach) Hedlund * G. MIEHE e H ros/rept/f Ch herb pulv e H ros/rept aH sem/rept ad PL frut . aH scap a G rhiz/scap a Ch 1 succ a H scap a Ch | succ a Ch | succ/caesp a Ch] succ a Grad a Grad a Grad aHsem a Hros a Ch suff a H sem/scap aH scap a G rad/scap aHsem aHsem a Ch suff/L a Ch suff aHsem ag Chherb pulv/G rhiz aHsem a G rhiz/f Ch herb ulv a G rhiz/f Ch herb pulv a H caesp a H caesp a G rhiz/H sem a H scap aH sem a Grad s NP caesp a G rhiz a G rhiz sd PL frut a Mes P scap X IV BL.2145, M.1100, Sm.s.n. M.1118 By.660, M.906, P.51, Z.1660 B.5679 Z..1679 H.22, M.1020, Z.1609 By. 620 Mc.270 M.862, 960, W.273, 301 By.13, 245, 655 M.1001a M.848, 883a Z.1634, 1770 M.1052 Z.1740A, 1845 By.197, 399 M.827, P.s.n. Z.1760 M.829 P.s.n. dH.2909, Sch.2363, W.383, Z.1740 dH.2901, $.4487 Z.1662 Z.658, 695 BL.2171, D.346, Z.789 Mc.379 M.1039, 1077, Mc.377, Z.568, Sm.s.n. M.1027 BL.2187, By.432, Sm.s.n. M.1006, Sm.s.n. Sm.s.n. M.968, P.s.n. M.s.n. S.2064, 4477 Mc.267, S.4627 Mc.291 Mc.55, 423, 1783 M.824 S. foliolosa (Wall.) Spach S. hedlundii Schneid. S. microphylla Wenzig * S. ursina (Wenzig) Decne. Soroseris hookeriana (C. B. Clarke) Stebb. * S. hookeriana (C. B. Clarke) Stebb. subsp. hookeriana S. pumila Stebb. Spathoglottis ixioides (D. Don) Lindl. * Spiraea arcuata Hook.f. * S. bella Sims S. canescens D. Don * Spiranthes sinensis (Pers.) Ames * Stellaria congestiflora Hara S. decumbens Edgew. * S. decumbens var. polyantha Edgew. & Hook.f. S. decumbens var. pulvinata Edgew. & Hook.f. S. graminea L. S. lanata Hook.f. ex Edgew. & Hook.f. S. sikkimensis Hook.f. ex Edgew. & Hook.f. Streptopus simplex D. Don * Sunipia bicolor Lindl. Swertia cuneata D. Don S. dilatata C. B. Clarke S. hookeri C. B. Clarke * S. kingii Hook.f. S. multicaulis D. Don Sympagis petiolaris (Nees) Bremek. Symplocos paniculata (Thunb.) Mig. * S. sumuntia Buch.-Ham. ex D. Don Tanacetum atkinsonii (C. B. Clarke) Kitam. * T. gossypinum Hook.f. & Thoms. ex C. B. Clarke * Taraxacum mitalii van Soest T. monochlamydeum Hand.-Mazz. T. pseudostenoceras van Soest Taxus baccata subsp. wallichiana (Zucc.) Pilger PLANTS SOUTH OF MT EVEREST a Mi P scap Ill a Mi P scap I a Mi P scap Il a Mi P scap III a Hros X aH ros Xx a H ros Xx obl. a E Ch succ IV a NP caesp III, V a NP caesp II a NP caesp II a Grad Vv aH rept VIII, X eHrept/efChherb X pulv aH rept xX af Ch herb pulv/a VIII, X H rept aH rept IV, VII aH rept ? T rept 4 a Grad Ill obl. e E Ch succ I aH sem VII, VIII aH scap Vv aH sem/G rad VII a H sem/G rad Vv a G rad/H caesp VII II s/aMiPscap/caesp_ I s MiP scap I aHsem V, VI a H ros/scap IX, X a G rad/H ros V-VII a G rad/H ros Vv a G rad/H ros V, VI b N/Mi P scap I 259 M.893 P:8.0: By.7, 839, 1155, D.451, S.6580, 7143, 7147 Mc.276 M.998, W.254 P.G.247, Sms: 0. By.382, P.G.217, Sm.s.n. By.366, M.806, L.016 By.439, M.933, P.s.n., Sm.s.n. S.4628 S.7122 H.62-64, M.835 W.246 M.1022, Sm.s.n. BL.2028, By.387, M.971 M.s.n. M.897 W.373 W.380 By.138, BL.2017, L.0158, vi 88 f Sch.2312 M.949, 1057, P.160, R.328 M.918 S.7087 L.0419, W.290 By.504, 524, D.339, H.44, P.G.209, G.285, W.234, Sm.s.n. Z.1743 S.4623 S.4483 P.s.n. By.1176, H.39, Sm.s.n., M.1031, 1070, P.G.211, G.295, Sch.2315, W.179, Z.1509 Z.270 Z;,1 163 Z.1603 f.M., Sch.2056, 2939, S.4496 260 Thalictrum alpinum L. * T. cultratum Wall. * T. elegans Wall. ex Royle T. reniforme Wall. * T. virgatum Hook.f. & Thoms. * Thermopsis barbata Royle * Thesium himalense Royle ex Edgew. Tiarella polyphylla D. Don Tofieldia himalaica Baker Trachydium aff. roylei Lindl. + Trigonotis multicaulis (DC.) Benth. ex C. B. Clarke T. ovalifolia (Wall.) Benth, ex C. B. Clarke T. rotundifolia (Wall. ex Benth.) Benth. ex C. B. Clarke * Triosteum himalayanum Wall. Triplostegia glandulifera Wall. ex DC. Tripogon filiformis Nees ex Steudel Tripterospermum volubile (D. Don) Hara Trisetum scitulum Bor T. spicatum (L.) K. Richter Trollius pumilus D. Don Tsuga dumosa (D. Don) Eichler * Typhonium diversifolium Wall. ex Schott * Urtica dioica L. Utricularia striatula Sm. Vaccinium dunalianum Wight V. glauco-album Hook.f. ex C. B. Clarke V. nummularia Hook.f. & Thoms. ex C. B. Clarke * V. retusum (Griff.) Hook.f. ex C. B. Clarke Valeriana hardwickii Wall. V. jatamansii Jones * V. stracheyi C. B. Clarke Verbascum thapsus L. Veronica cana Wall. ex Benth. V. cephaloides Pennell V. lanuginosa Benth. ex Hook.f. * V. himalensis D. Don V. umbelliformis Pennell Viburnum erubescens Wall. ex DG. V. grandiflorum Wall. ex DC. * G. MIEHE aHros aHsem aHsem aHsem aHsem aGrad aGrad aHsem a G rhiz aGrad a H caesp a H scap a H caesp a H scap a Grad a H caesp ael HL a H caesp a H caesp a H ros b Mes P scap a G rhiz a hyd H rept s E Ch frut s NP caesp s fac E Ch frut rept fac E/s NP caesp/Ch frut aH scap aHsem a H scap aHsem aH scap aH scap aH rept/G rhiz a H scap a H scap a Mi P caesp a NP caesp D.394, 400, BL.2111, Z.644, Sm.s.n. $.7140, Sm.s.n.° P.s.n., W.405 By.176, 191, M.890, S.7146 BL.2169, D.323, P.s.n., Sch.2099 By.511, 552, 949, dH.2956, Sch.2053, 2054, 2174, Z.1646 P.42, Z.501, 682 P.53, Z.448, 768 M.964 By.561 Beer 10016, S.653 Mc.357, S.7088 BL.2204, S.652 L.0654, P.s.n., Z.679 H.12, W.364 M.808, 817c Sch.2068 P.G.303 By.1071, M.1021a, P.R.349 BL.2080 f.M., Sch.2340, 2360 By.225 f.M. P.267 BL.2021 Sch.2163 S.4605, Z.266, 382 P.B.75, S.4475 By.320, D.358, Sch.2077, Z.1672 Z.263 ZaAisil By.182, Sch.2061, P.56, 57 P.260 W.191 BL.2176, M.1079, Mc.375, Sw.88-—99 V. nervosum D. Don * Vincetoxicum hirundinaria subsp. glaucum (Wall. ex Wight) Hara * Viola biflora L. * V. pilosa Blume V. thomsonii Oudemans Waldheimia glabra (Decne.) Regel * Youngia japonica (L.) DC. Y. racemifera (Hook.f.) Babcock & Stebb. Ypsilandra yunnanensis W. W. Smith & Jeffrey Zanthoxylum nepalense Babu * Z. oxyphyllum Edgew. * PLANTS SOUTH OF MT EVEREST a Mi P caesp Il a G rad/scap Vv a Hros II, III, V aH ros Ti aH ros I a H ros/G rhiz VIII, X aH sem Sea a Hsem/a G rad Ill a Grad V, VII a/s NP caesp II ad PL frut/NP II caesp Alphabetical list of ferns and fern allies Adiantum pedatum L. Anogramma leptophylla (L.) Link Araiostegia pulchra (Don) Copel. Arthromeris himalayensis (Hook.) Ching A. wallichiana (Sprengel) Ching Asplenium ensiforme Wall. ex Hook. & Grev. A. varians Wall. Athyrium drepanopterum (Kunze) A. Br. A. duthiei (Bedd.) Bedd. A. fimbriatum (Wall.) T. Moore A. mackinnoni (Hope) Christensen A. pectinatum (Wall. ex Mett.) Bedd. A. schimperi Moug. ex Fée A. schizochlamys Ching A. spinulosum (Maxim.) Milde Cheilanthes albomarginata C. B. Clarke C. farinosa (Forsk.) Kaulf. C. subvillosa Hook. Coniogramme affinis Hieron. C. caudata (Wall. ex Ettingsh.) Ching C. intermedia Hieron. Crypsinus erythrocarpus (Mett. ex Kuhn) Tagawa e rhiz III, V a caesp I a rhiz II a rhiz I a rhiz I e caesp I, Il a caesp Ill a/e caesp I a rhiz Ill a/e caesp I a caesp I, Il a rhiz I, Il a rhiz I a caesp Vv a rhiz Ill a caesp I a caesp a a caesp Vv e rhiz IV e rhiz IV e rhiz IV e rhiz I, Il 261 Mc.35, S.4523, Z.795 Sch.2249, Z.675, 714 By.110, W.366, Z.579, 652, 702, 1774, Sm.s.n. Onyon 10, $.4479, Z.397 S.4486 By.448, H.36, dH.2999, M.973, Mc.315, P.R.345, W.174, 181, Z.1453, Sm.s.n. Z.748 M.976, W.319 Beer 25356 Z.158 BL.2015, Z.465, 1695, 1866 Z.1688, 1779 Z.1754 Z.674 dH.2888, Z.256 Z.1871 Z.1896 WsdOd Z.1019 Z.646 Z.1870 Z.1682 Z.1665 Z.678, 1752 M.885A, 912A, 927 Z.677, 1685, 1794 Z.761 W.354 Z.1823 Lilee Z.454 Z.1749 Z.1689B 262 Cryptogramma brunoniana Wall. ex Hook. & Grev. Cystopteris dickieana Sim C. filix-fragilis (L.) Bernh. Drynaria mollis Bedd. Dryopteris acuto-dentata Ching D. barbigera (T. Moore ex Hook.) Kuntze subsp. barbigera D. chrysocoma (Christ) C. Chr. D. fibrillosa (C. B. Clarke) Hand.-Mazz. D. marginata (Wall. ex C. B. Clarke) Christ D. serratodentata (Bedd.) Hayata D. sublacera Christ D. wallichiana (Spreng.) N. Hylander Equisetum diffusum D. Don Gymnocarpium oyamense (Bak.) Ching Lepisorus clathratus (C. B. Clarke) Ching L. contortus (Christ) Ching L. kashyapii (Mehra) Mehra L. loriformis (Wall.) Ching L. variabilis Ching & S. K. Wu Leucostegia delavayi (Bedd.) Ching ex Christensen Lycopodium clavatum L. L. veitchii Christ Mecodium polyanthos (Sw.) Copel. Notholaena marantae (L.) Desv. Oleandra wallichii (Hook.) C. Presl Onychuum contiguum Hope Osmunda claytonia var. vestita (Wall.) Milde. Phymatodes crenatopinnata (C. B. Clarke) Ching P. ebenipes (Hook.) Ching P. malacodon (Hook.) Ching P. stewartii (Bedd.) Ching Plagiogyria communis Ching Polypodium amoenum Wall. ex Mett. P. microrhizoma C. B. Clarke ex Baker G. MIEHE a caesp e caesp a caesp a rhiz a caesp a caesp a caesp a caesp a caesp a caesp a caesp e caesp a G rhiz a rhiz e/a rhiz e rhiz e rhiz e rhiz e/a rhiz a rhiz b Ch herb rept b Ch herb rept a rhiz a caesp e rhiz a rhiz e caesp e rhiz e rhiz e rhiz e rhiz e caesp e/a rhiz e rhiz Ill, V, VII Vil Vv IVS yy. VI VI II, V1 1 TIE TV. VII I I, Il 1,18 I Ill 1 BEY I, WI-—V PA EV Vv M.986, W.256, Z.1562 M.1010A, Z.1618 Z.639, Sm.s.n. Z.1639 M.885B, 961 M.886 Z.1642, 1684 Z.1681, 1776 Z.273 M.1010B, Z.1542, 1567 Z.385, 488 Z.455, 1832 W.430 LA? Z.507, 638 Z.311, 396 Z.721 W.387, Z.427, 1890 W.352 Z.703, 1895 dH.2965, M.840, 885c, 925, Z.265, 629, 1650, 1713 Z.754 dH.2882, Z.1658 Z.715 Z.1689B (Candollea 15: 206 [1956]) Z.715B, 1689 W.298, Z.637, 1334, 1624 Z.738 Z.257, 460 2.153 M.912B PLANTS SOUTH OF MT EVEREST 263 Polystichum acanthophyllum e caesp Ein Z.384 (Franchet) Christ P. discretum (Don) Diels a caesp I Z.455 P. lachenense (Hook.) Bedd. e caesp VI-X M.1010c, Z.1449, 1595 P. mehrae Fraser-Jenkins & e caesp II N.9 Khullar P. neolobatum Nakai e caesp I Z.1680 P. nepalense (Spreng.) C. Chr. a caesp I, 11, 1V Z.1742 P. shensoense Christ a caesp III M.975 P. stimulans C. Presl e caesp IV Z.1666, 1737 P. thomsonii (Hook.f.) Bedd. a caesp IV W.356, Z.399, 747, 826b, 1736 Pteridium aquilinum (L.) Kuhn a rhiz II, II, V M.s.n. Pteris cretica L. e rhiz I Z.1738, 1847 P. dactylina Hook. e rhiz I Z.261 4805-1137 P. subquinata Wall. ex Hope e caesp IV Z.1826 P. wallichiana J. Agardh e caesp 16s Z.1873 Selaginella chrysocaulos (Hook.f. I A We & Grev.) Spring Thelypteris levingei (C. B. a rhiz Ill Z.1686 Clarke) Ching Tricholepidium normale (D. e rhiz IV Z.1750 Don) Ching Vittaria flexuosa Fée e rhiz IV Z.1748 Woodsia andersonii (Bedd.) a caesp ? W.355 Christ W. elongata Hook. a caesp IV Z.680, 1751, 1775, 1809 W. lanosa Hook. a caesp Z W.358 Acknowledgements In the summer and autumn 1982 I had the opportunity to accompany Professor Dr Matthias Kuhle, Department of Geography, University of Géttingen, during a three months expedition to the Khumbu Himal. I have to thank him first for giving me the chance to carry out field work and to collect plants. The expedition was sponsored by the DFG. My own small collection of plants was determined during a four weeks stay at BM in December 1982 and January 1983 and in March and April 1985, supported by the British Council, by the kind permission of the Keeper of Botany, Mr J. F. M. Cannon, and the Head of the General Herbarium, Mr A. O. Chater, to whom I am highly indebted. Only by the help of all members of staff and the clearly arranged unique collection of the herbarium I was able to check all life forms of plants collected in the Khumbu Himal and the adjacent valleys. In many questions I was able to profit of the experience of Mr J. D. A. Stainton. Special thanks are also extended to Dr E. Launert for his help in every situation, and to Mr J. F. M. Cannon and Mr J. R. Laundon for their considerable editorial efforts. References and other relevant literature Academia Sinica, Comprehensive Observation Team. 1967. The vegetation of central Tibet. Beijing. Washington, DC. Ahrendt, L. W. A. 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 3. — Berberidaceae. Candollea 15: 153-155. Alston, A. H. G. & Bonner, C. E. B. 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 5. — Pteridophyta. Candollea 15: 193-220. Andrews, C. 1983. Photographs and notes on tourism and deforestation in the Solu Khumbu, Nepal. Mount. Res. Dev. 3: 182-185. Baehni, C. 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 2. — Cruciferae. Candollea 15: 149-151. 264 G. MIEHE —— 1958. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 13. — Juglandaceae, Santalaceae, Thymelaceae et Saxifragaceae. Candollea 16: 215-227. —— 1959. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 16. — Euphorbiacées et Buxacées. Candollea 17: 69-70. ——., Bonner, C. E. B. & Vautier, S. 1951. Plantes récoltées par le Dr Wyss- -Dunant au cours de r expédition suisse 4 l’Himalaya en 1949. Candollea 13: 213-236. Bajracharya, D. 1983. Deforestation in the food/fuel context. Historical perspectives from Nepal. Mount. Res. Dev. 3: 227-240. Beer, L. List of seed collections, Kasuwa, Barun, Iswa Khola, autumn 1975. Typescript, BM, London. ——, Lancaster, C. R. & Morris, D. Seed collection Barun, Kasuwa, Iswa Khola, autumn 1971. Typescript, BM, London. Bernardi, L. 1963. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 18. - Monochlamydeae, Dialypetalae et Metachlamydeae. Candollea 18: 243-283. Bjonness, I.-M. 1980a. Animal husbandry and grazing, a conservation and management problem in * Sagarmatha National Park, Nepal. Norsk geogr. Tidsskr. 34: 59-76. — 1980b. Ecological conflicts and economic dependency on tourist trekking in Sagarmatha (Mt. Everest) National Park, Nepal. An alternative approach to park planning. Norsk geogr. Tidsskr. 34: 119-138. —— 1983. External economic dependency and changing human adjustment to marginal environment in the high Himalaya, Nepal. Mount. Res. Dev. 3: 263-272. Boesch, H. 1964. Zwei Jahre Wetterbeobachtungen in Ostnepal (1961-1963). Geographica helv. 3: 170-178. Bonner, C. E. B. 1959. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 15. — Ericaceae. Candollea 17: 61-67. Bor, N. L. 1958. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 7. - Gramineae. Candollea 16: 137-142. Brockmann-Jerosch, H. 1936. Futterlaubbaume und 2 bets apnatinen Ber. d. Schweizer. Bot. Ges. 46: 594-613. Byers, A. C. & Byers, E. A. 1984. UNN, Nepal Mountain Hazards Mapping Project, Phase II, Khumbu (Mt. Everest Basin) Nepal. Typescript, Dept. of Geography, University of Colorado, Boulder. Caine, N., Ives, J. D., Kienholz, H. & Messerli, B. 1982. A buried podsol near Namche Bazar, Solu-Khumbu, Nepal. Mount. Res. Dev. 2: 405-406. Cannon, M. J. & Cannon, J. F. M. 1984. A revision of the Morinaceae (Magnoliophyta-Dipsacales). Bull. Br. Mus. nat. Hist. (Bot.) 12: 1-35. Chang, D. H. S. 1981. The vegetation zonation of the Tibetan plateau. Mount. Res. Dev. 1: 29-48. Coburn, B. A. 1983. Managing a Himalayan world heritage site. Nature Resources 19 (3): 229-240. Dhar, O. N. & Narayanan, J. 1965. A study of precipitation distribution in the neighbourhood of Mt. Everest. Ind. J. Met. Geophys. 16: 229-240. Dierl, W. 1978. Grundziige einer 6kologischen Tiergeographie der Schwarmer Ostnepals (Lepidoptera: Sphingidae). Khumbu-Himal 3: 313-359. Disselhorst, G. 1968. Beitrage zur Okologie der V6gel Zentral- und Ost-Nepals. Khumbu Himal 2: 1-417. Dittmann, E. 1970. Statistische Untersuchungen zur Struktur der Niederschlage in Nepal. Khumbu Himal 7: 47-60. Dobremez, J.-F., Jest, C., Toffin, G., Varnatian, M.-C. &Vigny, F. 1974. Carte écologique du Népal 1:250000. Région Kathmandu- Everest. Paris. —— & Shakya, P. R. 1977. Carte écologique du Népal. Biratnagar-Kanchenjunga 1:250,000. Paris. Liste des echantillons collectés au Népal. Mission été 1970. Université Grenoble I, Laboratoire du Biologie Végétale. Typescript, BM, London. Dyhrenfurth, G. O. 1959. Die bergsteigerische ErschlieBung der Mt. Everest Gruppe. Jn T. Hagen (Ed.) Mount Everest. Aufbau, Erforschung und Bevélkerung des Mt. Everest Gebietes: 113-168. Ziirich. Ellenberg, H. 1954. Steppenheide und Waldweide. Ein vegetationskundlicher Beitrag zur Siedlungs- und Landschaftsgeschichte. Erdkunde 8: 188-194. —— 1956. Grundlagen der SAE nS ener ene I. InH. Walter, Einfiihrung in die Phytologie 4: 1-136. Stuttgart. — 1966. Leben und Kampf an den Baumgrenzen der Erde. Naturw. Rdsch. Stutt. 19: 133-139. — 1978. Die Vegetation Mitteleuropas mit den Alpen in 6kologischer Sicht. Stuttgart. — & Mueller-Dombois, D. 1967.°A key to Raunkiaer plant life forms with revised subdivisions. Ber. Geobot. Inst. ETH. Stiftung Riibel 37: 56-73. PLANTS SOUTH OF MT EVEREST 265 Fraser-Jenkins, C. R. 1984. An introduction to fern genera of the Indian subcontinent. Bull. Br. Mus. nat. Hist. (Bot.) 12: 37-76. Fiirer-Haimendorf, C. V. 1959. Die Sherpas des Khumbu-Gebietes. Jn T. Hagen (Ed.) Mount Everest, Aufbau, Erforschung und Bevélkerung des Mt. Everest Gebietes: 169-218. Zurich. —— 1964. The Sherpas of Nepal. Buddhist highlanders. London. Funke, F. W. 1969. Religidses Leben der Sherpa. Khumbu Himal 9: 1-369. Grierson, A. J. C. & Long, D. G. 1983. Flora of Bhutan. Including a record of plants from Sikkim. Edinburgh. Gruber, U. F. 1969. Tiergeographische, Skologische und bionomische Untersuchungen an kleinen Saugetieren in Ost-Nepal. Khumbu Himal 3: 197-312. Hackel, H., Hackl, K. & Kraus, H. 1970. Tagesgange des Energiehaushaltes der Erdoberflache auf der Alp Chukhung im Gebiet des Mt. Everest. Khumbu Himal 7: 71-133. Haffner, W. 1967. Ostnepal-Grundziige des vertikalen Landschaftsaufbaus. Khumbu Himal 1: 389-426. —— 1979. Nepal Himalaya. Untersuchungen zum vertikalen Landschaftsaufbau Zentral- und Ostnepals. Erdwissenschaftliche Forschung 12. Hagen, T. 1969. Report on the geological survey of Nepal. Preliminary reconnaissance. Denkschriften d. Schweizer. Nat. forsch. Ges. 86 (1). Hara, H. (Ed.) 1966. The flora of eastern Himalaya. 1. Report. Results of the botanical expedition to eastern Himalaya organized by the University of Tokyo 1960 and 1963. Tokyo. — (Ed.) 1968. Photo album of plants of eastern Himalaya. Tokyo. —— (Ed.) 1971. The flora of eastern Himalaya. 2. Report. Results of the Botanical expedition to eastern Himalaya organized by the University of Tokyo 1960 and 1963. Tokyo. ——., Chater, A. O. & Williams, L. H. J. 1982. An enumeration of the flowering plants of Nepal 3. London. ——,, Stearn, W. T. & Williams, L. H. J. 1978. An enumeration of the flowering plants of Nepal 1. London. ——, Tuyama, T., Murata, G., Kanai, H. & Togashi, M. 1963. Spring flora of Sikkim Himalaya. Osaka. —— & Williams, L. H. J. 1979. An enumeration of the flowering plants of Nepal 2. London. Heuberger, H. 1956. Der Weg zum Tscho Oyu. Kulturgeographische Beobachtungen in Ostnepal. Mitt. Geogr. Ges. Wien 98: 1-28. —— 1976. Kulturgeographische Stockwerke und Wanderbewegungen im Hochgebirge am Beispiel des Mt. Everest-Gebietes. 40. Dt. Geographentag, Tagungsbericht wiss. Abh. 771-774. Higuchi, K. (Ed.) 1984. Meteorological data in Khumbu and Shorong Himal, east Nepal (revised tables). J. Jap. Soc. Snow Ice: 132-134. Hinrichsen, D., Lucas, P. H. C., Coburn, B. & Upreti, B. N. 1983. Saving Sagarmatha. Ambio 12: 203-205. Holtmeier, F.-K. 1974. Geodkologische Beobachtungen und Studien an der subarktischen und alpinen Waldgrenze in vergleichender Sicht. Erdwissenschaftliche Forschung 8. Hornstein, F. v. 1951. Wald und Mensch. Waldgeschichte des Alpenvorlandes Deutschlands, Osterreichs und der Schweiz. Ravensburg. Inoue, J. 1976. Climate of Khumbu Himal. Seppyo, J. Jap. Soc. Snow Ice 38: 66-73. Joshi, D. P. 1982. The climate of Namche Bazar: a bioclimatic analysis. Mount. Res. Dev. 2: 399-403. Kanai, H., Shakya, P. R. & Shresta, T. B. 1975. Vegetation survey of central Nepal. Jn Ohashi (Ed.), The flora of eastern Himalaya 2: 415-423. Tokyo. Kitamura, S. 1967. A list of Compositae collected in Nepal by M. Numata and K. Yoda, Himalayan expedition of Chiba University, 1963. J. College Arts Sci. Chiba University 5: 85-86. Kleinert, C. 1983. Siedlung und Umwelt im zentralen Himalya. Jn U. Schweinfurth (Ed.) Geoecological Research 4. Wiesbaden. Komarkova, Wald, Jones, Byers & Byers. 1985. Catalog of Nepalese collections, Annapurna, Dhaulagiri, Everest. Typescript, University of Colorado Herbarium, Boulder. Kraus, H. 1964. Das Klima von Nepal. Khumbu Himal 1: 301-321. Kuhle, M. 1978. Obergrenze von Frostbodenerscheinungen. Zeitschr. Geomorph. I, 22: 350-356. —— 1984a. Auswertungen von Spacelabaufnahmen aus Hochasien. Geogr. Rdsch. 36: 298-301. —— 1984b. Hanglabilitat durch Rutschungen und Solifluktion im Verhaltnis zum Pflanzenkleid in den Alpen, den Abruzzen und im Himalaya. Entwicklung und léndlicher raum 18 (3): 3-7. Larcher, W. 1963. Zur Frage des Zusammenhanges zwischen Austrocknungsresistenz und Frostharte bei Immergriinen. Protoplasma 57: 569-587. Lidén, M. 1985. List of Corydalis specimens of the BM. Typescript, Inst. for Systematic Botany, University of Géteborg, Goteborg. Limberg, W. 1973. Transhumante Wanderhirten im Nepal-Himalaya. Erdwissenschaftliche Forschungen 5: 182-184. 266 G. MIEHE — 1982. Untersuchungen tiber Besiedlung, Landbesitz und Feldbau im Solu-Khumbu (Mount Everest- Gebiet, Ost-Nepal). Khumbu Himal 12: 1-154. Ludlow, F. & Stearn, W. T. 1975. New Himalayan and Tibetan species of Corydalis (Papaveraceae). Bull. Br. Mus. nat. Hist. (Bot.) 5: 45-69. Maie, M. & Shismada, K. 1965. Meteorological data in the Numbur area. Jn M. Numata (Ed.), Ecological study and mountaineering of Mt. Numbur in eastern Nepal 1963: 124-154. Chiba. Mehra, P. N. & Bir, S. S. 1964. Pteridophytic flora of Darjeeling and Sikkim Himalayas. Res. Bull. Panjab Univ. II, 15 (1-2): 69-182. Meusel, H., Jager, E. & Weinert, E. 1965, 1978. Vergleichende Chorologie der zentraleuropdischen Flora. Text and maps. 2 vols. Jena. Miehe, G. 1982. Vegetationsgeographische Untersuchungen im Dhaulagiri- und Annapurna-Himalaya. Dissertationes Botanicae 66 (1-2). Vaduz. —— 1984a. Vegetationsgrenzen im extremen und multizonalen Hochgebirge (zentraler Himalaya). Erdkunde 38: 268-277. 1984b. Waldnutzung im Himalaya, Beispiel Dhauladhar, Himachal Pradesh, Indien. Praxis Geog- raphie 10: 36-40. in press. Neue Befunde zur Vegetationshéhenstufung am Mt. Everest. 45. Dt. Geographentag, Tagungsbericht wiss. Abh. Stuttgart. Mitsudera, M. & Numata, M. 1967. Meteorology of eastern Himalaya. J. Coll. Arts Sci. Chiba Univ. 5: 75-83. Miller, F. 1958. Acht Monate Gletscher- und Bodenforschung im Everestgebiet. Berge Welt 12: 199-216. Murata, G. 1963. A list of plants collected by the Chiba University Rolwaling Himal Expeditions, 1963. Acta phytotax. geobot. Kyoto 25: 107-118, 26: 77-88, 27: 29-34, 106-113. — 1965. A list of Labiatae, collected by M. Numata and K. Yoda, Himalayan expedition of Chiba University 1963. J. Coll. Arts Sci. Chiba Univ. 4: 309. —— 1967. A List of Sympetalae (exclude Compositae and Labiatae) collected by M. Numata and K. aot the Chiba University Rolwaling Himal Expedition, 1963. J. Coll. Arts Sci. Chiba Univ. 5: 87-92. Nishida, M. 1966. Pteridophyta collected by Dr. M. Numata and Dr. K. Yoda, Chiba University Scientific Expedition to Nepal Himalaya 1963. J. Coll. Arts Sci. Chiba Univ. 4: 571-581. —— 1968. A supplementary note of Himalayan Pteridophytes. Ann. Rep. Foreign Students Coll. Chiba Univ. 3: 43-45. —— 1974. Pteridophyta collected by Mr. P. R. Shakya and Mr. M. Ohsawa, Himalayan expeditions of Chiba University 1971. J. Coll. Arts Sci. Chiba Univ. B7: 25-33. Numata, M. 1965a. Grassland vegetation in eastern Nepal. Jn M. Numata (Ed.), Ecological study and mountaineering of Mt. Numbur in eastern Nepal: 74—94. Chiba. —— 1965b. Weed flora and communities in eastern Nepal. Jn M. Numata (Ed.), Ecological study and mountaineering of Mt. Numbur in eastern Nepal: 98-99. Chiba. —— 1966. Vegetation and conservation in eastern Nepal. J. Coll. Arts Sci. Chiba Univ. 4: 559-569. —— 1967. Notes on botanical trip in eastern Nepal. J. Coll. Arts Sci. Chiba Univ. 5: 57-74, 243-258. —— 1983a. Ecological studies in the Nepal Himalayas. Jn M. Numata (Ed.), Biota and ecology of eastern Nepal: 1-13. Chiba. — 1983b. Semi-natural pastures and their management. Jn M. Numata (Ed.), Biota and ecology of eastern Nepal: 455-465. Chiba. —— 1983c. Structure and dynamics of vegetation in eastern Nepal. Chiba. Ohashi, H. (Ed.) The flora of eastern Himalaya, 3. Report. Results of the botanical expeditions to eastern Himalaya in 1972. Tokyo. ——., Ohsawa, M. & Shakya, P. R. 1973. On the occurence of deciduous broad-leaved forests in the cold temperate zone of the humid Himalayas in eastern Nepal. Jap. J. Ecol. 23: 218-228. Okuri, J. 1965. List of Cyperaceae and Gramineae collected by M. Numata, Himalayan expedition of Chiba University. J. Coll. Arts Sci. Chiba Univ. 4: 315-318. Oppitz, M. 1968. Geschichte und Sozialordnung des Sherpa. Khumbu Himal 8: 1-170. Pawson, I. G., Stanford, D. D. & Adams, V. A. 1984. Effects of modernization on the Khumbu region of Nepal: changes in population structure, 1970-1982. Mount. Res. Dev. 4: 73-81. Poelt, J. Verzeichnis der Aufsammlungen J. Poelt, Nepal 1962. Zusammengestellt von A. Schreiber. Typescript, Bayerische Staatssammlung Miinchen. Polunin, O. & Stainton, J. D. A. 1984. Flowers of the Himalaya. Oxford. Rathjens, C. 1982. Geographie des Hochgebirges. 1. Der Naturraum. Stuttgart. Rauh, W. 1939. Uber polsterfoérmigen Wuchs. Nova Acta Leopoldina II, 7 (49): 267-508. PLANTS SOUTH OF MT EVEREST 267 Rolla, S. R. 1963. The Indian Cho Oyo Expedition, 1958: observations of a botanist member. J. Bombay nat. Hist. Soc. 60: 400-409. Schilling, A. D. 1975. Everest trek. Typescript, BM, London. —— 1976. Everest trek. Autumn 1976. Typescript, BM, London. —— 1977a. Everest trek 1977. Typescript, BM, London. —— 1977b. Autumn trek 1977. Typescript BM, London. —— 1978a. Everest trek 1978. Typescript, BM, London. —— 1978b. The roads to Everest. J. R. hort. Soc. 103: 259-264, 309-314. —— 1981. East Nepal (Kangchenjunga area) collections. Autumn 1981. Typescript, BM. London. Schneider, E. 1965. Begleitworte zur Karte Khumbu Himal I und zur Namensgebung. Khumbu Himal 1: 430-446. Schwarz, R. 1983. Simulationsstudien zur Theorie der oberen Baumgrenze. Erdkunde 37: 1-11. Schweinfurth, U. 1957. Die horizontale und vertikale Verbreitung der Vegetation im Himalaya. Bonn. —— 1983. Man’s impact on vegetation and landscape in the Himalayas. Jn W. Holzner, M.J. A. Werger & I. Ikusima, Man’s impact on vegetation: 297-309. Den Haag. —— 1984. The Himalaya: complexity of a mountain system manifested by its vegetation. Mount. Res. Dev. 4: 339-344. Smith, G. F. 1985. Provisional list of plants collected above 4500 m (15000') in Khumbu in July-August 1984. Typescript, Gottingen. Smith, H. 1958, 1960. Saxifraga in the Himalaya. Bull. Br. Mus. nat. Hist. (Bot.) 2: 83-129, 227-260. Stainton, J. D. A. 1963. Field notes east Nepal 1963. Typescript, BM, London. —— 1964. Notes on a journey in east and central Nepal 1964 with brief notes on a journey in east Nepal 1956 and in central Nepal. Typescript, BM, London. — 1972. Forests of Nepal. London. Sutton, S. 1978. Plant collectors in Nepal. Jn H. Hara, W. T. Stearn & L. H. J. Williams, An enumeration of the flowering plants of Nepal 1: 13-21. London. Swan, L. W. Plants collected in the Barun Khola by L. W. Swan in 1954. Typescript, BM, London. —— 1961. The ecology of the high Himalayas. Scient. Am. 205 (4): 68-78. Tamura, M. 1967. Ranunculaceous plants collected by Chiba University Rolwaling Himal expedition in 1963. J. Coll. Arts Sci. Chiba Univ. 5: 141-142. Teschke, G. C. 1977. Anthropologie der Sherpa. Khumbu Himal 11: 1-183. Tibet Autonomous Region Scientific and Technical Committee. 1980. An enumeration of the vascular plants of Xizang (Tibet). Tilman, H. W. 1938a. Mount Everest. Cambridge. —— 1938b. The Mount Everest expedition of 1938. Geographical J. 92: 481-498. —— 1952. Nepal Himalaya. Cambridge. Troll, C. 1964. Die klimatische und vegetationsgeographische Gliederung des Himalaya-Systems. Khum- bu Himal 1: 353-388. Vautier, S. 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952-1954 (partie botanique). 6. — Polygonaceae. Candollea 15: 221-228. —— 1959. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 14. Labiatae. Candollea 17: 41-52. Walter, H. & Walter, E. 1953. Das Gesetz der relativen Standortskonstanz, das Wesen der Pflan- zengesellschaften. Ber. dt. bot. Ges. 66: 227-235. —— & Breckle, S. W. 1983. Okologische Grundlagen in globaler Sicht. Stuttgart. —— —— 1984. Spezielle Okologie der tropischen und subtropischen Zonen. Stuttgart. — & Lieth, H. 1966. Klimadiagramm-Weltatlas. Jena. Weibel, R. 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952-1954 (partie botanique). 4. — Primulaceae. Candollea 15: 157-165. —— 1958a. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 8. — Crassulaceae. Candollea 16: 143-145. —— 1958b. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 9.-— Compositae. Candollea 16: 179-194. —— 1958c. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 10. — Violaceae. Candollea 16: 195-196. —— 1958d. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 11. — Caprifoliaceae. Candollea 16: 197-199. — 1960. Résultat des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 17. — Melastomaceae, Myrsinaceae et Campanulaceae. Candollea 17: 81-85. 268 G. MIEHE Weigel, J. 1969. Systematische Ubersicht tiber die Insektenfresser und Nager Nepals nebst Bemerkungen zur Tiergeographie. Khumbu Himal 3: 149-196. Weischet, W. 1965. Der tropisch-konvektive und auBertropisch-advektive Typ der vertikalen Nieder- schlagsverteilung. Erdkunde 19: 6-14. Wraber, T. 1972. Determination list (Barun Khola, Dudh Kosi). Typescript, BM, London. Yoda, K. 1967, 1968. A preliminary survey of the forest vegetation of eastern Nepal. J. Coll Arts Sci. Chiba Univ. 5: 99-140, 277-302. Zimmermann, A. 1953. Pflanzen an den obersten Grenzen der Vegetation. Berge Welt 8: 130-136. —— 1955. Expedition Gaurisankar 1954. List of collected plants. Typescript, BM, London. —— 1956. Résultats des expéditions scientifiques genevoises au Népal en 1952 et 1954 (partie botanique). 1. —Itinéraires. Candollea 15: 127-147. Maps: Khumbu-Himal (Nepal) 1:50000. 2.Ed. 1978 Freytag-Berndt & Artaria: Wien. (= Forschungsunterneh- men Nepal Himalaya. E. Schneider, F. Ebster). Shorong/Hinku 1:50000. 2.Ed. 1979. Freytag-Berndt & Artaria: Wien. (= Nepal Kartenwerk der Arbeitsgemeinschaft fiir Vergleichende Hochgebirgsforschung 5). Khumbu Area, Nepal. Mountain Hazards Index Map. 1: 50000. 1985. Hallwag: Bern. (The United Nations University, Highland-Lowland Interactive Systems Project. Field survey by A. Zimmermann, M. Bichsel, H. Kienholz). British Museum (Natural History) An Enumeration of the Flowering Plants of Nepal H. Hara, W. T. Stearn, A. O. Chater & L. H. J. Williams Vol. 1. The Gymnosperms and Monocotyledons 1978, 275 X 215mm, 154pp, 7 figs. Paperback. 0 565 00777 7 £30.00 Vol. 2. The Dicotyledons (Part). 1979, 220pp. 0 565 00810 2 £30.00 Vol. 3. The Dicotyledons (Part). 1982, 226pp, 1 fig. 0 565 00854 4 £35.00 Titles to be published in Volume 16 Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) By N. K. B. Robson The lichen genus Ramalina in Australia By G. N. Stevens An annotated list of vascular plants collected in the valleys south of Mt Everest By G. Miehe Title to be announced Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk Printed in Great Britain by Henry Ling Ltd, Dorchester Bulletin of the British Museum (Natural History) Further genera of the Biddulphiaceae (diatoms) with interlocking linking spines Robert Ross & Patricia A. Sims Botany series Vol 16 No 4 26 November 1987 The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, and an Historical series. Papers in the Bulletin are primarily the results of research carried out on the unique and ever-growing collections of the Museum, both by the scientific staff of the Museum and by specialists from elsewhere who make use of the Museum’s resources. Many of the papers are works of reference that will remain indispensable for years to come. Parts are published at irregular intervals as they become ready, each is complete in itself, _ available separately, and individually priced. Volumes contain about 300 pages and several volumes may appear within a calendar year. Subscriptions may be placed for one or more of the series on either an Annual or Per Volume basis. Prices yary according to the contents of the individual parts. Orders and enquiries should be sent to: Publications Sales, British Museum (Natural History), Cromwell Road, London SW7 5BD, England. World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.) © British Museum (Natural History), 1987 The Botany series is edited in the Museum’s Department of Botany ___..... Keeper of Botany: Mr J. F. M. Cannon oe a 1 RAY ic i ey Editor of Bulletin: Mr J. R. Laundon | Po EEE tip eel fs, Assistant Editor: Dr A. J. Harrington | / \ ‘ Editor’s Assistant: - Miss M. J. Short | eee > ZNO 1087 his Lov NOUV 1764 6 seid a a 5 ‘ *. ) i i Fs + J : tn a baw kor er se | ti) so” See ISBN 0 565 08017 2 ISSN 0068-2292 Botany series Vol 16 No 4 pp 269-311 British Museum (Natural History) Cromwell Road London SW7 5BD Issued 26 November 1987 Further genera of the Biddulphiaceae (diatoms) with interlocking linking spines Robert Ross and Patricia A. Sims Department of Botany, British Museum (Natural History), Cromwell Road, London SW7 5BD Contents Sa a RP ED ee aR OT aie cape Oe TP LN Ras AMR FELONS MAE Can The ARMAND 269 | WE UA GeO Yer CY et 7 DN A Pe eaters on te alae Aner a Uae Rien Oa EE AA acs One mR 00 NE Sy. od Coe a 269 PER OT SOGCIERE NS BBO TECOLOS 555 oi. 25 oot 50s 0 alien 354344 cahawegcaii aun tee ges teamed ecsaes cans 270 CUS git te a Rie Bre SS a deh CAA AR SAPPY OPE TE ORR yote REED Peete ORT Ente seiner a oF rf Taxonomic account IRR 8 5 Ty ae aan a AiO See BP ed eee Oa PR ERERET Cotte ore PASE FORAY Ee ae enor 271 II. Pseudorutilaria (Grove & Sturt ex De Toni & Levi) RSOVe A StUTCEN Loe 0 OMe ae eae eo an eect Pe es 276 PEE OM ONE ROSSER: Ae SIONS Ooo ice EN ee ee 284 PV Maluingd R Ross Ss Pe Ae SUNS ooo rhwicise ates era a et a 285 Vat TROVE Tes IRAISS Ge Bets SIONS ceo fa te ae ec eae Sa LUA cud ater uns 286 Vir Dextradonaior Ro Ross &.P.. As Sis ecceres hes as ee i es ee 288 TDP SANG SATA OS 0 9 Gk os cited saat fori uidlectene eset yruR es AR Ue UN alan TOR ay te ae meee nets 289 PCM IIR TOCCOA ES isha iia fossa vai ani cee aia e sna alee cboald cate ipa kc dare ewien a eas suse ae 292 Ur go) RA Sear ne ae 8 Ee Pe REIT SR aE TaN oA RE ORS Rh RRS REAR Oe RP at 292 POORER OUMONS OF PALE cree coo cote ee ronope es engin at ose sok SETH Coa reuhon a etter elas da nem Saniee otic 295 ir 1 Ea! ARR kg SO EER CRS Lee vat oir Om MA oe ye aye et err AMAT Rs ery nt mee mete aren nen 299 Synopsis Descriptions are given of a number of genera of the Biddulphiales, all except one with interlocking linking spines. Solium Heiberg, a monotypic genus, is distinguished from Trinacria. Pseudorutilaria (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni is close to Solium but has some unique characteristics. Five new species from the Eocene of the southern oceans are described. Monile R. Ross & P. A. Sims, a monotypic genus from the Miocene of southern France with similarities to both Pseudorutilaria and Biddulphia, is described as new. A new genus, Maluina R. Ross & P. A. Sims, is proposed for Hemiaulus centralitenuis R. Ross & P. A. Sims, and another monotypic new genus, Bonea R. Ross & P. A. Sims, is established for a tripolar diatom from the Eocene of the U.S.S.R. close to Strelnikovia R. Ross & P. A. Sims. New information about Dextradonator eximius (Grunow) R. Ross & P. A. Sims is presented. The position of genera within the family Biddulphiaceae is discussed but presents some difficulties, especially because of possible parallel evolution. The limits of the order Biddulphiales are also considered. Introduction In a recent paper (Ross & Sims, 1985) we dealt with a group of related genera in the diatom family Biddulphiaceae characterized by the possession of interlocking linking spines and well developed pseudocelli. In the course of preparing this we examined all the other members of the family with interlocking linking spines but concluded that none of these were close to the group with which we were dealing there. They belonged to four genera and our studies of these with the scanning electron microscope produced new information recorded here. One of these genera is Solium Heiberg (1863), for our examination of S. exsculptum Heiberg, its type species, showed that it should not be placed in Trinacria Heiberg (1863), the genus to which it was transferred by Hustedt (1930). Another is Pseudorutilaria (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni (1894), of which we found five undescribed species in Eocene material from the Bull. Br. Mus. Nat. Hist. (Bot.) 16 (4): 269-311 Issued 26 November 1987 270 R. ROSS AND P. A. SIMS Falkland Plateau in the South Atlantic. This genus is apparently close to Solium, but it has remarkable structural features on its central portion that clearly distinguish it. We re-examined the diatom that we earlier described as Hemiaulus centralitenuis R. Ross & P. A. Sims (in Ross, Sims & Hasle, 1977) and concluded that it could not be retained in Hemiaulus Heiberg (nom. cons. prop., see Ross, 1985), but that it was necessary to establish for it a new genus, which we have called Maluina. The fourth genus with interlocking linking spines is Dextradonator R. Ross & P. A. Sims (1980), and we present here additional information about D. eximius R. Ross & P. A. Sims. We also describe here another new genus, Bonea, based on a single new species with interlocking linking spines. This, which comes from the Eocene of the U.S.S.R., has many similarities with the genera we described in our previous paper, especially Strelnikovia R. Ross & P. A. Sims (1985), but it had not come to our notice at the time when that paper was prepared. Another new genus described here, Monile, is also based on a single new species, one that comes from the Miocene of southern France. It does not have interlocking linking spines, but it is included here because, although there is little to separate it from Biddulphia Gray (1821), it has many features in common with Pseudorutilaria, as well as some in which it resembles Kera- tophora Pantocsek (1889) and Strelnikovia R. Ross & P. A. Sims. Its affinities are puzzling. Sources of specimens and records In the distribution lists for the species considered in this paper there are many records taken from literature, and all the information available about these is to be found in the references given. We have endeavoured to give a complete account of the known distribution of the species dealt with, but we realize that our coverage of Russian literature is incomplete, although probably sufficient to give a reasonably adequate picture. Many of the specimens that we have ourselves examined come from localities of which we gave details in our recent paper (Ross & Sims, 1985). The additional localities from which we have seen specimens are discussed below. U.S.S.R. Northern Urals One of the samples that the British Museum (Natural History) received from Mr Brigger was labelled ‘N. Urals, U.S.S.R., Oligocene’. We have no further information about this material but presume that it is from one of the samples that are the basis of the records from the lower Oligocene of the northern and central Urals in Diatomovyi Analiz (Proshkina-Lavrenko et al., 1949a). Germany Greifswalder Oie, Rostock This is one of the localities from which Schulz (1927) reported lower Eocene diatoms. The specimen from this locality that we have seen was mounted by Hustedt and came from Schulz’s sample. Denmark All the specimens from Denmark that we have examined come from the Moler formation of lower Eocene (Ypresian) age. The precision with which the locality is indicated varies; the specimens are said to come from: Jutland, Glynggre, Skive, Limfjord, Island of Mors, Nykjobing, Ejerslev, Island of Fur. France Saint Laurent-la Vernéde, Gard The Miocene deposit from this locality is described by Lefébure (1935). We are not aware of any more precise information about its geological age. GENERA OF THE BIDDULPHIACEAE PAB Barbados We have examined specimens from two localities in Barbados additional to those from which we reported specimens in our previous paper (Ross & Sims, 1985). These are Bissex Hill and Malvern Hill. No information about the level of these samples within the Oceanic Beds, which range from middle Eocene to lower Miocene, is available to us. ‘Lamont E 8’ This is the only label on one sample received from Mr Brigger. Neither the California Atademy of Sciences nor the Lamont-Doherty Geological Observatory have any record giving locality data for this. Its content suggests very strongly that it comes from the middle to upper Eocene of the Falkland Plateau area. Mr Brigger received a number of such samples from the Lamont- Doherty Geological Observatory and in is presumably one of these. Terminology The terminology used in the descriptions of the genera and species is that proposed by Ross et al. (1979), supplemented by terms that we adopted in our most recent paper (Ross & Sims, 1985: 285) and one proposed by von Stosch & Simonsen (1984: 11), viz. ‘interstrial pores’. In addition we use here two new terms, ‘subocellus’ and ‘pseudo-antiligula’. In Solium Heiberg, Pseudorutilaria (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, and Monile R. Ross & P. A. Sims there is a cluster of perforations through the valve on the distal side of the upper parts of the elevations; in Monile this also extends onto the summit of the elevations (Fig. 64). Vela are present in these in Solium (Fig. 5) and Pseudorutilaria (Fig. 42), and probably also in Monile. There is, however, no transition between the small areolae in these clusters and the areolae on the rest of the valve, and in this they differ from the pseudocelli of such genera as Biddulphia S. F. Gray (Ross & Sims, 1971: pl. 1 fig. 2), Briggera R. Ross & P. A. Sims (1985: pl. 5 fig. 6), and Trigonium Cleve (Ross & Sims, 1971: pl. 2 fig. 6). On the other hand there is no thickened ring of silica around these clusters of small areolae such as surrounds the ocelli of Amphitetras Ehrenberg (Ross & Sims, 1971: pl. 5 fig. 2) and the other genera included in the Eupodiscaceae and the ocelluli of the Cymatosiraceae (Hasle et al., 1983). For these clusters of small areolae with a definite but unthickened margin we use the term ‘subocellus’. We have also used this term for the small cluster of more tightly and uniformly packed perforations on the distal side of the tip of the elevations in Maluina R. Ross & P. A. Sims (Fig. 68) as this cluster also has no thickened rim; it may well be, however, that this cluster consists of porelli, not areolae. In Maluina centralitenuis (R. Ross & P. A. Sims) R. Ross & P. A. Sims there is at each pole of the valve a downward projection of the mantle, i.e. one parallel to the pervalvar axis. Similar projections are present on the girdle bands of many species and are termed ‘antiligulae’. This term cannot, however, be used for structures that are part of the valve; hence we use for them the term ‘pseudo-antiligula’. Taxonomic account I. SOLIUM Heiberg, Consp. crit. Diat. Dan.: 52 (1863). Hemiaulus subgen. Solium (Heiberg) Van Heurck, Treat. Diat., transl. Baxter: 455 (1896). Frustules united in short, straight, inseparable chains, the pervalvar axis from about half to about twice as long as the diameter of the valve. Valves multipolar, with short projections at each pole cut off from the central portion by transverse sulci or internal costae. Elevations at each pole moderately stout, not expanded above, with a subocellus on the distal side near the tip, linking spines on the proximal side of the tip expanded above and interlocking, except those of the end cells of chains. A well developed marginal ridge between the elevations. Mantle with the upper part areolate and with anastomosing external costae that continue onto the elevations, the lower part hyaline and deeply concave. Areolae irregularly scattered on the central portion of the valve, decussate or irregularly scattered on the projections, each occluded by a vola. 272 R. ROSS AND P. A. SIMS Scattered hollow spines or occluded processes sometimes present on the central portion of the valve. One labiate process on the central portion of the valve near the base of a projection, its internal opening sessile and straight, its external part tubular. Type SPECIES: Solium exsculptum Heiberg (loc. cit.) Heiberg (1863) established the genera Trinacria and Solium for tripolar and quadripolar diatoms with the frustules united in chains by elevations with linking spines. Bipolar diatoms forming similar colonies he placed in Hemiaulus. Hemiaulus Ehrenberg (1844) has been shown to be a later synonym of Eucampia Ehrenberg (1839), and the conservation of Hemiaulus in the sense in which Heiberg and all subsequent authors have used it has recently been proposed (Ross, 1985). H. L. Smith (1872) did not consider that the number of poles was a sufficient basis for distinguishing genera, and he regarded Trinacria and Solium as synonyms of Hemiaulus, as did Van Heurck (1896), who treated Trinacria and Solium as subgenera of Hemiaulus, and Schiitt (1896), who was the only one of these authors to publish any new combinations under Hemiaulus. Most authors, however, maintained the three genera as separate until Hustedt (1930) united Trinacria and Solium under the name Trinacria, a genus he regarded as separate from Hemiaulus. We would agree with the view that the number of poles is not by itself a sufficient basis for generic separation, and we are not here concerned with the question of whether or not there are other differences on which to base a separation of Trinacria from Hemiaulus. The differences, however, between Solium exsculptum Heiberg and Trinacria regina Heiberg, the type species of the two genera, are such as to justify placing them in separate genera. In Trinacria regina the linking spines are never expanded above and interlocking, there is no subocellus but only a poorly developed pseudocellus, the mantle is vertical throughout, and the areolae are occluded by cribra; in Solium exsculptum.the linking spines are normally expanded above and interlock- ing, there is a well developed subocellus, the lower part of the mantle is strongly concave, and the areolae are occluded by volae. We attach particular importance to the presence and nature of the subocellus and the concavity of the lower part of the mantle as features separating Solium from Trinacria because they suggest, as we argue below (p. 291), that Solium may be more closely related to such members of the Eupodiscaceae as Eupodiscus J. W. Bailey (1851), Triceratium Ehrenb. (1839) sensu stricto, and Amphitetras Ehrenb. (1840), than to Biddulphia Gray (1821) or Hemiaulus Heiberg (1863). Trinacria, on the other hand, is much closer to Hemiaulus Heiberg and the majority of genera of the Biddulphiaceae. The genus to which Solium is most closely related is Pseudorutilaria (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni and we discuss their resemblances and differences after our description of that genus. Solium exsculptum Heiberg is the only known species of the genus. The two others that have been referred to it, S. jeremiae (Truan & Witt) De Toni (1894) and S. princeps M. Peragallo (in Tempére & Peragallo, 1912) are both quadripolar species of Trinacria. S. jeremiae was originally described as Trinacria jeremiae Truan & Witt (1888), but S. princeps cannot be called Trinacria princeps as that name is pre-occupied. It has, however, been described independently as a new species, Trinacria cristata Gombos (1982). 1. Solium exsculptum Heiberg, Consp. crit. Diat. Dan.: 52 (1863). (Pl. 1, Pl. 12 figs 86-88) Hemiaulus exsculptus (Heiberg) Schiitt in Engler & Prantl, Naturl. Pflanzenfam. 1 (1b): 97 (1896). Trinacria exsculpta (Heiberg) Hustedt in Rabenhorst, Krypt.-Fl. Deutschl., Osterr. u. d. Schweiz 7 (1): 889 (1930). Valves with 4 or 5 projections separated from the central portion by deep sulci; length of side in 4-polar specimens 30-110 wm, maximum diameter of 5-polar specimens 35—45 um, height at centre of valve 7-5—16 um; central portion with straight sides between the projections, weakly. domed; projections semi-circular to narrowly triangular with obtuse apices, slightly domed and rising towards the elevations. Elevations arising at the apices, not expanded above; height to GENERA OF THE BIDDULPHIACEAE Paes their tops 18—38 wm. Areolae poroid, 0-3—0-7 ~m in diameter, occluded by volae, with a raised external rim from which one or two very short cylindrical spines sometimes arise, irregularly scattered and 5-6 in 10 wm on the central portion of the valve, the projections, the elevations, and the upper part of the mantle, or decussate and rather closer, 7-8 in 10 wm, on the projections. Sparse scattered hollow spines or occluded processes sometimes present on the central portion of the valve of end cells. Subocelli on the upper part of the distal face of the elevations, circular and c. 4 wm in diameter, surrounded by a narrow hyaline unthickened rim, their areolae in radial rows, c. 32 in 10 wm. 3—6 linking spines on the proximal sides of the tips of the elevations, usually expanded above and interlocking with those of the sibling valve, sometimes (on end cells) tapering upwards, c. 4 wm long; c. 6 short spines c. 0-5 wm tall on the distal side of the tips of the elevations. One labiate process on the central portion of the valve 3-5 jm proximal to a sulcus, opposite the centre of the sulcus or, more frequently slightly offset in a clock-wise direction when the valve is viewed from its external side, its inner opening parallel to the sulcus, its outer tube almost vertical to strongly inclined distally with its tip touching the surface of the adjacent projection. forma exsculptum (Figs 1-3, 5, 6, 86, 88) Valves quadripolar. Paleocene. Indian Ocean, 14° 46-7’ S., 88° 54-4’ E., 4780 m depth. Vityaz station 6744-40 (Mukhina, 1974, 1976). Cape basin, South Atlantic Ocean, 29° 29-055’ S., 3° 30-74’ E., 4805 m depth. Deep Sea Drilling Project site 524 (Gombos, 1984). Paleocene — lower Eocene. Barents Sea, 100-500 m depth, between Franz-Josefs Land and Novaya Zembla (Grunow, 1884). (Date uncertain, more probably Paleocene. ) Lower Eocene. Jutland, Denmark (BM 45388, 45391, 74094; PH coll. Schulze arr. 3192; Benda, 1972). Glyng¢gre, Jutland, Denmark (BM 30847). Skive, Jutland, Denmark (BM 14744, 69223; Tempére & Peragallo, 1891, 1913). Harre, Jutland, Denmark. Drill hole (Fenner, 1985). Limfjord, Jutland, Denmark (BM 9840, 13588, 45644, 46747, 54353, 54354, 54515, coll. Adams Bess. 1154; Miller, 1969). Island of Mors, Denmark (BM 9856, 9857, 10411, 10413, 12927, 14297, 27798, 30863, 31535, 32718, 32719, 38292, 44007, 44008, 44009, 44509, 54351, 54357, 60821, 60822, 60823, 60824, 60825, 60826, 64013, 68374, 68375, coll. Adams Bess. 1788, F1096, GC3483, J587, J2952, SEM CB15.281-—289; PH coll. Boyer S-1-18, Gen. coll. 12724; BRM Hd1/12; Heiberg, 1863; Kitton, 1870; Cleve & MOller, 1878; Tempére & Peragallo, 1889b, 1907). Nykjobing, Island of Mors, Denmark (BM 12926, 31604, 54798, coll. Adams C514; Kitton, 1870; Cleve & Moller, 1878). Ejerslev, Island of Mors, Denmark (BM coll. Adams TS286, TS287, TS416). Island of Fur, Denmark (BM 10412, 14271, 14272, 27806, 35882, 45389, 53564, 60779, 64014, 68656, coll. Adams Bess. 358, F116, L22, SEM B28.988—991, B29.88—93; PH Gen coll. 10736; BRM Hd1/13; Heiberg, 1863; Kitton, 1870; Tempére & Peragallo, 1889a, 1911). Henmoor, Niedersachsen, Germany (Schulz, 1927). Dornbusch, Hiddensee, Rostock, Germany (Schulz, 1927). Greifswalder Oie, Rostock, Germany (BRM 428/70; Schulz, 1927). London clay, near base (locality not indicated), England (Shrubsole & Kitton, 1881). Prionersk, Kaliningrad oblast, U.S.S.R. (Strel’nikova, Kaplan & Travina, 1978). Middle Volga basin, U.S.S.R. (Glezer et al., 1974). Sengilei, Ulyanovsk oblast, U.S.S.R. (Glezer, 1966). Region of Atlym, Berezovo, Kazym (i.e. 62°-64° N., 65°-67° E.), Tyumen’sk oblast, U.S.S.R. (Rudkevich, Rubina & Permyakov, 1957; Galerkina, 1959). North-west of Fergana, Uzbek S.S.R., U.S.S.R. (Glezer et al., 1974). Middle Eocene. Region of Atlym, Berezovo, Kazym, Tyumen’sk oblast, U.S.S.R. (Rudkevich, Rubina & Permyakov, 1957; Galerkina, 1959). Junction of rivers Ob and Chemashevskoi, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Plain between the rivers Ob and Pur, Tyumen’sk oblast, U.S.S.R. (Paramonova, 1964). Tazovsk Peninsula, Tyumen’sk oblast, U.S.S.R (Paramonova, 1964). 274 R. ROSS AND P. A. SIMS Middle—upper Eocene. Norwegian Sea, North Atlantic Ocean, 67° 47-11’ N., 5° 23-26’ E., 1297 m depth. Deep Sea Drilling Project site 338 (Dzinoridze et al., 1978). Upper Eocene. Norwegian Sea, North Atlantic Ocean, 67° 12-47’ N., 6° 18-38’ E., 1217 mdepth. Deep Sea Drilling Project site 340 (Dzinoridze et al., 1978). Chelyabinsk oblast, U.S.S.R. (Proshkina-Lavrenko et al., 1949b). Sverdlovsk oblast, U.S.S.R. (Proshkina-Lavrenko et al., 1949a, b). Ivdel area, well 130, Sverdlovsk oblast, U.S.S.R. (Glezer, 1966). Serov area, wells 762, 771, Sverdlovsk oblast, U.S.S.R. (Glezer, 1966). Koptelovo area, well 255, Sverdlovsk oblast, U.S.S.R. (Glezer, 1966). Luchinkin, Sverdlovsk oblast, U.S.S.R. (Vozzhennikova, 1960). ‘Kamichev’, presumed to be Kamyshlov, Sverdlovsk oblast, U.S.S.R. (BM 35045, 65831, 65832, 78364, coll. Adams TS743, TS748, TS749, SEM B27.529-530, CB13.717—725, CB15.276—280, CB15.368-—375, 4106-4116; Cheneviére, 1934). Kimkya, basin of the river Severnaya Sos’va, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Tyumen’, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Lar’yak, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Region of Atlym, Berezovo, Kazym, Tyumen’sk oblast, U.S.S.R. (Rudkevich, Rubina & Permyakov, 1957; Galerkina, 1959). Junction of rivers Ob and Chemashevskoi, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Nyda, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Plain between the rivers Ob and Pur, Tyumen’sk oblast, U.S.S.R. (Paramonova, 1964). Basin of the Arka-Tab-Yakha, Tyumen’sk oblast, U.S.S.R. (Glezer, 1966). Tazovsk Peninsula, Tyumen’sk oblast, U.S.S.R. (Paramonova, 1964). Pudino, Tomsk oblast, U.S.S.R. (Vozzhennikova, 1960). Vasyugan, Tomsk oblast, U.S.S.R. (Glezer, 1966). Parabel’, Tomsk oblast, U.S.S.R. (Glezer, 1966). River Shet-Irgiz, bore no. 6, Kazakhstan, U.S.S.R. (Glezer, 1969). Region of Ayak-Kuduk, Kazakhstan, U.S.S.R. (Shibkova, 1968). Region of Kara Tau, Kazakhstan, U.S.S.R. (Shibkova, 1968). Region of Lake Balkash, Kazakhstan, U.S.S.R. (Shibkova, 1968). Upper Eocene—Oligocene. Tyumen’, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Lar’yak, Tyumen’sk oblast, U.S.S.R. (Vozzhennikova, 1960). Lower Oligocene. Eastern slopes of the northern and central Urals — Tyumen’sk and Sverdlovsk oblasts. (Proshkina-Lavrenko et al., 1949a). Northern Urals, Tyumen’sk oblast, U.S.S.R. (BM SEM CB15.264—265, CB15.273-275). forma pentagona Jousé* (Figs 4, 87) Solium exsculptum [‘exculptum’| var. pentagonalis Kitton ex Walker & Chase, Notes on some new and rare diatoms [I]: 5, pl. 1 fig. 10 (1886). Valves 5-polar. Paleocene. Cape Basin, South Atlantic Ocean, 29° 29-055’ S., 3° 30-74’ E., 4805 m depth. Deep Sea Drilling Project site 524 (Gombos, 1984). Lower Eocene. Isle of Mors, Denmark (BM coll. Adams F1096). Isle of Fur, Denmark (BM 10412; Kitton, 1870; Walker & Chase, 1886). Upper Eocene. ‘Kamichev’, presumed to be Kamyshlov, Sverdlovsk oblast, U.S.S.R. (BM 65833, SEM CB16.491-495, CB16.789-790). Plain between the rivers Ob and Pur, Tyumen’sk oblast, U.S.S.R. (Paramonova, 1964). Tazovsk Peninsula, Tyumen’sk oblast, U.S.S.R. (Paramonova, 1964). Lower Oligocene, Northern Urals, Tyumen’sk oblast, U.S.S.R. (BM SEM CB15.260-—263, CB15.271- 212). Although both H. L. Smith (1872) and Van Heurck (1896) included Solium in Hemiaulus, neither of them published the combination Hemiaulus exsculptus; this was first done by Schitt * We have not been able to trace the valid publication of this name, which is used by Paramonova (1964) but assume that it has been given valid publication in some work not available to us. If not, it is vaiidly published here by the reference given to a validly published varietal name. In this connexion, it should be pointed out that epithets do not have priority outside their own rank. GENERA OF THE BIDDULPHIACEAE 215 (1896), whose publication appeared later in the year than Van Heurck’s. The specific epithet has frequently been misspelled ‘exculptum’ or, in combination with Trinacria, ‘exculpta’. The 5-polar form of this species occurs in a number of localities along with the quadripolar form, although normally less commonly. It has never been recorded in the absence of the quadripolar form. As it does not seem to have a distinct distribution in either space or time, there is no case for separating it taxonomically at a higher rank than forma, but we have recognized it at this level as some authors may wish to note that it is present in material on which they are reporting. The size range of the specimens of this form that we have seen corresponds to the lower part only of the size range of the quadripolar specimens. Walker & Chase (1886) attribute the name Solium exsculptum var. pentagonalis to Kitton, saying ‘the variety with five angles is mentioned by Kitton’, but he did not publish the name; he merely remarked: ‘in the Fuur deposit valves with five angles are not uncommon’ (Kitton, 1870). This species shows little variation in any other characters than the number of poles except that specimens with prominent hollow spines or occluded processes occur only in certain localities. These structures are not present on specimens from the lower Eocene of Denmark but are usually to be found on those from ‘Kamichev’ (presumed to be Upper Eocene from Kamyshlov, Sverdlovsk oblast, U.S.S.R.). A specimen from material received from the late Mr A. L. Brigger and labelled ‘North Urals, U.S.S.R. Oligocene’ has what appears to be the base of one such spine or occluded process, its upper part, like the external part of the labiate process, having been broken off. The broken base of such a structure may be present on one of the specimens of the tropical Indian Ocean illustrated by Jousé (1977: tabl. 67 fig. 10) and one or two are present on the upper Eocene specimen from the western Siberian lowlands illustrated by Glezer et al. (1974: tabl. 23 fig. 4). None, however, are visible on the illustrations of specimens from the late Paleocene of Cape basin, South Atlantic (Gombos, 1984: pl. 8 figs 9-11), nor on the upper Eocene specimen from NW Siberia illustrated by Paramonova (1964: tabl. X fig. 2). The end valves (Figs 1, 2), i.e. those not inseparably linked to their sibling valves by interlocking linking spines, differ from the others not only in their linking spines but also in the inclination of the external part of their labiate processes. On the valves within the chain these are so strongly inclined that they almost or quite touch the surface of the projection distal to their point of insertion (Fig. 3); this makes them very difficult to detect when specimens in girdle view are examined with the light microscope. On end valves, however, the external tubes of the labiate processes are inclined at less than 45 degrees to the pervalvar axis, and they may be almost vertical (Fig. 1). The frequency with which end valves are encountered is such as to suggest that the chains in this species were short. We have seen two lower Eocene specimens from Glyng¢gre, Jutland, Denmark, that have the appearance of being initial valves. These have sides of 91 wm and 94 wm. The first has no costae delimiting three of its four projections whilst the other is somewhat asymmetrical, two adjacent projections being larger than the other two. Both of these specimens have the labiate process opposite the centre of a side of the valve instead of opposite a projection, the normal position. The specimens illustrated by A. Schmidt (1890: taf. 152 figs 24-25), of which he says ‘nach Witt’s Ansicht vielleicht zu Solium exsculptum zu rechnen’ are not this species but Trinacria cristata Gombos (1982). We have examined the specimens that are the originals of these figures (BM coll. Adams J3016— both specimens are mounted on this slide). Proshkina-Lavrenko et al. (1949b) treat Trinacria semseyi Pant. as a synonym of Solium exsculptum but this is a manifest error. T. semseyi, which comes from the upper Eocene of Kuznetsk, Penza oblast, U.S.S.R., isa tripolar species which has the characters of Trinacria, not Solium. The earliest records of this species are from the upper Paleocene of the South Atlantic (Gombos, 1984) and the tropical part of the Indian Ocean (Mukhina 1974, 1976). All the subsequent records are from northern Europe or adjacent seas, north-western Siberia, the central Volga basin, and Kazakhstan. It persisted throughout the Eocene in this area, and continued into the lower Oligocene in north-western Siberia. It is not rare in its lower Eocene localities nor in the upper Eocene from north-western Siberia, but elsewhere it is apparently infrequent in its occurrence. This rather peculiar distribution, in low latitudes in the Paleocene and in high northern latitudes from the lower Eocene to the Oligocene, seems more likely to bea 276 R. ROSS AND P. A. SIMS reflection of the inadequacy of the fossil record than a complete record of its history. It is presumably because it is only in north-western Siberia that it has been found to occur commonly in the upper Eocene that Fenner (1985) says that it is characteristic for the late Paleocene and earliest Eocene, whereas Glezer (in Glezer et al., 1974: 140) remarks that it did not decrease in the upper Eocene. A misprint in a previous paper (Ross, Sims & Hasle, 1977) that we have only recently detected suggests that Solium exsculptum occurs in Eocene samples from the Falkland Plateau taken by R/V Vema. It does not. The first line of the upper of the two paragraphs on p. 180 of that paper which begins: ‘Hemiaulus februatus and Trinacria exsculpta ...’ should read: ‘Hemiaulus centralitenuis, H. reflexispinosus, and H. prolongatus from. . .’ We have seen three specimens of this species labelled as coming from localities that we have not included when listing its distribution. These are: PH Gen. coll. 89190, Jackson’s Paddock, Oamaru, New Zealand; BM coll. Adams TS478, San Redondo, California, U.S.A. BRM Hd2/27, Santa Monica, California, U.S.A. As there is no other record of the occurrence of this species in any of these much studied deposits, we regard these specimens as strays and have not accepted these records. II. PPEUDORUTILARIA (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, Syl. Alg. 2: 854 (1894). Subgen. Pseudo-rutilaria Grove & Sturt in J. Quekett microsc. Club I, 2: 324 (1886), nom invalid (Arts 9 a ae le Rutilaria subgen. Pseudorutilaria Grove & Sturt ex De Toni & Levi in Notarisia 2: 349 (1887). Frustules rectangular, united in inseparable chains. Valves bipolar or tripolar, the bipolar valves having a straight or curved apical axis, with long projections each cut off from the central portion by a transverse sulcus or internal costa and crossed by a number of other sulci or internal costae; valve surface undulate. Elevations at each pole not very tall, not expanded above, with a clearly defined subocellus on their distal side, interlocking linking spines at their tip on the proximal side, and anastomosing low external costae on their lower part and on the mantle below them. Mantle with a deep furrow above a narrow hyaline band, the furrow usually continuous around the whole valve but sometimes present only where the valve margin is convex. Areolae irregularly arranged on the valve surface and the upper part of the mantle, each with a simple volate velum. A well-developed marginal ridge, normally bearing many strong marginal spines, these sometimes fused with those of the sibling valve. An irregular ring of occluded processes on the central portion of the valve more or less alternating with troughs with raised margins that clasp the inclined upper parts of the occluded processes of the sibling valve usually present but absent on end valves and often on the shortest valves. One or two labiate processes normally present on the central portion of the valve, more or less on the transapical axis in bipolar species, their internal openings straight or slightly curved and sessile, their external tubes stout and very similar to the occluded processes but reaching to the margin of the sibling valve. TYPE SPECIES: Pseudorutilaria monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni (loc. Sit), The first account of this genus was given by Grove & Sturt (1886). This was headed ‘Pseudo- rutilaria. Nov. subgenus’, but there was no positive indication of the genus to which it was to be attributed. The statement: ‘We have formed this subgenus for the reception of the species described below, which, while resembling Rutilaria in other characteristics, is without the central clasping process, so important a feature of that genus’, and the fact that the account of it immediately follows that of Rutilaria, make it plain that they intended it as a subgenus of that genus. However, the /nternational Code of Botanical Nomenclature provides that ‘the name of a subdivision of a genus is a combination of a generic name and a subdivisional epithet connected by aterm. . . denoting its rank’ (Art. 21.1) and that ‘a combination (autonyms excepted) is not validly published unless the author definitely indicates that the epithet or epithets concerned are to be used in that particular combination’ (Art. 33.1). Grove & Sturt did not give such an GENERA OF THE BIDDULPHIACEAE ME indication. They used the name ‘Ps.-rutilaria monile’ in the text for the one species that they included in this subgenus, and ‘Pseudo-rutilaria monile’ in the legend of the plate on which it is illustrated. The subgeneric name Rutilaria subgenus Pseudorutilaria was validly published by De Toni & Levi in 1887, as also was the specific name Rutilaria monile. They attributed both to Grove & Sturt and gave a Latin translation of Grove & Sturt’s account of the two taxa. De Toni (1894) later treated Pseudorutilaria as a genus, attributing this name also to Grove & Sturt. At the same time he validly published the combination Pseudorutilaria monile for the type species. Grunow (1887), Tempére & Peragallo (1890), and Schmidt (1893) used the name Pseudorutilaria monile earlier than this but did not give the generic name valid publication. The most striking and unusual feature of Pseudorutilaria is the ring of occluded processes and troughs with raised margins that clasp the distal ends of the occluded processes of the sibling valve. This feature is well shown in Fig. 18. The external tubes of the labiate processes resemble the occluded processes and are clasped by the raised margins of the troughs in the same way; they, however, reach the margin of the valve (see Fig. 32), whereas the occluded processes stop short of it. There are many points of similarity between this genus and Solium Heiberg: the subocelli of the two genera resemble each other closely; both have interlocking linking spines on the proximal sides of the elevations and low anastomosing external costae on the elevations and the upper part of the mantle, although these costae are less well developed in Pseudorutilaria than in Solium; in both there is a concavity of the mantle above a narrow hyaline rim; both have projections cut off from the central portion of the valve by sulci or internal costae, although in Pseudorutilaria the projections are crossed by additional transverse sulci or internal costae; both normally have one labiate process inserted towards the margin of the central portion of the valve, with an elongate sessile internal opening and a stout external tube. The two genera are clearly closely related. In some species of Pseudorutilaria, e.g. P. incompleta R. Ross & P. A. Sims, the projections are cut off from the central portion of the valve and are crossed by sulci, inward folds of the valve (Figs 39-41). In others, e.g. P. monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, all the transverse structures are solid internal costae with their free edges laterally expanded and level with the margins of the valve, which itself has an inwardly projecting rim (Figs 11, 12). In P. hendeyi R. Ross & P. A. Sims, however, the proximal transverse structures are sulci and the more distal ones are costae (Figs 48-50, 54). The only species until now added to the genus since its original publication is Pseudorutilaria monomembranacea Schrader in Schrader & Fenner (1976: 994). This species, however, has neither terminal elevations with subocelli and interlocking linking spines nor occluded processes and associated clasping troughs. Strel’nikova (in Dzinoridze et al., 1979: 63) erected for this species the new genus Praecymatosira, a genus which she considered was related to Cymatosira Grun., and this seems to be its correct position. In this paper we describe five new species of Pseudorutilaria from Eocene material from the south-western Atlantic. One of these is the species misidentified by Hajés (1976: 829) as P. monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni. Also, the diatom misidentified as Eunotogramma weissei var. productum Grove & Sturt by Fenner (1977: 519) is another of the species described here as new. The genus has so far been found only in middle to upper Eocene, or possibly lowest Oligocene, deposits in the south-western Atlantic, the Tasman Sea, and New Zealand. This suggests that its species may well prove to be useful stratigraphic markers. Key to the species a ae vO ER rsp ge) ia toi cs vg Sem cana idceiealgtc uxt4 tes Weta s eaeeeron. Wuevaded woe cate: 6. hendeyi(p. 283) 1b. Valves bipolar: ; 2a. Central portion pentagonal and with acrenate margin....................0656 5. incompleta(p. 282) 2b. Central portion circular and with an entire margin: 3a. Margins of the projections crenate or undulate: 278 R. ROSS AND P. A. SIMS 4a. Margins of the projections crenate, indented opposite every transverse costa SAR ER Ne eee ats ye eI eae ARSED UR SEPT y BRE EAS SRE TENA P 1. monile(p. 278) 4b. Margins of the projections undulate, undulations more widely spaced than the EE ANSVOISE COSUAC: stn. fone vs aes oy avs raes seas cease hoeee ees ee rere 4. hannai(p. 281) 3b. Margins of the projections with one gibbosity, elsewhere entire: 5a. Gibbosity of the margins of the projections adjacent to the central portion BN en, Cee oe Reno Oe a ERT ee eC NTT TURNS Pe ERATE 2. clavata(p. 279) 5b. Gibbosity of the margins of the projections about midway between the central PORTON ANG TNE APOK aos csc ions os scan gcatie oes vic ches resem ained tascen 3. nodosa (p. 280) 1. Pseudorutilaria monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, Syll. Alg. 2: 854 (1894). (Pl. 2, Pl. 12 fig. 89) ‘Pseudo-rutilaria’ monile Grove & Sturt in J. Quekett microsc. Club II, 2: 324, pl. 18 figs 7-8 (1886), nom. invalid (Arts 21.1, 33.1). Rutilaria monile Grove & Sturt ex De Toni & Levi in Notarisia 2: 349 (1887). Valve bipolar, straight, 75-190 wm long, 17-24 wm broad; central portion circular, separated from the projections by transverse costae; projections about half the breadth of the central portion at their proximal ends, tapering very slightly towards the apices, margins crenulate, indented opposite each transapical costa, costae 6—7 uw m apart, laterally expanded at their free edges, which are level with the inwardly projecting valve margin. Height to the top of the elevations 10-14 wm. Linking spines 3-4 on each elevation, extending rather more than half-way around the summits, c. 3 wm tall. Areolae poroid, with a slightly raised external rim, diameter c. 0-3 wm, scattered over the whole valve surface and the upper part of the mantle, sparser on the central portion than on the projections, where they are c. 8in 10 um. Ring of c. 6 occluded processes on the central portion, absent on end valves. A stout, flattened marginal spine with a serrate tip between each pair of marginal costae, the spines of sibling valves opposite and abutting against one another, but absent on end valves. Labiate process one, situated in the ring of occluded processes on the central portion, more or less on the transapical axis, the labiate processes of sibling pairs of valves on the same side. Middle Eocene. South-western Atlantic, Falkland Plateau, 49° 52-194’ S., 40° 50-713’ W., 1844 m depth. Deep Sea Drilling Project site 512 (Gombos, 1983). Upper Eocene. Oamaru, Otago Province, New Zealand (BM 9241, 9242, 11094, 11140, 11164, 11243, 33316 (Jackson’s Paddock), 33317 (Jackson’s Paddock), 46542, 46577, 46606, 46633, 52709, 52759, 52760, 60842, 63397, 63398, 64687, 65090, 66566, 71539 (lectotype), 74150, 74151, 76748, 77754 (Jackson’s Paddock), 77756, 77758 (Allen’s Farm), 77760, coll. Adams Bess. 979, C499, G80, G97 (Cormack’s Siding), G104 (Cormack’s Siding), G624, GC2614, GC3208, J548, J818 (Jackson’s Pad- dock), TS23, TS271, TS292, TS433 (Allen’s Farm), TS949 (Bain’s Farm, lower), SEM B4.414—415 (Division Hill), B4.425—428 (Division Hill), CB14.200—208, 44519-44537; BRM Ht1/1, Ht1/2, Ht1/3 (Jackson’s Paddock); Grove & Sturt, 1886; Schmidt, 1893; Tempére & Peragallo, 1890 (Bain’s Farm, Totara); Tempére & Peragallo, 1908 (Bain’s Farm); Laporte & Lefébure, 1929; Tsumura, 1964 (Flume Creek). South-western Atlantic, 49° 48-67’ S, 36° 39-53’ W, 5095 m depth. Deep Sea Drilling Project site 328 (Gombos, 1976). Upper Eocene—lower Oligocene. South-western Atlantic, Falkland Plateau, 51° 00-28’ S, 46° 58-30’ W., 2589 m depth. Deep Sea Drilling Project site 511 (Gombos & Ciesielski, 1983). Although this species was described and figured by Grove & Sturt in 1886, the name Pseudo- rutilaria monile that they used for it was not validly published as they did not give valid publication to Pseudo-rutilaria as a generic name. As explained above (p. 276), they said that Pseudo-rutilaria was a subgenus but nevertheless made the combination with that name and not with Rutilaria Grev., which, as they implied but did not definitely state, was the genus of which Pseudo-rutilaria was a subgenus. It was De Toni & Levi (1887) who first gave valid publication to a name for the species. It is clear from their protologue that it was based entirely on Grove & Sturt’s account and its type material is thus the specimens that those authors had studied. GENERA OF THE BIDDULPHIACEAE 279 Although both Grove’s and Sturt’s collections are in the British Museum (Natural History), none of the specimens in either of them is the original of either of their figures. One cannot say which, if any, were in their hands at the time when their paper was written; some certainly were not. There are a number of other slides mounted by Grove elsewhere in the collections at the Museum, and one of these, BM 71539 from the collection of Thomas Saxton, is a strewn slide from Oamaru labelled in Grove’s hand. Amongst the names of species on the label is ‘Rut. monile’, and it can therefore be regarded as certain that this slide was prepared and labelled before Grove & Sturt published their account of the species and that the specimens on it are part of the original material; on all the other slides labelled by Grove the name is given as Pseudo-rutilaria monile. We therefore designate this slide as lectotype. We have seen under the scanning electron microscope one specimen of this species that we interpret as an end valve. This specimen, which is 120 um long, has no occluded processes nor troughs on its central portion and it lacks marginal spines. This species has a known range from the middle Eocene—Oligocene boundary in deposits from the south-western Atlantic and it also occurs in the late upper Eocene deposit at Oamaru, Otago Province, New Zealand, from which it was originally described. The records of it from the upper Eocene of the Tasman Sea by Hajos (1976) are based on a misidentification; the species she illustrated is Pseudorutilaria clavata R. Ross & P. A. Sims. 2. Pseudorutilaria clavata R. Ross & P. A. Sims, sp. nov. (Pl. 3; Pl. 12 figs 90, 91) Pseudorutilaria monile sensu Hajos in Init. Rep. Deep Sea Drilling Proj. 35: 829, pl. 14 figs 10-12 (1976), non Grove & Sturt. Valva bipolaris, recta, 65-155 wm longa, 22—30 um lata; portio centralis circularis, ab projecturis sulcis transversis separata; projecturae sulcis transitae, utraque margine inter sulcum juxta portionem centralem et proximum gibboso, apices versus primo gradatim angustatatum in apices clavatos rotundatos expansa, margine integro. Duo sulci magis proximales in utraque projectura 10-12 wm distantes, alii 6-8 wm distantes. Altitudo ad vertices elevationum 12-14 um. Spinae ligantes in utraque elevatione 4-5, circum paullo plus quam dimidium proximalem verticis positae, supra valde expansa, 2-0-—3-5 xm altae. Areolae poroides, volis occlusae, extrinsecus margine elevato, diametro c. 0-3 wm, in fronte et parte superiori limbi irregulariter dispersae, 6—8 in 10 wm sed in portione centrali distantiores, in latere proximali elevationum carentes vel rarissimae. Areolae subocellorum 40-50 in 10 wm. Annulus ex c. 6 processibus occlusis constans in portione centrali, in valvis curtissimis carens. Spinae marginales acutae, eis valvae continguae non oppositae, c. 5 in 10 wm. Rimoportula una, in annulo processuum occlusorum et plus minusve in axe transapicali posita, eae valvarum contiguarum idem latus versus extensae. Taenia sola visa (?valvocopula) indivisa, c. 8 wm alta, seriebus verticalibus areolarum parvarum c. 20 in 10 wm, areolis c. 15 in 10 wm. Typus: BM 81136, ex stratis eocaenicis de profundis maris Atlantici australis. Valve bipolar, straight, 65-155 wm long, 22-30 wm broad; central portion circular, separated from the projections by transverse sulci; projections crossed by sulci, gibbous between the central portion and the next transverse sulcus, distal to this tapering and then expanded to clavate rounded apices, their margins entire. Two proximal transverse sulci on each projection 10-12 xm apart, the others 6-8 um apart. Height to the top of the elevations 12—14 wm. Linking spines on each elevation 4—5, extending rather more than half-way round the proximal side of the summit of the elevation, strongly expanded above, 2-0—-3-5 um tall. Areolae poroid, occluded by volae, with a slightly raised external rim, diameter c. 0-3 wm, scattered over the valve surface and the upper part of the mantle, 6-8 in 10 wm, rather sparser on the central portion, absent or very sparse on the proximal side of the elevations. Areolae of the subocellus 40-50 in 10 wm. Ring of about 6 occluded processes on the central portion, absent in the shortest valves. Marginal spines pointed, not opposite those of the sibling valve, c. 5 in 10 wm. Labiate process one, situated in the ring of occluded processes and more or less in the transapical axis, the labiate processes of sibling valves on the same side. The only girdle band seen (? valvocopula) undivided, c. 8 um deep, with vertical rows of areolae, rows c. 20 in 10 wm, areolae c. 15 in 10 wm. 280 R. ROSS AND P. A. SIMS Middle Eocene. South-western Atlantic, Falkland Plateau, 51° 08’ S., 54° 22’ W., 1525 m depth. Vema cruise 17, core 107, (BM SEM CB12.623-—628, CB13.172—173, CB13.313-—326, 5861-5878), 50 cm (BM 81137, SEM CB14.098-101, CB14.121-124, CB14.930-—931, 44263-44268, 44631-44638), 175 cm (BM 81136). Middle—upper Eocene. South-western Atlantic, Falkland Plateau, 53° 01’ S., 52° 52’ W., 2880 m depth. Vema cruise 18, core 104, 330 cm (BM SEM CB13.585-—594, 5891-5894, 5901-5909). Upper Eocene. South-western Pacific, South Tasman Rise, 47° 59-84°S, 147° 45-85’ E., 1600 m depth. Deep Sea Drilling Project site 281 (Hajés, 1976). South-western Pacific, Tasman Sea, 43° 54-60’ S., 154° 16-96’ E., 4750 m depth. Deep Sea Drilling Project site 283 (Hajos, 1976). This species is easily distinguished from Pseudorutilaria monile by the outline of the valve; the margin is gibbous between the proximal sulcus and the next and distal to this it is entire, not crenulate. These characters are well shown on Hajés’s figures (1976: pl. 14 figs 10-12), making it obvious that the specimens from the upper Eocene of the south-western Pacific that she identified as P. monile are actually P. clavata. The girdle band, probably a valvocopula, shown on Fig. 19, is the only one that we have seen on any member of the genus. It would seem that this band is uninterrupted; all the ends of the three pieces of it appear to be breaks and it would seem that no part is missing. We have seen two specimens of Pseudorutilaria clavata that are about 65 wm long. Both of these consist of two sibling valves inseparably joined by interlocking linking spines on their elevations (Fig. 20). There is one labiate process but no occluded processes on the central portion of each valve of these specimens. It would seem that in this species, as in P. incompleta R. Ross & P. A. Sims (see p. 282 below), these structures are not developed on the shortest specimens. There are, however, indications of a ring of tooth-like spines at the centre of the valve similar to those on the end valves of P. incompleta, but the position of the specimens does not allow certain observation of this. Pseudorutilaria clavata has a range from the middle Eocene to the upper Eocene. It has been found in samples from the Falkland Plateau that date from the middle Eocene and from the middle—upper Eocene boundary, and in others from the Tasman Sea area of the south-western Pacific that are upper Eocene in date. 3. Pseudorutilaria nodosa R. Ross & P. A. Sims, sp. nov. (Pl. 4; Pl. 12 fig. 92) Valva bipolaris, recta curvatave, 145-195 wm longa, 32-34 wm lata; portio centralis circularis, ab projecturis costis internis transversis separata; projecturae costis transitae, apices paullo clavatos rotunda- tos versus gradatim angustatae, sed medio gibbosae, marginibus alibi integris. Costae 7-10 wm distantes, eae medio propriores distantiores. Altitudo ad vertices elevationum c. 15 wm. Spinae ligantes 2—4 in utraque elevatione, circum dimidium proximalem verticis positae, c. 4 wm altae. Areolae poroides, extrinsecus margine paullo elevato, diametro c. 0-3 wm, in fronte et parte superiori limbi irregulariter dispersae, c. 6 in 10 wm. Areolae subocellorum 40-45 in 10 wm. Annulus ex 6-7 processibus occlusis constans in portione centrali. Spinae marginales acutae, eis valvae continguae non oppositae, 5-7 wm distantes, sed prope apices crebriores. Rimoportula una, in annulo processum occlusorum et plus minusve in axe transapicali posita, eae valvarum contiguarum idem latus versus extensae. Typus: BM 81138, ex stratis eocaenicis de profundis maris Atlantici australis. Valve bipolar, straight or curved, 145-195 um long, 32-34 wm broad; central portion circular, separated from the projections by transverse internal costae; projections crossed by costae, gradually tapering but gibbous about half-way to the slightly clavate rounded apices, margins entire. Transverse costae 7-10 wm apart, those near the apices the more closely spaced. Height to the top of the elevations c. 15 wm. Linking spines on each elevation 2—4, extending half-way around the summit on its proximal side, c. 4 wm tall. Areolae poroid, with a slightly raised external rim, diameter c. 0-3 ~m, scattered over the valve face and the upper part of the mantle, c. 6in 10 wm. Areolae of the subocellus 40—45 in 10 wm. Ring of 6-7 occluded processes on the central portion. Marginal spines pointed, not opposite those of the sibling valve, 5-7 wm apart but closer near the apices. Labiate process one, situated in the ring of occluded processes on the GENERA OF THE BIDDULPHIACEAE 281 central portion and more or less in the transapical axis, the labiate processes of sibling valves on the same side. Middle Eocene. South-western Atlantic, Falkland Plateau, 51° 08’ S, 54° 22’ W, 1525 m depth. Vema cruise 17, core 107, 50 cm (BM SEM CB15.059-071), 175 cm (BM 81138). Middle—upper Eocene. South-western Atlantic, Falkland Plateau, 53° 01’ S, 52° 52’ W, 2880 m depth. Vema cruise 18, core 104, 330 cm (BM SEM CB11.813-814, 5879-5883). ? ‘Lamont E8’. (Locality and age uncertain, see p. 271) (BM SEM 2715-2728). The swelling about half-way along each projection is the distinguishing feature of this species, and it is to this that the specific epithet applies. Pseudorutilaria clavata is the only other species of the genus in which the margins of the projections are not undulate or crenate. It, too, has a swelling on the projections, but that is situated immediately adjacent to the central portion. In P. clavata, also, the transverse structures are sulci, not the costae that are found in P. nodosa (Figs 24, 26). None of the specimens of this species that we have seen has any trace of a velum in its areolae. The certain records of this species are from samples from the Falkland Plateau that date from the middle Eocene and from the middle—upper Eocene boundary. We have also seen specimens from the sample labelled ‘Lamont E8’, which we believe to come from the same area and to be of about the same age. 4. Pseudorutilaria hannai R. Ross & P. A. Sims, sp. nov. (Pl. 5; Pl. 12 fig. 93) Valva bipolaris, curvata, 150 wm longa, 32 wm lata; portio centralis circularis, ab projecturis costis transversis separata; projecturae costis transitae, ad fines suas proximales latitudine 2/3 ejus portionis centralis, apices angustos rotundatos versus decrescentes, marginibus undulatis, non crenulatis, undis 15-18 um distantibus. Costae c. 8 wm distantes. Altitudo ad vertices elevationum c. 13-5 wm. Spinae ligantes 2—3 in utraque elevatione, circum dimidium proximale verticis positae, supra valde expansae, c. 4 pm altae. Areolae poroides, extrinsecus margine paullo elevato, diametro c. 0-5 wm, in fronte et parte superiori limbi irregulariter dispersae, 6-8 in 10 wm, in portione centrali distantiores. Annulus ex 3-4 processibus occlusis constans in portione centrali. Spinae marginales 2—3 in 10 wm, pro parte maxima clavatae et eis valvae contiguae non oppositae et nonnullae cum eis implexae, aliquae irregulariter formatae positaeque. Rimoportula una, in annulo processuum occlusorum et plus minusve in axe transapicali, positae, in valvis nonnullis carentes. Typus: BM 81300, ex stratis eocaenicis de profundis maris Atlantici australis. Valve bipolar, curved, 150 wm long, 32 um broad; central portion circular, separated from the projections by transverse costae; projections crossed by costae, at their proximal ends about *4 the width of the central portion, tapering to narrow rounded apices, their margins undulate, not crenulate, the crests of the undulations 15—18 wm apart. Costae c. 8 wm apart. Height to the top of the elevations c. 13-5 wm. Linking spines 2—3 on each elevation, extending half-way round the proximal side of the summit, strongly expanded above, c. 4 wm tall. Areolae poroid, with a slightly raised external rim, diameter c. 0-5 ym, scattered over the valve face and the upper part of the mantle, 6-8 in 10 wm, rather sparser on the central portion. Ring of 3—4 occluded processes on the central portion. Marginal spines mostly clavate and alternating and in places interlocking with those of the sibling valve, 2-3 in 10 wm, but some irregular in shape and arrangement. Labiate process one, situated in the ring of occluded processes, and more or less in the transapical axis, absent on some valves. Middle—upper Eocene. South-western Atlantic, Falklands Plateau, 53° 01’ S., 52°52’ W., 2880 m depth. Vema cruise 18, core 104, 330 cm (BM 81300, SEM CB13.595—599, CB13.601-—607, 5910-5912). This species is the rarest of the genus and we have seen only one specimen which we have examined with the scanning electron microscope and then mounted to serve as the type. The undulate outline, which is not related to the position of the transverse costae (Fig. 34), distinguishes it from the other bipolar species. The one specimen that we have seen consists of two joined sibling valves. One of these has one labiate process, the other appears to have no 282 R. ROSS AND P. A. SIMS labiate process. We have not seen a specimen of any other species of the genus without a labiate process. This specimen comes from a middle to late Eocene sample from the Falkland Plateau. 5. Pseudorutilaria incompleta R. Ross & P. A. Sims, sp. nov. (Pl. 6; Pl. 13 figs 94, 95). Eunotogramma weissei var. productum [‘producta’] sensu Fenner in Init. Rep. Deep Sea Drilling Proj. 39: 519, pl. 27 figs 3-5 (1977), non Grove & Sturt. Valva bipolaris, angulo 30°—45° inter projecturas, margine ubique crenulato, 50-155 wm longa, 25-45 wm lata; portio centralis pentagonalis, ab projecturis sulcis transversis sed non indentationibus marginis separata; projecturae sulcis transitae, apices angustos rotundatos versus gradatim angustatae. Sulci 6-7 um distantes. Altitudo ad vertices elevationum 8-5—11 wm. Spinae ligantes in utraque elevatione 1-2, plerumque supra expansae et implexae, c. 2 wm altae, aliquando spina una decrescens c. 3-5 um alta. Areolae poroides, volis occlusae, extrinsecus margine elevato, diametro c. 0-3 yw, in fronte et parte superiori limbi irregulariter dispersae, 6-10 in 10 wm. Areolae subocellorum c. 50 in 10 wm. Annulus ex 8-12 processibus occlusis constans in portione centrali, aliquando annulo spinarum humilium cuneatarum substitutus, in valvis curtissimis carens. Spinae marginales latae, supra expansae, eas valvae contiguae tangens, margine superiori irregulariter indentatae. Rimoportula una, paullo extra annulum processuum occlusorum et latus dorsale valvae versus posita. Typus: BM 81139, ex stratis eocaenicis de profundis maris Atlantici australis. Valves bipolar, with the projections at an angle of 30°—45° to one another, margin crenulate throughout, 50-155 wm long, 25-45 um broad; central portion pentagonal, separated from the projections by transverse sulci but not by indentations of the margin; projections crossed by sulci, gradually tapering to the narrow rounded apices. Transverse sulci 6—7 xm apart. Height to the summit of the elevations 8-5—11 wm. Linking spines 1-2 on each elevation, normally expanded above and interlocking, c. 2 wm tall, sometimes a tapering spine c. 3-5 um tall. Areolae poroid, occluded by volae, with a raised external rim, diameter c. 0-3 wm, scattered over the valve surface and the upper part of the mantle, 6-10 in 10 um. Areolae of the subocellus c. 50 in 10 wm. Ring of 8-12 occluded processes on the central portion, sometimes replaced by a ring of low cuneate spines, absent in the shortest valves. Marginal spines broad, expanded above, abutting against those of the sibling valve and with irregularly indented upper margins. Labiate process one, slightly outside the ring of occluded processes on its dorsal side. Middle Eocene. South-western Atlantic, Falkland Plateau, 51° 08’ S, 54° 22’ W, 1525 m depth. Vema cruise 17, core 107 (BM 81139, SEM B26.67—77, B26.645-651), 50 cm (BM 81140, CB14.095-097, CB14.102—110, CB14.125—128, CB14.909-929, CB15.041—057), 170 cm (BM 81141). South-western Atlantic, south-west edge of the Sao Paulo Plateau, 28° 17-22’ S, 41° 05-28’ W, 3203 m depth. Deep Sea Drilling Project, site 356 (Fenner, 1977). Middle—upper Eocene. South-western Atlantic, Falkland Plateau, 47° 45-7’ S, 57° 38-5’ W, 3650 m depth. Conrad cruise 12, core 237 (BM SEM 5444-5450, 23545-23554, 33543-33549). ? ‘Lamont E8’. (Locality and age uncertain, see p. 271) (BM SEM 2706-2709, 2711-2714, 2729- 2730). Amongst the specimens of this species that we have examined with the scanning electron microscope there are two which differ from the others in a number of ways: the ring of occluded processes and flanged troughs on the central portion is replaced by a rather smaller hyaline area surrounded either by eight or nine low cuneate spines or by a low flange-like costa (Figs 39, 45); the external tube of the labiate process is straight, not curved; the marginal ridge is undulate but bears no spines; there is one spine on each elevation and this is not expanded above but tapers to a point. We interpret these as separation valves. A third specimen consisting of a single valve mounted with its inner side uppermost seems also to be a separation valve. It has a raised (seen as depressed from the inside) area in the centre similar in size to the raised hyaline area of the other two and it has no marginal spines. On shorter valves neither occluded processes and their associated troughs nor cuneate spines are present on the central portion of the valve. Of specimens we have examined with the light microscope, there is one that is 100 wm in length that has neither, and another of the same GENERA OF THE BIDDULPHIACEAE 283 length, the holotype, which has a ring of occluded processes and troughs. A further specimen 80 pm long has neither occluded processes nor cuneate spines and there is no sign of any of these on the three specimens figured by Fenner (1977) under the name Eunotogramma weissei var. productum Grove & Sturt; these are 50 wm, 60 wm, and 80 um long. E. weissei var. productum differs from Pseudorutilaria imcompleta in a number of ways: it has entire margins; its projections do not taper towards the apices and are much narrower proximally relative to the breadth of the central portion; they are crossed by costae expanded laterally at their free edges; there is a ring of c. eight labiate processes on the central portion; the elevations bear pseudocelli but no linking spines. Although neither linking spines nor subocelli can be seen on Fenner’s figures, they do not show that the specimens did not possess them, and in all other respects the specimens that she illustrates have the characters that distinguish P. incompleta from E. weissei var. productum; they clearly are the former. Pseudorutilaria incompleta occurs in samples from the Falkland Plateau and from the south-west edge of the Sao Paulo Plateau that date from the middle Eocene. It also occurs in a sample from the edge of Falkland Plateau (Conrad cruise 12, core 237) the age of which is less well established but is almost certainly middle or upper Eocene. We have also seen specimens from ‘Lamont E8’, a sample for which there are no firm data but which we believe to come from the Falkland Plateau area and to be of about the same age as the other samples in which the species occurs. 6. Pseudorutilaria hendeyi R. Ross & P. A. Sims, sp. nov. (Pio 73-PE AS he. 96) Valva tripolaris, marginibus concavis, crenulatis, 95-185 wm ab apice ad apicem; portio centralis hexagonalis, ab projecturis sulcis transversis sed non indentationibus marginis separata; projecturae apices rotundatos versus angustatae, prope portionem centralem sulcis, apices versus costis, transitae. Sulci costaeque c. 8-5 wm distantes. Crenae marginis c. 11 zm distantes, eae in lateribus oppositis projecturae non semper oppositae. Altitudo ad vertices elevationum c. 10 wm. Spinae ligantes 3—4 in utraque elevatione, supra expansae et implexae. Areolae poroides, diametro c. 0-3 wm, in fronte irregulariter dispersae vel in seriebus obliquis indistinctis positae, 6-8 in 10 wm, prope centrum valvae et in parte superiori limbi distantiores vel carentes. Area irregularis diametro c. 10 ~m ad centrum valvae porellis c. 30 in 10 wm instructa. Annulus ex c. 12 processibus occlusis constans ad centrum valvae, aream porellorum cingens. Spinae marginales c. 4 wm distantes, irregulariter positae, eae valvae contiguae non oppositae, supra dendriticales et hic illic implexae. Rimoportula una vel duae approximatae in annulo picecssuua occlusorum posita. Typus: BM 81142, ex stratis eocaenicis de profundis maris Atlantici australis. Valves tripolar, with concave crenulate margins, 95-185 um from apex to apex; central portion hexagonal, separated from the projections by sulci but not by indentations of the margin; projections tapering to the rounded apices, crossed by sulci grading distally into costae. Sulci and costae c. 8-5 wm apart. Crenulations of the margin c. 11 wm apart, not always opposite on the two sides of a projection. Height to the top of the elevations c. 10 wm. Linking spines 3—4 on each elevation, expanded above and interlocking. Areolae poroid, diameter c. 0-3 um, scattered or in indistinct oblique rows on the valve surface, 6—8 in 10 wm, sparser to absent near the centre of the valve and on the upper part of the mantle. An irregular area, diameter c. 10 um, of porellic. 30 in 10 um in the centre of the valve. Ring of c. 12 occluded processes on the central portion surrounding the area of porelli. Marginal spines c. 4 wm apart, irregularly spaced, not opposite those of the sibling valve, dendritically branched above and in places interlocking. One, or two adjacent, labiate processes in the ring of occluded processes. Middle Eocene. South-western Atlantic, Falkland Plateau, 51° 08’ S., 54° 22’ W., 1525 m depth. Vema cruise 17, core 107, 50 cm (BM 81143, SEM CB14.517-—520), 170 cm (BM 81142). Middle—upper Eocene. South-western Atlantic, Falkland Plateau, 53° 01’ S., 51° 52’ W., 2880 m depth. Vema cruise 18, core 104, 330 cm (BM SEM CB.13.576-584, CB13.617—618, 5884-5890, 5895-5900). This is the only species of Pseudorutilaria with an area of porelli in the centre of the valve (Fig. 51). It would seem that the valve is thinned in this area as it is broken away in most of the 284 R. ROSS AND P. A. SIMS specimens we have seen. In none of these is there any trace of vela in the areolae and in all of them the subocelli have been broken away. Pseudorutilaria hendeyi has only been found in samples from the Falkland Plateau that date from the middle Eocene to the middle—upper Eocene boundary. Ill. MONILE R. Ross & P. A. Sims, gen. nov. Valvae bipolares, sulcis transitae, ad apices elevationibus munitae. Elevatio utraque annulo ex dentibus radialibus humilibus constanti circum verticem suum instructa et subocellum ferens. Limbus ad apices, centro valvae oppositus et in locis aliquis ubi ambitus valvae convexus est concavus, alibi verticalis. Areolae poroides, cribris occlusae. Rimoportulae paucae, ad centrum valvae irregulariter aggregatae. Taenia sola visa areolata, antiligulam ferens. SPECIES TYPICA: Monile laurentii R. Ross & P. A. Sims, infra. Valves bipolar, crossed by sulci, with elevations at the apices. Elevations with a ring of radial low teeth around their summits and each bearing a subocellus. Mantle concave at the apices, opposite the centre of the valve and in other places where the outline of the valve is convex, elsewhere vertical. Areolae poroid, occluded by cribra. Labiate processes few, in an irregular group at the centre of the valve. The only girdle band seen areolate, antiligulate. This genus, like Pseudorutilaria monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, has something of the appearance of a row of beads. We have therefore chosen as its generic name the noun used in apposition as a substantival epithet for that species, a word meaning necklace. Monile resembles Pseudorutilaria in many ways but differs in important respects: its cells are not united in inseparable filamentous colonies, although the flattened summits of the elevations suggest that they did form filaments (Figs 63, 64); there are no occluded processes with their associated troughs at the centre of the valve (Fig. 61); there are no interlocking linking spines on the elevations (Figs 63, 64); also the subocelli extend on to the summits of the elevations. Pseudorutilaria is found earlier in the fossil record than Monile, and the nature and degree of the relationship between the two is far from clear. It is discussed below (p. 291). 1. Monile laurentii R. Ross & P. A. Sims, sp. nov. (Pl. 8; Pl. 13 figs 97, 98) Valva bipolaris, longa angustaque, marginibus sinuosis, apicibus subclavatis, ex portione centrali circulari tholiformi et projecturis ab portione centrali sulcis separatis, sulcis transitis et inter sulcos tholiformibus, constans, elevatione ad apicem utremque munita, 63-220 wm longa, 20-25 um lata. Elevationes verticales, cylindricae, supra expansae, verticibus applanatis, utraque annulo ex dentibus humilibus radialibus linearibus constanti circum verticem instructa, et subocellum in vertice et in parte superiori expansa ferens. In utraque elevatione sub subocello costae humiles irregulariter anastomosantes, reticu- lum debiliter effectum facientes. Altitudo ad vertices elevationum 18—40 wm. Limbus ad apices, centro valvae oppositus et proximaliter sed non distaliter in projecturis ubi ambitus valvae convexus est, proxime super marginem concavus, alibi verticalis. Crista marginalis humilis, hyalina ab elevationibus ad por- tionem centralem extensa. Areolae poroides, cribris occlusae, extrinsecus margine paullo elevato, diam. c. 0-7 wm, in fronte, in elevationibus et in parte superiori limbi ubi ambitus valvae convexus est irregulariter dispersae, 4—6 in 10 wm; areolae subocellorum 15-20 in 10 wm, diametro c. 0-1 wm. Pori interstriales in fronte irregulariter dispersi. Spinae superficiales ad 4 wm longae in fronte praeter partes distales projecturarum, ex marginibus elevatis areolarum orientes. Rimoportulae 4, ad centrum valvae irregulari- ter aggregatae, unaquaeque apertura interna recta paullo elevata et, extrinsecus, ex tubo crasso 4 wm minimum alto et, pro parte maxima, basi fultura munito constans. Taenia sola visa c. 8 wm alta, ad unum apicem antiligula valida instructa, seriebus verticalibus areolarum ornata, seriebus areolisque 10-12 in 10 pm. Typus: BM coll. Adams H941, specim. I, 12, ex depositione miocenica ad ‘St. Laurent-la-Vernéde’, Gard, Gallia. Valve bipolar, long and narrow, with sinuous margins and sub-clavate apices, consisting of a circular domed central portion separated by sulci from the projections that are themselves GENERA OF THE BIDDULPHIACEAE 285 crossed by sulci and domed between them, and with an elevation at each apex, 63-220 wm long, 20-25 wm broad. Elevations vertical, cylindrical, expanded above, with flattened summits bearing a ring of low radial linear teeth around their margins, and with a subocellus on their summits and expanded upper part. A poorly developed network of low, irregularly anastomos- ing costae on the elevations. Height to the top of the elevations 18-40 um. Mantle strongly concave immediately above the margin at the apices, opposite the centre of the valve and proximally but not distally on the projections where the outline of the valve is convex, elsewhere vertical. A low hyaline marginal ridge extending from the elevations to the central portion. Areolae poroid, occluded by cribra, with a slightly raised external rim, diameter c. 0-7 wm, scattered over the valve surface, the elevations, and the upper part of the mantle where the valve outline is convex, 4—6 in 10 wm; areolae of the subocelli 15-20 in 10 wm, diameter c. 0-1 wm. Scattered interstrial pores on the valve surface. Superficial spines up to 4 wm long arising from the raised rim of the areolae on the valve surface, except on the distal parts of the projections. Labiate processes 4, in an irregular group in the centre of the valve, each with a straight, slightly raised internal opening and, externally, a stout tube at least 4 wm tall, mostly with a single buttress at the base. One girdle band seen, about 8 xm deep, with a well developed antiligula at one apex and with vertical rows of areolae, rows and areolae 10-12 in 10 um. Miocene. St Laurent-la-Vernéde, Gard. France (BM 81303, coll. Adams H941, TS961, SEM CB13.634—- 650, 36674-36682). IV. MALUINA R. Ross & P. A. Sims, gen. nov. Valva bipolaris, ad apices elevationibus aliquanto crassis, non supra expansis, ad summum utraeque elevationis in latere suo distali subocello minimo. Limbus verticalis. Areolae poroideés, parvae, in partibus valvae sparsae carentesve. Spinae ligantes 2—4 in latere proximali verticis utraeque elevationis, supra expansae et implexae. Rimoportula una, prope centrum valvae posita, infra sessilis longa angusta, extra tubo curto recto. SPECIES TYPICA: Maluina centralitenuis (R. Ross & P. A. Sims) R. Ross & P. A. Sims, comb. nov. infra. Valve bipolar, with moderately stout elevations at the apices, not expanded above. On the distal side of the top of each elevation a very small subocellus. Mantle vertical. Areolae poroid, small, sparse or absent on parts of the valve. Linking spines 2—4 on the proximal sides of the summits of each elevation, expanded above and interlocking. Labiate process single, situated near the centre of the valve, internally sessile, long and narrow, externally with a short, straight tube. We have chosen a name for this genus from the only locality where it has been found, the Falkland (Malvinas) Plateau. It is clear that the species for which this genus is erected here cannot be retained in Hemiaulus Heiberg. The interlocking linking spines and, even more obviously, the subocellus with its very close areolae or porelli and its definite margin are inconsistent with that position for it. Also, it cannot be transferred to any of the other genera with a small number of interlocking linking spines on each elevation. Briggera R. Ross & P. A. Sims and Dicladiopsis De Toni have pseudocelli with indefinite margins and expanded tips to their elevations; Solium Heiberg and Pseudorutilaria (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni have projections cut off from a central portion by sulci or costae, an indented mantle, and anastomosing costae on the elevations and mantle. These differences, in our view, preclude placing Hemiaulus centra- litenuis R. Ross & P. A. Sims in any of these genera. 1. Maluina centralitenuis (R. Ross & P. A. Sims) R. Ross & P. A. Sims, comb. nov. (Pl. 9) Hemiaulus centralitenuis R. Ross & P. A. Sims in Beih. Nova Hedwigia 54: 182, pl. 2 figs 7-9 (1977). Valves elliptical, 20-55 um long, 15-25 um broad, 10-12 wm tall at the domed centre, with more or less strongly developed acute pseudo-antiligulae at the apices and a domed centre about % the width of the valve in diameter. Elevations arising at the apices, cylindrical and tapering slightly upwards, c. 5 wm in diameter; height to the top of the elevations 15—25 wm. Valve 286 R. ROSS AND P. A. SIMS surrounded by a flange-like marginal ridge. Areolae poroid, with a weak external rim, c. 0-15 jzm in diameter, in radial rows on the central dome and sometimes also around the outer part of the valve face, rows 25—30 in 10 wm, areolae 30-35 in 10 wm; areolae also dense, 25—30 in 10 wm, on the distal side and the lower part of the proximal side of the elevations, elsewhere on the elevations sparse; areolae absent in a broad area around the central dome. A very small, transversely widened subocellus on the distal side of each elevation at its tip, the areolae or porellic. 60in 10 wm. Linking spines 2—4, interlocking, expanded immediately above their base, narrowed above this and then expanded into a clavate tip rounded at the apex, 6-7 um long. Labiate process about half-way between the centre of the valve and the edge of the central dome, slightly nearer one apex than the other. Middle Eocene. South-western Atlantic, Falkland Plateau, 51° 08’ S., 54° 22’ W., 1525 m depth. Vema cruise 17, core 107, 50 cm (BM SEM CB14.074-077, CB14.080-—083, CB14.502-—505, 5818-5826; Ross, Sims & Hasle, 1977). Middle—upper Eocene. South-western Atlantic, Falkland Plateau, 53° 01’ S., 52° 52’ W., 2880 m depth. Vema cruise 18, core 104, 330 cm (BM 78000, holotype; Ross, Sims & Hasle, 1977). Since we first described this diatom as a species of Hemiaulus (in Ross, Sims & Hasle, 1977), we have been able to examine additional, and less eroded, specimens with the scanning electron microscope. These have provided us with more information about the structure of the valves and hence we have provided a new description here. In 1977 we drew attention to the small area at the tips of the elevations that was broken away in the specimens that we had then seen and suggested that a pseudocellus might have been present there. We now know that this area is occupied by the structure of the sort which we have here called a ‘subocellus’ (Fig. 68). This species, and genus, is known only from cores taken from the Falkland Plateau. The samples in which it has been found come from the middle Eocene and the middle—upper Eocene boundary. V. BONEA R. Ross & P. A. Sims, gen. nov. Frustula rectangularis, in catenis inseparabilibus conjuncta. Valvae tripolares, ad apices elevationibus humilibus. Elevatio unaquaeque vertice annulo perfecto spinarum ligantium implexarum circumcincto, et pseudocellum ferens. Limbus verticalis. Areolae poroides, rotis occlusae. Rimoportulae paucae, prope centrum valvae positae. Cingulum ex taeniis duis areolatis, probabiliter indivisis constans. SPECIES TYPICA: Bonea simulans R. Ross & P. A. Sims, infra. Frustules rectangular, united in inseparable chains. Valves tripolar, with low elevations at the corners. Elevations with their summits surrounded by a complete ring of interlocking linking spines, and bearing a pseudocellus. Mantle vertical. Areolae poroid, occluded by rotae. Labiate processes few, situated near the centre of the valve. Girdle bands two, areolate, probably undivided. We have named this genus after Mr E. C. P. Bone of Portslade, who has assisted us materially by providing us with many selected specimens of species that we have studied. Bonea has many points of resemblance to Strelnikovia R. Ross & P. A. Sims (1985) but differs in having a complete ring of interlocking linking spines on the summits of the elevations (Figs 73, 75). These linking spines, moreover, have bases that extend circumferentially, not radially as in Strelnikovia; in this, Bonea resembles more closely Briggera R. Ross & P. A. Sims (1985) and Dicladiopsis De Toni (1894). Keratophora Pantocsek (1889) and Thaumatonema Greville (1863), which, like Bonea, have a complete ring of interlocking linking spines around the summits of their elevations, also have the bases of these extending radially, as in Strelnikovia. The tall, tubular, hyaline elevations of these two genera distinguish them even more sharply from Bonea, which, nevertheless, is related to the group of genera that consists of them along with Strelnikovia, Briggera, and Dicladiopsis. GENERA OF THE BIDDULPHIACEAE 287 1. Bonea simulans R. Ross & P. A. Sims, sp. nov. (Pl. 10; Pl. 13 figs 99, 100) Valva tripolaris, lateribus rectis vel paullo convexis, fronte ad centrum paullo tholiformi, inter centrum et medium uniuscuiusque lateris convexa, elevationes versus concava et ad vertices earum ascendenti; longitudo lateris 115-125 wm. Elevationes sub-triangulares, interdum a limbo depressione angusta horizontali separatae. Altitudo ad vertices elevationum c. 30 wm. Spinae ligantes in unaquaeque elevatione 15—20, supra expansae et ambitu irregulari. Crista marginalis humilis, hyalina ab elevatione ad elevationem extensa, spinas marginales acutas, 10-12 wm altas, c. 5 wm distantes ferens; hae spinae oppositae et superimpositae sed non implexae. Areolae poroides, rotis radiis 7-10 instructis occlusae, radiis ramulo uno vel ramulis duobus conjunctis, diametro 0-8—1-5 wm, in fronte in seriebus radialibus dispositae, seriebus 2-5—4 in 10 wm, areolis 2—4 in 10 wm, sed prope centrum valvae distantiores; series areolarum media laterum versus distantiores quam eae apices versus directae; medio in valva area hyalina irregularis, diametro c. 10 ~m; in limbo prope marginem series una horizontalis areolarum 2-5-3 in 10 wm et, prope cristam marginalem, alter series similaris. Pori interstriales in fronte irregulariter dispersi. Pseudocellus in vertice et latere distali uniuscuiusque elevationis et in limbo sub elevatione, interdum area hyalina verticali in limbo et, quum depressio horizontalis inter limbum et elevationem adest, area hyalina super hanc interruptus; areolae pseudocellorum in seriebus dispositae, seriebus in vertices elevationum plus regularibus quam alibi; series areolaeque 8-10 in 10 um; areolae in vertices elevationum diametro 0-2—0-4 wm, minor quam eae alibi diametro 0-5—1-0 wm. Rimoportulae 2-4, prope aream centralem hyalinam irregulariter dispositae, unaquaeque apertura interna recta, curta, paullo elevata, extrinsecus ex tubo verticali c. 6 4m alto appendice acuta lanceolata ad apicem instructo constans. Valvocopula parte exteriori c. 20 wm alta, ad omnes apices antiligula rotundata humili instructa, et areolis irregulariter dispersis 2-5—4 in 10 wm ornata; pleura parte exteriori c. 10 wm alta et parte interiori hyalina c. 4 wm alta, pars exterior serie areolarum c. 5 in 10 wm secus marginem proximalem et, huic distalibus, seriebus duabus paullo irregularibus areolarum c. 4in 10 wm. Typus: BM 81301, ex stratis eocaenicis ad ‘Carlovo, U.S.S.R.’ Valve tripolar, with straight or slightly convex sides, the face slightly domed in the centre, convex between the centre and the middle of each side but concave towards the elevations and rising to their summits, length of side 115-125 wm. Elevations sub-triangular, sometimes separated from the mantle by a narrow horizontal depression. Height to the summit of the elevations c. 30 wm. Linking spines 15—20 on each elevation, expanded and rather irregularly shaped above, 2—4 um tall. A low hyaline marginal ridge running from elevation to elevation and bearing acute marginal spines 10-12 wm tall and c. 5 wm apart, the spines opposite and overlapping but not interlocking. Areolae poroid, occluded by rotae with 7-10 spokes, the spokes connected by one or two side branches, diameter 0-8—1-5 wm; radial rows of areolae on the valve face, rows 2:5—4 in 10 um, areolae 2—4 in 10 wm, except near the centre of the valve, where they are sparser; the rows directed towards the middles of the sides more widely spaced than those directed towards the apices; an irregular central hyaline area c. 10 «wm in diameter. Mantle with a horizontal row of areolae 2:5—3 in 10 wm close to the margin and another close to the marginal ridge. Scattered interstrial pores on the valve surface. Pseudocellus on the summit and distal face of the elevations and on the mantle below the elevations, where it is sometimes interrupted by a vertical hyaline area, a hyaline area also above the transverse furrow separating mantle and elevation when this is present; areolae of the pseudocelli arranged in rows, more regular on the summits of the elevations than elsewhere, rows and areolae 8-10 in 10 um; areolae on the summits of the elevations 0-2—0-4 wm in diameter, smaller than those on the distal face and the mantle, which are 0-5—1-0 wm in diameter. Labiate processes 2—4, irregularly distributed a little outside the central hyaline area, with a slightly raised, short, straight internal opening and, externally, a vertical tube c. 6 wm tall with a lanceolate acute appendage on its tip c. 2 um long. Valvocopula with a pars exterior c. 20 wm deep, with a slightly rounded antiligula at each apex and irregularly scattered areolae 2-5—4 in 10 xm; pleura with a pars exterior c. 10 wm deep and a hyaline pars interior c. 4 um deep, the pars exterior with a row of areolae c. 5in 10 um along its proximal edge and two somewhat irregular rows c. 4 in 10 wm distal to this. ? Lower Eocene. ‘Carlovo’. U.S.S.R. (BM 81301, SEM 25006-25016, 36571-36580). (For this locality see Ross & Sims 1985: 281) ‘288 R. ROSS AND P. A. SIMS Bonea simulans resembles Triceratium exornatum Greville (1865), which was described from the Cambridge Estate, Barbados; it also occurs in samples from Springfield and Joe’s River on that island. The material from the Cambridge Estate has been dated as middle Eocene and that from Joe’s River as coming from the upper Eocene—Oligocene boundary (Holmes & Brigger, 1979). There is no firm date available for the material from Springfield. We have examined the holotype of 7. exornatum (BM 3079) and a number of other specimens of the species with the light microscope but have found no trace of linking spines or marginal spines on any of these. There are a few other minor differences: the areolae of T. exornatum are 4-5—5 in 10 wm on the parts of the valve surface between the centre and the elevations, whereas those of B. simulans are 3—4 in 10 wm on that part of the valve; there is only one row of areolae on the mantle in T. exornatum, but two in B. simulans (Fig. 74); in T. exornatum a transverse furrow separating elevation from mantle is never present; we could not detect any labiate process on T. exornatum. These are the only differences between the species; in outline and contour, in shape of the elevations, in arrangement of the areolae, and in the position and size of the pseudocelli they are alike. We have chosen the epithet simulans for the species described here because of this resemblance. It should, perhaps, be pointed out here that T. exornatum is not correctly placed in the genus Triceratium Ehrenberg (1839). It belongs either in Sheshukovia Glezer (1975) or in Biddulphia Gray (1821); we have reservations about whether these two genera are distinct. As we said earlier (Ross & Sims, 1985: 281), there is uncertainty about the locality from which the material labelled ‘Carlovo’ came and about its geological age. This is the only deposit in which we have found Bonea simulans. Glezer (in Glezer et al., 1974) records Triceratium exornatum from the early to late Eocene of the west Siberian plain (pp. 112, 118, 119, 122, tabl. XXI fig. 1) and from the late Eocene of Kazakhstan (pp. 127, 129, tabl. XXXII fig. 8). The two figures are of the same specimen, that on tabl. XXXII rotated 120° anti-clockwise as compared with that on tabl. XXI. It is impossible to tell from these illustrations whether the specimen shown is T. exornatum or Bonea simulans; it might be either. VI. DEXTRADONATOR R. Ross & P. A. Sims in Bacillaria 3: 115 (1980). Additional information about Dextradonator eximius (Grunow) R. Ross & P. A. Sims, the type species of this genus, is given in the description below. 1. Dextradonator eximius (Grunow) R. Ross & P. A. Sims in Bacillaria 3: 118, figs 1-4 (1980). (PI. 11) Syringidium eximium Grunow in Van Heurck, Syn. Diat. Belg.: pl. 106 figs 1, 3 (1883). Frustules cylindrical, united in inseparable chains, the apical axis 3-10 times as long as the diameter. Valves circular or almost so, the surface flat to concave, diameter 30-75 xm, height to the centre of the valve 20-30 um. Elevations two, marginal, cylindrical, 5-12 wm in diameter, height to the top of the elevations 55-75 wm. A hyaline marginal ridge c. 10 wm tall and turned inwards along its upper edge extending from elevation to elevation. Mantle vertical, hyaline. Areolae poroid, c. 0-8 ~m in diameter, vela eroded on all specimens seen, in radial rows on the valve surface, rows and areolae 9-10 in 10 wm, extending up the proximal side of the elevations; a few small areolae or pores, 0-2—0-4 um in diameter on the distal side of the elevations near their summits. An irregular hyaline area containing 1-3 areolae in the centre of the valve, diameter 4—6 zm. Two flattened, acute linking spines on each elevation, partially encircling the elevation of the sibling valve, those on one ofa pair of sibling valves arising from the distal side of the tops of the elevations, those on the other from the proximal side, linking spines 15—30 wm tall. No labiate process seen. Cingulum consisting of two entire bands; the valvocopula 60-75 um deep, its pars interior hyaline and overlapping a rounded internal ridge at the margin of the valve, its pars exterior with a hyaline distal margin 2-5 xm deep, elsewhere with scattered large areolae c. 2 ~m in diameter and 3-5 wm apart; pleura hyaline, 15—30 wm deep. GENERA OF THE BIDDULPHIACEAE 289 Middle Eocene. Cambridge Estate, Barbados (BM 10453, 13789; Grunow in Van Heurck, 1883). Middle—Upper Eocene boundary. Indian Ocean, 10° 25’ S, 63° 15’ E, 3115 m depth, dredge sample. Dodo-123-D1 (BM SEM B26.387-—392, CB11.301—302, CB19.911—912, 46522-46532). Middle Eocene—lower Miocene. Barbados (BM coll. Adams TS 266; PH coll. Shulze arr. 196, 1352, 2404, 2406, 2791). Bissex Hill, Barbados (BM coll. Adams TS 891). Malvern Hill, Barbados (PH coll. Febiger 83, 131). Mount Hillaby, Barbados (BM coll. Adams GC 3088). Newcastle, Barbados (BM coll. Adams F 1236). Since we published our previous account of this species (Ross & Sims, 1980), we have been able to examine more specimens and to add appreciably to our knowledge of it. The most remarkable feature is the inflexed marginal ridge (Fig. 84). This is also present in Dextradonator jeremianus R. Ross & P. A. Sims, the other species of the genus, but we know of nothing similar in any other genus. The stratigraphic range of this species is uncertain. All the specimens that can be firmly dated come from the middle Eocene or the middle—upper Eocene boundary, but there can be no certainty about the level in the Oceanic Beds from which come the specimens from Bissex Hill, Malvern Hill, Mount Hillaby, and Newcastle, and those with no locality detail other than ‘Barbados’. In addition to the specimens that we have definitely identified as Dextradonator eximius, there is one about which we are uncertain (Figs 82, 83). This comes from the Indian Ocean dredge sample that dates from the middle—upper Eocene boundary. It has a diameter of 94 um, appreciably larger than those about whose identity we are sure, and it differs from them in having a marginal ridge that is not turned inwards at its upper edge and that bears two flattened spines about 30 um tall opposite each other and half-way between the elevations. In all other respects it is identical with the specimens that we include in D. eximius with certainty. To decide whether it represents a separate species or whether the characteristics peculiar to it are those of specimens at the upper end of the size range of D. eximius needs a greater range of specimens than those currently available to us. Discussion As we pointed out in the introduction to this paper, we deal here with anumber of genera that do not form a closely related group, and hence their affinities and position in the classification system of diatoms can only be discussed piecemeal. There are, however, a few general points to be made before doing this. As all the genera dealt with in this paper belong to the order Biddulphiales, we need to say something about the position and circumscription of this order. Very different views on this have been put forward during the last decade. This is not the place to consider whether a taxon containing all the diatoms should have the level of a division or of a subdivision within the Chrysophyta, but the latter is, in our view, the lowest rank to be assigned to it. Some authors, the latest of them Round (1981), have considered that, at the next lower level, the diatoms should be divided into three groups: the centric ones, the pennate ones without a raphe, and those with a raphe. Others, most recently Simonsen (1979), consider that the pennate diatoms with and without a raphe should not be separated at this level. This difference of opinion does not, however, affect the circumscription of a taxon containing the centric diatoms, which include the Biddulphiales. Hendey (1964), on the other hand, rejects the separation of the centric and pennate diatoms, and divides the diatoms into a number of groups with the rank of suborder. We accept that the centric and pennate diatoms should be separated as distinct taxa at the level of class. Whether the pennates should constitute a single class or two, one without and one with a raphe, is not relevant here. Within the class Centrophycidae we include an order Biddulphiales. We thus agree with Glezer (1979) and Round (1981) about the rank of a taxon that includes Biddulphia but excludes Thalassiosira, Coscinodiscus, Aulacodiscus, and Rhizo- solenia, rather than with Simonsen (1979), for whom it is a suborder within the order Centrales. 290 R. ROSS AND P. A. SIMS For both Glezer and Round, the Biddulphiales have a much narrower circumscription than that attributed to his suborder by Simonsen; they both include in the order only part of the Biddulphiaceae as delimited by Simonsen, to which Glezer adds a few genera included by Simonsen in the Eupodiscaceae and the Diatomaceae. We would argue for a much wider circumscription of the order than either Glezer’s or Round’s, but asomewhat narrower one than Simonsen’s in that we consider that the Lithodesmiaceae should not be included. Also, we are not completely certain that either the Chaetoceraceae or the Stictodiscoideae belong in the order. There has, in addition, been one important change in the classification of the genera that we would definitely include within the Biddulphiales: the establishment of the family Cymatosir- aceae Hasle, von Stosch & Syvertsen (1983). We point out below the difficulties in classifying into families the remaining genera that we include in the order. We regard all the diatoms that we definitely include in the Biddulphiales as having polar valves, from unipolar, as in /sthmia Agardh, to multipolar, as in those species of Triceratium Ehrenberg for which De Toni (1894) erected the genus Nothoceratium. We consider that the ocelli of such genera as Euopodiscus J. W. Bailey and Rattrayella De Toni indicate that the basic morphology of these genera is polar in spite of their circular outline. Also, almost invariably throughout the order as we delimit it the internal opening of the labiate process is straight and projects as no more than a papilla into the interior of the frustule; it is not borne on a stalk as in the Thalassiosirales and Coscinodiscales. The only exceptions of which we are aware are Isthmia Agardh (see Navarro, 1981) and Biddulphiopsis von Stosch & Simonsen (1984), in both of which the labiate process is shortly stalked internally. Labiate processes of a similar form to that found in the Biddulphiales occur in some Melosiraceae, and all those found in pennate diatoms are of this type. Another character that is widespread among the Biddulphiales as we delimit them is the presence of interstrial pores. The genera in which these occur include not only Biddulphia (see Ross & Sims, 1971: pl. 1 fig. 2) and Biddulphiopsis (see von Stosch & Simonsen, 1984), but also Hemiaulus (see Ross, Sims & Hasle, 1977: pl. 5 fig. 32, pl. 8 fig. 52), Trigonium (see Ross & Sims, 1971: pl. 2 fig. 6), Amphitetras, and Pseudauliscus (unpublished observations). They do not seem to occur, however, in the Cymatosiraceae or the Chaetoceraceae, nor outside the Biddulphiales. ; Glezer (1979) restricts the order Biddulphiales to genera with poroid areolae, maintaining that the type of areola is of primary importance in the classification of the diatoms and that such structures as the ocellus and the pseudocellus are polyphyletic (see also Glezer, 1983). She accordingly maintains Zygoceros Ehrenberg, a genus with ocelli and loculate areolae (see Ross & Sims, 1971), as separate from Odontella Agardh, with ocelli but poroid areolae, whereas other authors, e.g. Simonsen (1974), have included the species with loculate areolae in Odontella. She places Zygoceros, along with Trigonium Cleve, which has loculate areolae but pseudocelli at the apices, in the Coscinodiscaceae, whilst assigning Odontella and Amphitetras Ehrenberg, another genus with ocelli and poroid areolae, to the Biddulphiaceae, along with Biddulphia Gray which also has poroid areolae but has pseudocelli at the apices. The number of characters shared by the genera we include in the Biddulphiales and not present in Coscinodis- cus Ehrenberg and its obvious allies indicates that this separation on a priori grounds is unsound. All the evidence seems to point to loculate areolae being polyphyletic, and the fact that, when areolae become close-packed, they must be of this type for mechanical reasons, reinforces this view. More recently, Glezer (1986) has proposed a new monotypic family for Triceratium Ehrenberg sensu stricto, and assigned it to the Pyxidiculales. She bases her arguments for this on the view that the pseudoloculi of Triceratium are areolae. However, as we showed earlier (Ross & Sims, 1971), and Miller & Collier (1978) confirmed, the pseudoloculi of Triceratium are formed by a network of external costae laterally expanded at their free edges and the areolae are small and poroid. The genus is closely related to Amphitetras, which Glezer (1979) includes in the Biddulphiales. As well as differing from Glezer on the circumscription of the Biddulphiales, we also do not agree with her classification of the genera that she includes within the family Biddulphiaceae. She separates the Biddulphiaceae into two subfamilies, one containing those genera with bipolar valves, the other those genera with tripolar and multipolar valves, again on a priori grounds. GENERA OF THE BIDDULPHIACEAE 291 There is, however, overwhelming evidence that the change from bipolar valves to tripolar ones has happened many times. The range of morphological variation in diatoms of the two shapes is very similar and there are many cases where the only differences between species other than the number of poles are in the detail of the arrangement and spacing of the areolae, characters that in any other context would be regarded as indicating separation at no more than the specific level. There are even a number of cases where bipolar and tripolar forms have been treated as varieties of the same species, Odontella retiformis (A. Mann) von Stosch (1985) being the most recent. Pseudorutilaria adds another example where bipolar and tripolar diatoms must be considered as belonging to the same taxon at the generic level. There is a nomenclatural point that must also be considered before discussing the rela- tionships of the genera described above. The family name Eupodiscaceae has long been used in the sense of a family based on the genus Eupodiscus J. W. Bailey, nom. cons., published in 1851. However, the family name was published by Kitzing in 1849 and is based on Eupodiscus Ehrenberg, nom. rej., the type of which is Eupodiscus germanicus (Ehrenberg) Ehrenberg, a synonym of Aulacodiscus argus (Ehrenberg) A. Schmidt. Eupodiscaceae is thus not the correct name for a family within the Biddulphiales. It is also an illegitimate name because Eupodiscus Ehrenberg is not only a rejected name but also an illegitimate superfluous substitute for Tripodiscus Ehrenberg. Solium and Pseudorutilaria resemble each other in most respects. The only basis for generic distinction between them is the presence in Pseudorutilaria of the troughs on the central portion, the raised edges of which grasp the occluded and labiate processes of the sibling valve. These we regard as a derived feature, and hence we postulate that Pseudorutilaria evolved from an ancestor which would, if it were known, be placed in, or very close to, the genus Solium. Whilst the fact that the subocelli of Solium and Pseudorutilaria consist of areolae and not porelli suggests that their affinities are with Biddulphia and Briggera rather than with Amphitetras and Odontella, many of their characters point in the other direction. Their mantles are concave, at least in part, as in Odontella and Amphitetras, not vertical as in Briggera and most species of Biddulphia, including its type. Anastomosing low external costae, such as occur on the elevations and mantles of Solium and Pseudorutilaria, are present throughout on Amphitetras but do not occur on Biddulphia or any of its allies. On the other hand, no species with a true ocellus has linking spines or internal costae. What is clear, however, is that Solium and Pseudorutilaria, in spite of their possession of linking spines, are not to be included in the Hemiauloideae sensu Sims (1986). The only genus that might be close to Solium and Pseudorutilaria is Monile. It closely resembles Pseudorutilaria in its shape of valve, in the many sulci crossing its projections, in the concavity of the valve mantle where the valve margin is convex, in the anastomosing costae on its elevations, in the arrangement of its poroid areolae, and in the presence of a subocellus. On the other hand, its areolae are occluded by cribra, not volae. It is only presence of a subocellus rather than a pseudocellus on its elevations, the flattened summits of these, and the ring of radial ridges on them that distinguishes it from Biddulphia, although the concavity of the mantle is unusual in that genus. The radial ridges on the summits of the elevations have the same pattern of arrangement as the linking spines of Strelnikovia R. Ross & P. A. Sims and Keratophora Pantocsek, genera which also have well developed pseudocelli. Which of these resemblances are indications of affinity and which are parallelisms cannot be decided on the basis of our present knowledge. The subocellus of Maluina is much smaller than that of Solium, Pseudorutilaria, and Monile and its areolae or porelli are much finer and more tightly and uniformly packed. These differences are so considerable as to suggest that Maluina may well not have any close affinity with those genera. Its elevations bear a small number of interlocking linking spines, but these are of a different shape from those in Solium and Pseudorutilaria (compare Fig. 67 with Figs 6 and 13) and it differs from those two genera in virtually every other respect. It seems to be an isolated genus and we can offer no suggestions as to its affinities. The similarities between Bonea and Strelnikovia are such as to indicate that they are related. The many interlocking linking spines and the large pseudocellus are the characters that point to 292 R. ROSS AND P. A. SIMS this most strongly. In Strelnikovia antiqua (Strel’nikova) R. Ross & P. A. Sims and S. inclinata R. Ross & P. A. Sims there is some sign of a vertical division of the pseudocellus at the apex of the valve (see Ross & Sims, 1985: pl. 21 fig. 5, pl. 24 fig. 6) and this is more strongly developed in Bonea (Fig. 73). We would therefore add Bonea to the group consisting of Briggera R. Ross & P. A. Sims, Dicladiopsis De Toni, Strelnikovia R. Ross & P. A. Sims, Keratophora Pantocsek, and Thaumatonema Greville. As we have already indicated (Ross & Sims, 1985; Sims, 1986), this group is much closer to Biddulphia than it is to Hemiaulus, and it should not be included in the Hemiauloideae. There are species of Biddulphia that differ from species of Briggera in nothing but the lack of interlocking linking spines, and other species that have all the characters of Strelnikovia except interlocking linking spines (see Ross & Sims, 1985: 288-290). There is the same similarity and difference between Bonea simulans R. Ross & P. A. Sims and Triceratium exornatum Greville. The significance of this for the supra-generic classification of the Biddul- phiaceae remains obscure, but it does suggest that the transition between diatoms with interlocking linking spines and those without them has occurred more than once. There is, though, no real indication of the direction in which it happened. One of us (Sims, 1986) has recently grouped together in the subfamily Hemiauloideae of the Biddulphiaceae the genera Hemiaulus Heiberg, Sphynctolethus Hanna, Pseudaulacodiscus Jousé, Ailuretta Sims, Cerataulina H. Peragallo, and Eucampia Ehrenberg. Trinacria Heiberg differs from Hemiaulus sensu stricto only in the absence of sulci or internal costae and in the number of poles and of labiate processes, and it also clearly belongs in the Hemiauloideae. Whether Dextradonator should also be included is more difficult to decide. It differs from Hemiaulus in the shape of the valve and of the linking spines, in the absence of sulci or internal costae, and in the pattern of areolation. Its very deep, coarsely areolate, entire valvocopula (Fig. 81) is another point of difference. This is a character that it shares with Abas R. Ross & P. A. Sims (1980), which it also resembles in shape of valve and pattern of areolation. We would for the present tentatively include both genera in the Hemiauloideae, with the proviso that they are not nearly so closely related to the rest of the subfamily as the other genera included in it are to one another. | Our main conclusion must be, then, that the information presented here makes the supra- generic classification of the genera that have been included in the Biddulphiaceae and the so-called Eupodiscaceae more difficult rather than less. That this is so is probably to be attributed to the fact that our knowledge of the Cretaceous and Paleogene diatom floras is still inadequate, and will perhaps always be so. Acknowledgements Our thanks are due to the late Mr A. L. Brigger and to Prof. R. W. Holmes of the University of California, Santa Barbara, for material presented to the British Museum (Natural History), which included many specimens studied whilst preparing this paper. We also wish to thank Dr Ruth Patrick and Dr C. W. Reimer of the Academy of Natural Sciences of Philadelphia, and Dr R. Simonsen of the Alfred-Wegener- Institut fiir Polar- und Meeresforschung, Bremerhaven, for the opportunity to examine collections in their charge. We have once again relied on information about the provenance and dating of samples collected by R/V Vema provided by Mrs Margaret Hanna of the California Academy of Sciences and Dr J. Fenner of the Geologisch-Palaontologisches Institut und Museum der Universitat Kiel. Type specimens of new species were mounted for us by Mr K. D. Kemp of East Brent, Somerset, and Mr S. J. Russell of our Department. We also thank the staffs of the photographic unit and electron microscope unit of the museum for their assistance. Mr A. Eddy kindly entered the manuscript on a word processor. References Bailey, J. W. 1851. Microscopical observations made in South Carolina, Georgia and Florida. Smithson. Contr. Knowl. 2 (8): 1-48, pl. 1-3. Benda, L. 1972. The diatoms of the Moler Formation of Denmark (Lower Eocene). A preliminary report. Beih. nov. Hedwigia 39: 251-266. Cheneviére, E. 1934. Sur un dépét fossile marin a diatomées situé 4 Kamischev (Russie centrale). Bull. Soc. fr. Microsc. 3: 103-107, pl. 5-8. GENERA OF THE BIDDULPHIACEAE 293 Cleve, P. T. & Moller, J. D. 1878. Diatoms. III Part, 9 pp. Uppsala. De Toni, G. B. 1894. Sylloge Algarum omnium hucusque cognitarum. 2 Bacillarieae. Sect. III. Cryptorhaphideae. pp. 819-1556. Patavii. —— & Levi, D. 1887. Algae novae. Notarisia 2: 333-353. Dzinoridze, R. N., Jousé, A. P., Koroleva-Golikova, G. S., Kozlova, G. E., Nagaeva, G. S., Petrushevs- kaya, M. G. & Strel’nikova, N. I. 1978. Diatom and Radiolarian Cenozoic Stratigraphy, Norwegian Basin; D.S.D.P. Leg 38. Initial Rep. Deep Sea Drilling Proj. 38: 289-427. & Strel’nikova, N. I. 1979. Opisanie Diatomovykh Vodoroslei [Descriptions of diatoms.] In Istoriya Mikroplanktona Norvezhskogo Morya (po materialam glubokovodnogo bureniya). /ssled. Fauny Morei 23 (3): 32-70. Ehrenberg, C. G. 1839. Hr. Ehrenberg las tiber jetzt wirklich noch zahlreich lebende Thier-Arten der Kreideformation der Erde. Ber. Akad. Wiss. Berlin 1839: 152-157. —— 1840. Uber noch zahlreich jetzt lebende Thierarten der Kreidebildung. 94 pp., 4 Taf. Berlin & Leipzig. Reimpr. in Phys. Math. Abh. K. Akad. Wiss. Berlin 1839: 81-174, Taf. I-IV (1841). —— 1844. Zum Schluss legte Hr. Ehrenberg der Akademie einige verlaufige Resultate seiner Unter- suchungen der ihm von der Siidpolreise des Capitain Ross, so wie von den Herren Schayer und Darwin zugekommenen Materialen tber das Verhalten des kleinsten Lebens in den Oceanen und den gréssten bisher zuganglichen Tiefen des Weltmeers vor. Ber. Akad. Wiss. Berlin 1844: 182-207. Fenner, J. 1977. Cenozoic diatom biostratigraphy of the equatorial and southern Atlantic Ocean. Jnitial Rep. deep Sea Drilling Proj. 39: 491-623. —— 1985. Late Cretaceous to Oligocene planktic diatoms. Jn H. M. Bolli, J. B. Saunders & K. Perch-Nielsen (Eds.), Plankton stratigraphy: 713-762. Cambridge. Galerkina, S. G. 1959. Stratigrafiya verkhnemelovykh u tretichnykh otlozhenii Berezovosko-Malo- Atlymskogo raiona. [Stratigraphy of the upper Cretaceous and Tertiary deposits of the Berezov, Mal, Atlym region.] Trudy vses. neft. nauchno-issled. geol.-razv. Inst. |VNIGRI] 140: 139-167. Glezer, Z. I. 1966. Kremnevye Zhgutikovye Vodorosli (Silikoflagellaty). Flora Sporovykh Rastenii SSSR. 7: 331 pp. Moskva. Translation: Silicoflagellatophyceae. Cryptogamic Plants of the U.S.S.R. 7: 363 pp. Jerusalem. —— 1969. Pozdneeotsenovye Kompleksy Diatomovykh, kremnevykh zhgutikovykh vodoroslei i ebriidei yugo-zapadnoi chasti Turgaiskogo Progiba. [Late Eocene complexes of diatoms, silicoflagellates and Ebriidae of the south-western part of the Turgai Gates.] Trudy vses. nauchno-issled. geol. Inst. 130 (4): 67-85. — 1975. K revizii roda Triceratium Ehr. sensu Hustedt, 1930 (Bacillariophyta). [Towards a revision of the genus Triceratium Ehr. sensu Hustedt, 1930.] Bot. Zh. SSSR 60: 1304-1310, tab. I-III. —— 1979. Evolyutsiya i sistematika poryadka Biddulphiales (Bacillariophyta). Paleont. Zh. 1979 (1): 110-120. Translation: Evolution and systematics of the order Biddulphiales (Bacillariophyta). Paleont. J. 13: 103-111. ; — 1983. Taksonomicheskaya znachimost’ priznakov u diatomovykh vodoroslei v svete razrabotki novoi klassifikatsii Bacillariophyta. [Taxonomic significance of characters in diatoms in the light of elaboration of a new Bacillariophyta classification.| Bot. Zh. SSSR 68: 993-1002. — 1986. O polozhenii roda Triceratium s. str. (Bacillariophyta) v sisteme diatomovykh vodoroslei. [On the position of the genus Triceratium s. str. (Bacillariophyta) in the diatom system.] Bot. Zh. SSSR 71: 1543-1545. ——, Jousé, A. P., Makarova, I. V., Proshkina-Lavrenko, A. I. & Sheshukova-Poretskaya, V. S. 1974. Diatomovye Vodorosli SSSR Iskopaemye i Sovremennye. [The diatoms of the USSR fossil and recent. | 1. 403 pp. Leningrad. Gombos, A. M., Jr. 1976. Paleogene and Neogene diatoms from the Falkland Plateau and Malvinas Outer Basin. Initial Rep. deep Sea Drilling Proj. 36: 575—687. — 1982. Early and middle Eocene diatom evolutionary events. Bacillaria 5: 225-242. — 1983. Middle Eocene diatoms from the South Atlantic. Initial Rep. deep Sea Drilling Proj. 71: 565-581. — 1984. Late Paleocene diatoms in the Cape basin. Initial Rep. deep Sea Drilling Proj. 73: 495-511. —— & Ciesielski, P. F. 1983. Late Eocene to early Miocene diatoms from the southwest Atlantic. Jnitial Rep. deep Sea Drilling Proj. 71: 583-634. Gray, S. F. 1821. A natural arrangement of British plants 1. London. Greville, R. K. 1863. Descriptions of new and rare diatoms. Series IX. Trans. microsc. Soc. Lond. II, 11: 63-76, pl. IV-V. — 1865. Descriptions of new and rare diatoms. Series XIX. Trans. microsc. Soc. Lond. II, 13: 1-10, pl. I-III. 294 R. ROSS AND P. A. SIMS Grove, E. & Sturt, G. 1886. On a fossil marine diatomaceous deposit from Oamaru, Otago, New Zealand. Part I. J. Quekett microsc. Club I, 2: 321-330, pl. XVII-XIX. Grunow, A. 1884. Die Diatomeen von Franz Josefs-Land. Denkschr. Akad. Wiss. Wien 48 (2): 53-112, Taf. I-V. — 1887. [Review of Grove & Sturt, 1886.] Bot. Zbl. 31: 131-133. Hajos, M. 1976. Upper Eocene and lower Oligocene Diatomaceae, Archaeomonadaceae and Silicoflagel- latae in southwestern Pacific sediments, DSDP Leg 29. Initial Rep. deep Sea Drilling Proj. 35: 817-883. Hasle, G. R., von Stosch, H. A. & Syvertsen, E. E. 1983. Cymatosiraceae, a new diatom family. Bacillaria 6: 9-156. Heiberg, P. A. C. 1863. Conspectus criticus Diatomacearum Danicarum. 135 pp., 6 Taf. Kébenhavn. Hendey, N. I. 1964. An introductory account of the smaller algae of British coastal waters. Part V: Bacillariophyta (diatoms). Fishery Invest., Lond. IV, 5. pp. i-xxii, 1-317, pl. I-XLV. Holmes, R. W. & Brigger, A. L. 1979. The marine fossil genus Entogonia Greville — a review. Bacillaria 2: 155-214. Hustedt, F. 1930. Die Kieselalgen Deutschlands, Osterreichs und der Schweiz. In Dr L. Rabenhorst’s Kryptogamen-Flora von Deutschland, Osterreich und der Schweiz 7 (1, 5): 785-920. Jousé, A. P. 1977. Atlas Mikroorganizmov b donnykh Osadkakh Okeanov. Diatomei, Radiolyarii, Silikoflyagellyaty, Kokkolity. Atlas of microorganisms in bottom sediments of the oceans. 32 pp., 160 Tabl. Moskva. Kitton, F. 1870. Diatomaceous deposits from Jutland. J. Quekett microsc. Club 2: 99-102. Kitzing, F. T. 1849. Species algarum. Lipsiae. Laporte, L. J. & Lefébure, P. 1929. Diatomées rares et curieuses. 1 P|. 1—XV. Paris. Lefébure, P. 1935. Diatomées contenues dans le dépét fossile marin de Saint-Laurent-la- Vernéde (Gard). Bull. Soc. fr. Microsc. 4: 44—57, pl. WI-IV. Miller, U. 1969. Fossil diatoms under the scanning electron microscope. A preliminary report. Sver. geol. Unders. Afh. C, 63 (5): 1-65. Miller, W. I. & Collier, A. 1978. Ultrastructure of the frustule of Triceratium favus (Bacillariophyceae). J. Phycol. 14: 56-62. Mukhina, V. V. 1974. Paleotsenovye Diatomovye ily v vostochnoi chasti Indiiskogo Okeana. [Paleocene diatomaceous ooze in the eastern part of the Indian Ocean.] Okeanologiya 14: 852-858. 1976. Species composition of the late Paleocene diatoms and silicoflagellates in the Indian Ocean. Micropaleontology 22: 151-158, pl. 1-2. ; Navarro, J. N. 1981. Asurvey of the marine diatoms of Puerto Rico. II. Suborder Biddulphiineae: Families Biddulphiaceae, Lithodesmiaceae and Eupodiscaceae. Botanica mar. 24: 615-630. Pantocsek, J. 1889. Beitrdége zur Kenntniss der fossilen Bacillarien Ungarns. II Theil. 123 pp., 30 pl. Nagy-Tapolcsany. Paramonova, N. V. 1964. Materialy po diatomovym vodoroslyam iz paleogenovykh otlozhenii severa zapadnoi Sibiri. [Materials on the diatoms from paleogene deposits of north-western Siberia.] Trudy vses. nauchno-issled. geol.-razv. Inst. 239: 232-246, 476-497. Proshkina-Lavrenko, A. I. (Ed.), Jousé, A. P. & Sheshukova, V. S. 1949a. Diatomovyi Analiz. Kn. 1. Obshchaya i paleobotanicheskaya kharakteristika diatomovykh vodoroslei. [Diatom analysis. General and paleobotanical characteristics of diatoms. ] 239 pp. Leningrad. ——.,, Kiselev, A. I., Poretskii, V. S. & Sheshukova, V. S. 1949b. Diatomovyi Analiz. Kn. 2. Opredelitel’ iskopaemykh i sovremennykh diatomovykh vodoroslei. Poryadki Centrales i Mediales. [Diatom analysis. Descriptions of fossil and recent diatoms. Orders Centrales and Mediales.] 238 pp., 101 Tab. Leningrad. Ross, R. 1985. Proposal to conserve Hemiaulus Heiberg against Hemiaulus Ehrenberg (Bacillariophyta). Taxon 34: 301-302. ——, Cox, E. J., Karayeva, N. I., Mann, D. G., Paddock, T. B. B., Simonsen, R. & Sims, P. A. 1979. An amended terminology for the siliceous components of the diatom cell. Beih. nov. Hedwigia 64: 513-533. & Sims, P. A. 1971. Generic limits in the Biddulphiaceae as indicated by the scanning electron microscope. In V. H. Heywood (Ed.), Scanning electron microscopy, systematic and evolutionary applications: 155-177. London. — 1980. Syringidium Ehrenb., Dextradonator Ross & Sims, nov. gen. and Abas Ross & Sims, nov. gen. Bacillaria 3: 115-127. 1985. Some genera of the Biddulphiaceae (diatoms) with interlocking linking spines. Bull. Br. Mus. nat. Hist. (Bot.) 13: 277-381. & Hasle, G. R. 1977. Observations on some species of the Hemiauloideae. Beih. nov. Hedwigia 54: 179-213. GENERA OF THE BIDDULPHIACEAE 295 Round, F. E. 1981. The ecology of algae. 653 pp. Cambridge. Rudkevich, M. Ya., Rubina, N. V. & Permyakov, A. I. 1957. Materialy k stratigrafii paleogena nizhnego Priob’ya. [Materials towards the stratigraphy of the Paleogene of the lower Ob basin.] Byull. Mosk. Obshch. Ispyt. Prir. 11, 62, Otdel Geol. 32 (1): 75-91. Schmidt, A. 1890. Atlas der Diatomaceenkunde. Heft 38. Taf. 149-152. Leipzig. —— 1893. Atlas der Diatomaceenkunde. Heft 46. Taf. 181-184. Leipzig. Schrader, H.-J. & Fenner, J. 1976. Norwegian Sea Cenozoic diatom biostratigraphy and taxonomy. Initial Rep. deep Sea Drilling Proj. 38: 921-1099. Schulz, P. 1927. Diatomeen aus norddeutschen Basalttuffen und -Tuffgeschieben. Z. Geschiebeforsch. 3: 66-78, 113-126. Schiitt, F. 1896. Bacillariaceae (Diatomeae). Jn A. Engler & K. Prantl, Naturl. PfiFam. 1 (1, b): 31-153. Shibkova, K. G. 1968. Diatomei paleogenovykh otlozhenii Yuzhnogo Kazakhstana. [Diatoms of Paleogene deposits of southern Kazakhstan.] Jn A. P. Jousé (Ed.), Iskopaemye Diatomovye Vodorosli S.S.S.R.: 21-26. Moskva. Shrubsole, W. H. & Kitton, F. 1881. The diatoms of the London Clay. J/ R. microsc. Soc. I, 1: 381-387. Simonsen, R. 1974. The diatom plankton of the Indian Ocean Expedition of R/V ‘Meteor’ 1964-1965. Meteor ForschErgebn. D, 19: 1-107. —— 1979. The diatom system: ideas on phylogeny. Bacillaria 2: 9-71. Sims, P. A. 1986. Sphynctolethus Hanna, Ailuretta gen. nov., and evolutionary trends within the Hemiauloideae. Diatom Research 1: 241-269. Smith, H. L. 1872. Conspectus of the families and genera of the Diatomaceae. II. Lens 1: 72-93. Stosch, H. A. von 1985. Some marine diatoms from the Australian region, especially from Port Phillip Bay and tropical north-eastern Australia. Brunia 8: 293-348. —— & Simonsen, R. 1984. Biddulphiopsis, a new genus of the Biddulphiaceae. Bacillaria 7: 9-36. Strel’nikova, N. I., Kaplan, A. A. & Travina, M. A. 1978. Paleogenovye diatomei, silikoflyagellyaty i ebriidei Kaliningradskoi oblasti. [Paleogene diatoms, silicoflagellates and ebriidei of Kaliningrad oblast.] Jn A. P. Jousé (Ed.), Morskaya Mikropaleontologiya (Diatomei, Radiolyarii, Silikoflyagellyaty, Foraminifery i izvestkovyi Nannoplankton): 57-66. Tempére, J. & Peragallo, H. 1889a. Diatomées, Fasc. 1. pp. 1-16. Paris. —— — 1889b. Diatomées. Fasc. 2. pp. 17-32. —— —— 1890. Diatomées. Fasc. 7. pp. 97-112. —— — 1891. Diatomées. Fasc. 15. pp. 225-240. —— — 1907. Diatomées du monde entier. Fasc. 1. pp. 1-16. Paris. —— —— 1908. Diatomées du monde entier. Fasc. 5. pp. 65-80. — —— 1911. Diatomées du monde entier. Fasc. 11. pp. 161-176. —— —— 1912. Diatomées du monde entier. Fasc. 22. pp. 337-352. — —— 1913. Diatomées du monde entier. Fasc. 27. pp. 417-432. Truan y Luard, A. & Witt, O. N. 1888. Die Diatomaceen der Polycistinenkreide von Jérémie in Hayti, Westindien. Berlin. Tsumura, K. 1964. Some kinds of Rutilaria and Pseudorutilaria. Bull. Yokohama City Univ. Soc. 16: 77-96, Pl. I-III. Van Heurck, H. 1883. Synopsis des diatomées de Belgique, Atlas, fasc. 6, pl. 104-132, 22 bis, 82 bis, 83 bis, 83 ter, 95 bis. Anvers. —— 1896. A treatise on the Diatomaceae, translated by Wynne E. Baxter. xx + 558 pp., 35 Pl. London. Vozzhennikova, T. F. 1960. Paleoalgologicheskaya kharakteristika mezokainozoiskikh otlozhenii zapad- no-sibirskoi nizmennosti. [Paleoalgological characteristics of the mesocaenozoic deposits of the West- Siberian plain.] Trudy Inst. Geol. Geofiz. sib. Otd. 1: 7-64. Walker, W. C. & Chase, H. H. 1886. Notes on some new and rare diatoms. (Series I). 7 pp., 2 Pl. Utica, New York. Description of plates Plates 1-11 are SEM micrographs. Plates 12-13 are LM micrographs. Plate 1. Solium exsculptum Heiberg. Figs 1-3, 5, 6: forma exsculptum, ‘Kamischev’, presumed to be Kamyshlov, Sverdlovsk oblast, U.S.S.R.; Fig. 4: forma pentagona Jousé, Northern Urals, Tyumen’sk oblast, U.S.S.R.; Figs 1, 3, 4: bar line = 10 wm; Figs 2, 5, 6: bar line = 2 wm. Fig. 1: valve exterior tilted and displaying four projections with elevations separated from central domed area by deep sulci, well-developed marginal ridge (arrowhead), numerous hollow spines (or occluded 296 R. ROSS AND P. A. SIMS processes), and single labiate process (arrow). Fig. 2: same specimen, detail of distal side of elevation with subocellus of areolae in radial rows positioned on upper part of elevation, linking spines, and concave mantle. Note also costae on mantle and elevation. Fig. 3: oblique view of frustule linked to single valve; fragments of girdle bands present. Fig. 4: oblique view of valve exterior of 5-polar form with concave mantle, sulci separating projections from central portion which has irregularly scattered areolae and an off-centre single labiate process. Fig. 5: distal side of linked elevations with clearly-defined subocelli apposed and single rows of poroid areolae between anastomosing costae. Note each areola is bordered by a rim with a single small spine and occluded by a vola. Fig. 6: proximal face of tips of two elevations linked by expanded and interlocking linking spines. Plate 2. Pseudorutilaria monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, Division Hill, Otago, New Zealand. Figs 7-9, 11: bar line = 20 wm; Figs 10, 12, 13, 14: bar line = 3 wm. Fig. 7: girdle view of two valves linked at elevations. Note central circular domed area, also regular rows of spines arising from marginal ridge between central portion and each elevation. Figs 8, 9: valve exterior with almost circular central portion and projections with crenulate margins gently tapering from central portion to each pole. Fig. 10: central portion with entire margin, sparse and scattered poroid areolae, irregular ring of troughs with raised margins (one broken occluded process (arrow) clasped within trough) mostly alternating with the fractured occluded processes (arrowhead). Fig. 11: two linked valves, oblique view of valve interior showing transapical costae positioned at each indentation of the valve margin. Fig. 12: valve interior with opening from single labiate process. Fig. 13: interlocking linking spines at tips of linked elevations on proximal sides. Fig. 14: oblique view of two linked elevations with a subocellus positioned just beneath each tip, linking spines on the proximal faces and solid spines arising from marginal ridge. Plate 3. Pseudorutilaria clavata R. Ross & P. A. Sims. Figs 15, 16, 18-23: south-western Atlantic, Falkland Plateau, 51° 08’S, 54° 22’ W, 1525 m depth, Vema cruise 17, core 107, 50 cm. Fig. 17: south-western Atlantic, Falkland Plateau, 53° 01’ S, 52° 52’ W, 2880 m depth, Vema cruise 18, core 104, 330 cm. Figs 15-17, 19, 20: bar line = 20 wm; Figs 18, 21—23: bar line = 5 wm. Figs 15-17: exterior views showing valve outline with almost circular central portion, gibbous portion on either side with each projection tapering to clavate rounded apices. Note marginal spines on marginal ridge and undulate valve face. Fig. 18: girdle view, two linked valves at centre, each with occluded processes and single labiate process (arrow) clasped within troughs with raised margins (two arrows) of opposing valve. Note girdle band with finely poroid areolae. Figs 19, 20: valve interior with four transverse sulci visible, positioned between central portion, gibbous portions, and projections. Note valve margin is entire, also note opening from single labiate process on fig. 20. Fig. 21: central portion with ring of occluded processes (broken at base), base of labiate process (arrow), also troughs with raised margins, one enclosing process of sibling valve. Fig. 22: distal side of low elevation with subocellus and four linking spines, one broken at base. Fig. 23: girdle view showing interlocking linking spines at tips of proximal faces of low elevations, apposed subocelli clearly defined, also flattened spines arising from marginal ridge. Note also deep furrow between mantle and hyaline marginal band, also anastomosing costae. Plate 4. Pseudorutilaria nodosa R. Ross & P. A. Sims. Figs 24-27, 29, 32: ‘Lamont E8’; Figs 28, 33: south-western Atlantic, Falkland Plateau, 53° 01’ S, 52° 52’ W, 2880 m depth, Vema cruise 18, core 104, 330 cm. Figs 24-28, bar line = 20 wm; Figs 29-33, bar line = 5 pm. Fig. 24: oblique view of two linked valves showing valve interior with numerous transverse costae. Fig. 25: oblique view of valve exterior showing undulate valve face, domed central portion, distinct marginal ridge and vertical mantle with deep furrow above narrow marginal band. Figs 26, 27: same specimens lying flat, displaying distinctive valve outline. Fig. 28: specimen with straight, not curved valves. Note position of gibbous portions and arrangement of internal costae identical with those above. Fig. 29: central portion, exterior view, with ring of fractured occluded processes alternating with troughs with fractured margins, also scattered poroid areolae. Fig. 30: linked elevations with subocelli on distal side and linking spines on proximal face, also long solid spines arising from marginal ridge. Fig. 31: single elevation with eroded linking spines. Fig. 32: two linked valves, central portion, with external openings from two labiate processes visible (arrows), one from each valve inclined and lying on opposing valve margin. Note also raised margins of troughs and indentation at base of mantle. Fig. 33: interior view, central portion, showing opening from single labiate process and transapical costae. Plate 5. Pseudorutilaria hannai R. Ross & P. A. Sims. South-western Atlantic, Falkland Plateau, 53° 01'S, 52° 52' W, 2880 m depth; Vema cruise 18, core 104, 330 cm. All figures are of the same specimen. Figs 34, 35: bar line = 30 wm; Figs 36—38: bar line = 5 wm. Fig. 34: two linked valves, interior view with central circular portion defined by transverse costae and curved projections regularly crossed by costae. Fig. 35: specimen in almost girdle view showing undulate margin and spines from marginal ridge of both valves alternating along complete length of valve. Fig. 36: GENERA OF THE BIDDULPHIACEAE 297 view at pole with interlocking linking spines on proximal face of each low elevation, subocellus missing on distal side and indentation or deep furrow between concave mantle and marginal hyaline band. Fig. 37: exterior opening of labiate process from upper valve (arrow) positioned between marginal spines. Fig. 38: central portion with slit-like opening from single labiate process (arrow). Note also scattered areolae and furrow between mantle and valve margin in central portion. Plate 6. Pseudorutilaria incompleta R. Ross & P. A. Sims. Figs 39—46: south-western Atlantic, Falkland Plateau, 51°08’ S, 54° 22' W, 1525 m depth, Vema cruise 17, core 107. Fig. 47: south-western Atlantic, Falkland Plateau, 47° 45-7’ S, 57° 38-5’ W, 3650 m depth, Conrad cruise 12, core 237. Figs 39-41: bar line = 20 um; Figs 42-44: bar line = 2 um; Figs 45—47: bar line = 5 wm. Figs 39, 40: valve exterior of separation valve and valve within chain displaying distinctive valve outline with crenulate margin, sulci producing undulate valve surface and the pentagonal central portion. Fig. 41: two linked valves tilted to show valve interior with sulci positioned between each crenulation and a large pentagonal central portion. Fig. 42: separation valve at pole with subocellus positioned just beneath tip of low elevation and single large spine on proximal margin. Fig. 43: distal side of linked elevations with discrete subocellus beneath tip of each elevation. Fig. 44: girdle view of same specimen with discrete subocelli and interlocking linking spines, areolae with vela, also marginal spines interdigitating. Fig. 45: central portion of separation valve with single straight labiate process, hyaline circular area surrounded by low flange-like costa, also scattered poroid areolae. Fig. 46: central portion, valve exterior with base of single labiate process (arrow), occluded processes and eroded troughs with raised margins arranged in circle. Fig. 47: central portion, valve interior, with opening from single labiate process, crenulate margin and internal sulci. Plate 7. Pseudorutilaria hendeyi R. Ross & P. A. Sims. Figs 48, 49, 53: south-western Atlantic, Falkland Plateau, 53° 01'S, 52° 52' W, 2880 m depth, Vema cruise 18, core 104, 330 cm. Figs 50-52, 54: south-western Atlantic, Falkland Plateau, 51°08’ S, 54°22’ W, 1525m depth, Vema cruise 17, core 107, 50 cm. Figs 48, 50: bar line = 20 wm; Figs 49, 51-54: bar line = 5 wm. Fig. 48: oblique view of triangular valve with central area defined by three sulci, long projections with undulate surface and rounded apices. Fig. 49: central portion fractured at centre but surrounded by remains of occluded processes, labiate processes and troughs. Fig. 50: interior view, two linked valves with transverse sulci proximally and costae distally across each projection and crenulate valve margin. Fig. 51: interior view, central portion, with openings from two labiate processes, scattered poroid areolae and patch of porelli. Fig. 52: fractured specimen, two linked valves, view looking into valve centre with occluded processes (arrow) and troughs with raised margins (arrowheads). Note marginal spines interdigi- tating. Fig. 53: single projection with low elevation. Exterior view showing position of linking spines on proximal face. Fig. 54: two linked valves, polar view of projection with low elevations, remains of well-defined subocelli and internal costae on upper valve. Plate 8. Monile laurentii R. Ross & P. A. Sims. St. Laurent-la-Vernéde, Gard, France. Figs 55—60: bar line = 25 wm; Figs 61-64: bar line = 5 wm. Figs 55, 56: single valves, exterior view, showing sinuous valve outline, domed central portion, sub-clavate apices and elevations with flattened summits. Fig. 57: oblique view, valve exterior, with vertical mantle on which there are three central and a polar indentation. Figs 58-59: interior view, single valves with transapical sulci positioned at each indentation of valve margin. Fig. 60: frustule with attached girdle band. Upper valve with undulate valve face, openings from four labiate processes on central portion and flattened summits to short, squat elevations. Fig. 61: central portion with exterior openings from four labiate processes projecting as fractured hollow spines, poroid areolae each with slightly raised rim with one or two small solid spines and occluded by cribrum. Fig. 62: interior view, central portion, with openings from four labiate processes. Fig. 63: proximal face of elevation with flattened summit encircled by marginal row of low teeth, areolae with cribra and interstrial pores. Fig. 64: distal side of elevation, expanded above, flattened summit bearing ring of low radial teeth on margin and subocellus on summit and expanded tip. Note anastomosing costae on elevation. Plate 9. Maluina centralitenuis (R. Ross & P. A. Sims) R. Ross & P. A. Sims. Figs 65-70: south-western Atlantic, Falkland Plateau, 51° 08’ S, 54° 22’ W, 1525 m depth, Vema cruise 17, core 107, 50 cm. Fig. 65: bar line = 20 um; Figs 66-70; bar line = 5 wm. Fig. 65: two linked valves with cylindrical elevations, vertical mantles with strongly developed pseudo- antiligulae at apices and domed centres. Fig. 66: detail of mantle with single row of poroid areolae and pseudo-antiligula, central domed area with radial rows of poroid areolae and single labiate process. Fig. 67: proximal face of cylindrical elevations with linking spines, expanded above and interlocking. Fig. 68: distal side of elevations, each with small subocellus at its tip. Note scattered areolae. Fig. 69: valve interior with large, slit-like opening from single labiate process, rows of fine areolae radiating from valve centre to margin of dome, also on outer valve face but sparse on vertical mantle. Fig. 70: valve exterior with small 298 R. ROSS AND P. A. SIMS circular hyaline central area, rows of fine poroid areolae radiating to base of domed centre, also opening from single labiate process projecting as a small tubular spine. Plate 10. Bonea simulans R. Ross & P. A. Sims. ‘Carlovo’, U.S.S.R. Figs 71, 72, 74: bar line = 20 wm; Figs 73, 75-77: bar line = 3 wm. Fig. 71: valve exterior, triangular valve with low elevations with flattened summits, marginal ridge between elevations bearing large spines, also vertical mantle. Fig. 72: valve interior showing weakly folded valve face with opening from single central labiate process. Fig. 73: elevation with summit surrounded by complete ring of expanded solid spines, those on the distal margin fractured. Note pseudocellus on summit and tip of elevation. Fig. 74: girdle view of frustule linked to two valves with two girdle bands present. Note pseudocellus on distal side of each elevation, also the interdigitating marginal spines. Fig. 75: distal side of elevations linked by numerous small expanded spines which interlock with those of the apposing elevation. Fig. 76: poroid areolae occluded by rotae with c. nine spokes, also single interstrial pore. Fig. 77: occlusion to smaller poroid areolae, also two interstrial pores. Plate 11. Dextradonator eximius (Grunow) R. Ross & P. A. Sims. Indian Ocean, 10° 25’ S, 63° 15’ E, 3115 m depth, dredge sample DODO-123-D1. Figs 78-83: bar line = 20 wm; Figs 84-85: bar line = 10 wm. Identity of specimen shown in Figs 82—3 doubtful (see p. 289). Fig. 78: two valves linked, upper with hyaline band (pleura) attached. Note linking mechanism involving two flattened and acute linking spines partially encircling elevation of sibling valve. Fig. 79: oblique view, same specimen, with hyaline marginal ridge turned inwards between elevations (arrows). Fig. 80: linked circular valves with broad vertical mantles, upper with attached valvocopula, lower with hyaline mantle and marginal ridge turned inwards. Fig. 81: single valve with attached valvocopula, cylindrical elevations each with two fractured linking spines. Fig. 82: linked valves, girdle view. Note two spines from each valve projecting beyond the girdle bands. Fig. 83: oblique view, valve face with small hyaline central area, radiating rows of fine poroid areolae extending to marginal ridge and slightly beyond base of elevations, also two long spines arising from marginal ridge. Fig. 84: distal side of elevation with few scattered areolae at tip, also circular valve face with marginal ridge turned inwards. Fig. 85: valve interior, showing central hyaline area, rows of fine poroid areolae radiating from centre, also broad valve margin. Plate 12. Figs 86—88: Solium exsculptum Heiberg. ‘Kamischev’ presumed to be Kamyshlov, Sverdlovsk oblast, U.S.S.R. Fig. 86: fo. exsculptum BM 81304: frustule with valves linked on either side, girdle view, 86-5 X 32 um. Fig. 87: fo. pentagona Jousé, BM 65833: 36 wm diameter. Fig. 88: fo. exsculptum BM coll. Adams TS748: full length one side, 67 wm. Fig. 89: Pseudorutilaria monile (Grove & Sturt ex De Toni & Levi) Grove & Sturt ex De Toni, BM 63397: Oamaru, New Zealand: length of specimens a: 136 wm, b: 97 wm, c: 130 wm. Figs 90, 91: Pseudorutilaria clavata R. Ross & P. A. Sims, BM 81136: south-western Atlantic, 51° 08’ S, 54° 22’ W, 1525 m depth; Vema cruise 17, core 107, 175 cm; Fig. 90: girdle view, two linked valves; apical axis 156 pm; Fig. 91: Holotype; apical axis 95 wm. Fig. 92: Pseudorutilaria nodosa R. Ross & P. A. Sims, Holotype, BM 81138: south-western Atlantic, 51° 08’ S, 54° 22’ W, 1525 m depth; Vema cruise 17, core 107, 50 cm; apical axis 146 wm. Fig. 93: Pseudorutilaria hannai R. Ross & P. A. Sims, Holotype, BM 81300: south-western Atlantic, 53° 01’ S, 52°52’ W, 2880 m depth; Vema cruise 18, core 104, 330 cm; apical axis 152 wm. Plate 13. Figs 94-95: Pseudorutilaria incompleta R. Ross & P. A. Sims, BM 81139: south-western Atlantic, 51° 08’ S, 54° 22’ W, 1525 m depth; Vema cruise 17, core 107. Fig. 94: Holotype; apical axis 98 pum. Fig. 95: two linked valves; apical axis 80 wm. Fig. 96: Pseudorutilaria hendeyi R. Ross & P. A. Sims, Holotype; BM 81142: south-western Atlantic, 51° 08’ S, 54° 22’ W, 1525 m depth; Vema cruise 17, core 107, 170 cm; apex to apex = 180 wm. Figs 97-98: Monile laurentii R. Ross & P. A. Sims, St Laurent-la-Vernéde, Gard, France. Fig. 97: Holotype, BM coll. Adams H941, specim. I, 12: apical axis 220 um. Fig. 98: BM 81303: apical axis 228 wm. Figs 99-100: Bonea simulans R. Ross & P. A. Sims, ‘Carlovo’, U.S.S.R.; Fig. 99: Holotype, BM 81301: length of side 119 um. Fig. 100: BM 81302: girdle view; length of side 140 um. PLATE | SR ecee a aa PLATE 5 303 PLATE 7 Ree? | PLATE 9 ee) he fe hd es “ "¥ ‘es. iw ee a “ : . oe ,, eenvr hb iy ’ x # * Haewnend setuvueter siinconasciabeanncete ne eae i) ea) e < o a i ‘ ‘ : : , ‘wh So oe sD PLATE 12 310 PLATE 13 311 4 a = British Museum (Natural History) An Enumeration of the Flowering Plants of Nepal H. Hara, W. T. Stearn, A. O. Chater & L. H. J. Williams Vol. 1. The Gymnosperms and Monocotyledons 1978, 275 x 215mm, 154pp, 7 figs. Paperback. 0 565 00777 7 £30.00 Vol. 2. The Dicotyledons (Part). 1979, 220pp. 0 565 00810 2 £30.00 Vol. 3. The Dicotyledons (Part). 1982, 226pp, 1 fig. 0 565 00854 4 £35.00 Titles to be published in Volume 16 Studies in the genus Hypericum L. (Guttiferae) 7. Section 29. Brathys (part 1) By N. K. B. Robson The lichen genus Ramalina in Australia By G. N. Stevens An annotated list of vascular plants collected in the valleys south of Mt Everest By G. Miehe Further genera of the Biddulphiaceae (diatoms) with interlocking linking spines By R. Ross & P. A. Sims Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk Printed in Great Britain by Henry Ling Ltd, Dorchester te, Std Fees | at ole a . + A ene pitas Se Ms Rua} Testes ieee be 8) Rat ay ocd) ‘i cate ‘ i Fan = wi 7 Aotet ey te gy iy tt, Paty’ Pita ‘ SS a =a as eters ase 5 oe ‘ Ta nb septate Dept er A Mita ot Vv PA ar net 4 aed 4 * y, A innleee Vaiss) wai sth toe dase eer i Bra + ; AG Boh Serageealt We ae NRCS iN Pat oth Ras Si DU ToS inearetie hit eta wah: 4 tte se a 4 . ove Cte ( oF SS = as PSS Sy =a i “a S q fi SAGA < wera Py ey th Ad 4 tee LP varie Aten, Paste Dey ct ate (3 Ker ell heh ats Aha eaves SAREE) i gna if * ae tt SSeaiek _ 4 ic Heat tetas ehclinaes) fy) yy mae : Se 4 vie. Latent anaes ty SS i Uy Ct AY Arkehe Ag alels Ms siea iS SN au si B “, 84"¢ ny A sail Se :