We. dete bb dee Le Ween Vena tine he Wedel: 14a Hematite Cee Fete Parente arte se eh ode Nahin eh Rebe ihe haeeth, rhe Oh po eubneal SCRE TP eieeerir oral wa mL tet yen, ‘ Mark PO rah at Een Tee ee ete RUST i eee Moet aren etek PASE COe renee Mw . ts wee Bary Ve" bey fs 4 " aN a ins aout st 4 ery Y te “ “s Apis han tsAe safle orersrtes yt ; : “ “ Peer ete areas te 8 rh Me bee ere Aah aan rer hes ana eRe te) ’ wan wrweyeaecr its ase, Barbet -« ene’ Seaton Notes a “* r ‘aie Ons ’ : atin ARR oN nar febee. ee r wes * Ne ad “ ba 2 poe Pee Ne gare, . x ~ + “ ras rmairge S ouke tryst inhi oe eae te abo te. ees £ Ay arenn etn tetera “ ees et ettinte canteen retard teen rid pewoer Sal nts SY are Ue LA eG ne ai A aide vi, ‘H: vi "C THE NATURAL HISTORY OF JUAN FERNANDEZ AND EASTER ISLAND eee DD BY AD Rh GARE SROLTSBEERG VOI il BO LAIN. WITH 116 PLATES UPPSALA 1920-1953 ALMQVIST & WIKSELLS BOKTRYCKERI AB LIBRARY NEW YORK BOTANICAL GARDEN Utgivet med understéd av Langmanska kulturfonden UPPSALA 1953 ALMQVIST & WIKSELLS BOKTRYCKERI AB Table of Contents. CHRISTENSEN, CARL and SKOTTSBERG, CARL. The Pteridophyta of the Juan Fernandez Islands CHRISTENSEN, CARL and SKOTTSBERG, CARL. Mhembems of Master Island. .-..5..... Fries, Ros. E. Die Myxomyceten der Juan BermamGez=lOSEIM ec. cis v0 2 = 2s ve ec oe Frigs, THORE C. E. Die Gasteromyceten der Juan Fernandez- und Osterinseln...... SKOTTSBERG, CARL. The Phanerogams of PS He TMS aa i atteraye ia eveheveteile enarsts erste neers MUNSTER STROM, K. Freshwater Algae from Juan Fernandez and Easter Island.... SKOTTSBERG, CARL. The Phanerogams of the anmebernandez dislandsy sca. cle ele > BrotHERvwsS, V. F. Musci Insulae-Paschalis BG6RGESEN, F. Marine Algae from Easter MeSTeinligparerociciess. cr etst's. ava yatete syreel ef ole esta versie SJ6sTEDT, L. GUNNAR. Ein neues Sargassum von der Osterinsel ZAHLBRUCKNER, A. Die Flechten der Juan iPierinarayclevaa) lnselhalengang dic ao erelcin occ cer BrotHERus, V. F. The Musci of the Juan Fernandez Islands ZAHLBRUCKNER, A. Die Flechten der Oster- insel nebst einem Beitrag zur Flechten- floraevon juan Bernandez..-.-....... Bove PETERSEN, JouHs. Marine Cyano- phyceae from Easter Island RoMELL, L. Bacidiomycetes from Juan Fer- TMZITIYG (EVA7 eA ICnGRO COREE DI CECRCECHCHE NER OROR ORE CuCH NCHOICRATC RS KEISSLER, KARL. Ascomyceten, Fungi im- 85 95 241 247 perfecti und Uredineen von Juan Fer- MENNGAs ooo son du aoOOOND DO OOOO Ob aNe6. SKOTTSBERG, CARL. The Vegetation of Easter UWAEHIKG! Golo g bcade Og OOD ODO ODS SoGoamad ¢ SKOTTSBERG, CARL. Pollinationsbiologie und Samenverbreitung auf den Juan Fernan- GEVZIGN AN gin. doco OOOO SOOO UO DE ETd.0O8 KEISSLER, K. Nachtrag zur Pilzflora von itiane Hemmandezpretnrryeratietrersretetaeietsuete Evans, ALEXANDER W. The thallose Hepa- ticae of the Juan Fernandez Islands .. SETCHELL, WILLIAM ALBERT. The Codiums of the Juan Fernandez Islands........ LEvRING, TorE. Die Meeresalgen der Juan Berman 6z=lrtSeliterstarerteler ists foucione etnies SKOTTSBERG, C. Marine Algal Communities of the Juan Fernandez Islands, with re- 473 487 593 601 marks on the Composition of the Flora 672 HeErRzoG, Tu. Die foliosen Lebermoose der Juan Fernandez-Inseln und der Osterinsel 697 LEVRING, TorE. Die Corallinaceen der Juan HennandezasimSelina sierra sieercmereictelenetels ete LEvRING, Tore. Einige Corallinaceen von der Oster-Insel SKOTTSBERG, C. Additional Remarks to FOOOGOC OOO DCOO do OmmAmS “Marine Algal Communities of the Juan Fernandez Islands” SKOTTSBERG, C. A Supplement to the Pte- ridophytes and Phanerogams of Juan Fernandez and Easter Island SKOTTSBERG, C. The Vegetation of the Juan Fernandez Islands aie je) e. elie e)'s \e 18)\e);0\ le) ekialiei leks (6) Ua ( } ir clay, ae - ee el La Ce a th Sabi nee Ath THE NATURAL HISTORY SOF JUAN FERNANDEZ AND BASPER TSEAND EDITED BY DR CARL’ SKOTTSBERG VOR! BOTANY - PARE. 3 1. C. CHRISTENSEN & C, SKOTTSBERG: The Pteridophyta . of the Juan Fernandez Islands, ~ 2. C. CHRISTENSEN & C. SKOTTSBERG: The Ferns of Easter Island. 3. Ros, E. Frigs: Die Myxomyceten der Juan Fernandez-Inseln, UPPSALA 1920 ALMQVIST & WIKSELLS BOKTRYCKERI-A-B. 1. The Pteridophyta of the Juan Fernandez Islands. By CARL CHRISTENSEN and CARL SKOTTSBERG. With 7 text figures and 5 plates. This paper is based on material collected by the undersigned, assisted by Mrs. SKOTTSBERG, 1916—17. ‘The study of the collection was begun in Sant- iago, where, thanks to the kindness of Mr. F. FUENTES, I had ample oppor- tunity to see the important herbaria in the Museo Nacional and also, with the kind permission of Professor F. JOHOW, could study his herbarium in the In- stituto Pedagojico. In order to solve some important questions and to classify some species which remained doubtful, I asked the prominent pteridologist, Dr. C. CHRISTENSEN, for assistance, and he willingly offered himself to make a thorough study of the entire collection. I have the great pleasure herewith to thank him for all the trouble he has taken in preparing the descriptions of the new species, in adding numerous critical notes on the systematic position and affinity and in making the fine and very exact drawings. The localities in Masatierra are enumerated from E to W along the north and south sides of the island, respectively; for Masafuera, the localities are enumerated from N to S along the east coast and then round to the west coast, followed by the observations made in the high mountain region, The Spanish names are used, with the following abbrevations: B. = Bahia, bay; C. = Cordon, mountain ridge; Co = Cerro, mountain; Pta = Punta, cape; Pto = Puerto, cove; Q. = Quebrada, valley with steep sides, gorge, canyon; V. = Valle, valley. — Maps will accompany Vol. I. Unless otherwise stated, all species have been observed by myself in all the places quoted. The numbers in brackets refer to our collection, of which specimens are kept in Stockholm (Riksmuseum), Gothenburg, Upsala etc., and in several foreign herbaria. A ! after the name of a collector signifies that I have seen the original specimen. An * before a latin name indicates a new addition to the flora of the group; if put before the name of one of the islands, it s'gnifies that the species was found for the first time in the island in question. The altitudes are in meters above sea level. I — 20206, The Nat. Hist. of Juan Fernandez and Easter Isl. ,Vol. II. 2 CARL CHRISTENSEN AND CARL SKOTISBERG The remarks on fertility refer to my observations in 1908’ and 1916—17; where I have reason to believe that a species occurs with ripe sporangia also at other seasons, these remarks are put in brackets. We have adopted the nomenclature followed by CHRISTENSEN in his »Index Filicum», where also explanations of the abbreviated book titles will be found. C. Skotisberg. Hymenophyllaceae. Trichomanes L. 1. T. Philippianum -Sturm, Enum. pl. cr. vasc. chil. 38 (1858); v. d. Bosch, Syn. Hymen. ed. Goddijn, Meded. Rijk’s Herb. Leiden no. 17: 24 f. 13 (1913); C. Chr. Ind. 647. — Sym T. dichotomum Philippi Anal. Univ. 1856: 169, Bot. Zeit. 1856: 650, Ann. sc. nat. II. 7: 110; Hemsl. 69; Johow 1893: 44, ° 1896: 173; zon Kze 1847. Fig... 1. Damp montane forests above 500 m., epiphytic on trunks of treeferns. (Fertile Jan.—Appril.) Masatierra: the high ridge between Q.* Laura and Q. Piedra agujeriada, on Dicksonia, c. 650 m (no. 594); C. Centinela, on trunks of ferns, 580 m (no. 616); NE. slope of La Damajuana, on Blechnum cycadifolium c, 500 m (no. 351); C. Salsipuedes, Q. Helechos, on Dicksonia, 660 m (no. 283). 7. dichotomum Phil. was described from specimens collected by GERMAIN (!). No locality except »Juan Fernandez» was stated. In Herb. Kew is a specimen collected by REED(!) and labelled » Valdivia», probably in PHILIPPI’s hand-writing (HEMSLEY, 1. c.). It is possible that this locality is false: REED made ex- tensive collections in both places and the localities may have been confounded. Until further investigations can be made, the species should be listed as endemic. T. Philippianum is one of the most distinct species of the genus. It has been described and figured in details by VAN DEN BOSCH, but his description and figure were not published until lately by GODDIJN. The numerous spe- cimens at hand correspond fairly well with this description, but the species grows much larger than his figure shows. The largest leaves are over 7 cm long including the 2 cm long stipe. The filiform, wide-creeping rhizome and the basal part of the stipes are sparsely furnished with reddish hairs, the plant otherwise being perfectly glabrous. The leaves are either scattered or fasciculate in small tufts of 2—7 on short branches from the rhizome. The lamina is exceedingly thin, yellowish green, the cells larger than in any other species of the genus, clearly seen 1 Botan. Ergebn. der Schwed. Exp. nach Patagonien und dem Feuerlande. IV. K. Svenska Vet.-Akad. Handl. Bd 51, No. 9 (1913). 2 For abbreviations, see p. 1. THE PIERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 3 through a weak lens. In shape and division the leaf varies considerably: generally it is oblanceolate, broadening upwards from the cuneate, shortly decurrent base; subdeltoid, deeply pinnatifid fronds, leaving a broad even wing to the midrib are, however, not uncommon. Segments erect, 2—2,; mm broad, obtuse or emarginate or deeply furcate at apex, very unequal in length, the larger ones often producing a secondary segment on the upper side near the base. Margins even, not at all thickened. Veins not regularly pinnatifid but rather repetitiously dichotomous, not atrofuscous as says BOSCH, terminating in the emargination; spurious veins none. Sori rarely terminal on the main branches ec. fle, Fig. 1. Z7ichomanes Philip/ianum Sturm. Plants showing different types of furcation, nat size, and enlarged sorus. of the veins but generally confined to the basal secondary segments, deeply immersed; indusium turbinate with a widened entire mouth; receptacle long exserted but easily broken off. VAN DEN Boscu placed the species in his genus Gonocormis, but it may be seriously doubted whether this position is natural. It looks very much different from all other species of that group, all of which belong to the Old World. In general habit and size it resembles the Polynesian 7. Aumi/le Forst., _ but it is certainly not nearly related to it. By its cellular structure it is quite unique. ; Area of distribution: Probably endemic in Masatierra. 2. T. Ingae C. Chr. nov. spec. Syn. 7. pyxzidiferum L., Hemsl. 69; Johow 1803: 45, 1896: 173. Eutrichomanes_e grege T. pyxidiferi L., rhizomate intricato, filiformi; — ie 4 CARL CHRISTENSEN AND CARL SKOTTSBERG foliis atroviridibus, planis, glabris, lanceolatis, maximis vix 5 cm longis et I cm latis, stipite I—1,; cm longo, e medio sursum anguste alato, rhachide tota alis sub 0,5; mm latis utrinque alata. Lamina bipinnatifida, pinnis 5—6-jugis, alternis, suberectis, infimis paulo abbreviatis, mediis 5—8 mm longis, inaequi- lateralibus, basi superiore segmentum soriferum gerentibus, parte exteriore in 2—4 segmenta sterilia contigua subtruncata vel leviter emarginata divisis. Mar- ginibus, praesertim versus apicem, minutissime et irregulariter crenato-dentatis. Venis distinctis, spuriis nullis. Cellulis parenchymaticis parvis, irregularibus, 4—6-angulatis, fere isodiametricis parietibus crassis, hyalinis. Soris singulis axillaribus, rarissime binis in pinna, tubo cylindrico anguste marginato, ore in- tegro vel levissime emarginato, paulo expanso; receptaculo exserto. Pig: 22. In moss mats on stones and on the ground, in dense forests, rare. (Fertile Aug.—Dec.) Fig. 2. Trichomanes Ingae C. Chr. a two fronds, nat. size; 6 fragment, X 21/2, ¢ sorus and d tip of sterile pinna, more enlarged. Masatierra, central part: V. Anson, NE. slope of El Yunque, 300 m (no. 162), in the same place where it was found in 1908; NE. slope of La Damajuana, c. 450 m; V. Colonial, Q Gutierrez (no. 68). — Found before only by MosELEY and by SKOTTSBERG, 1908. This new species, named in honour of Mrs. INGA SKOTYTSBERG, belongs to a number of closely related species which by most authors are united under the name 7. pyxzdzferum L. Convinced of this being a collective species that includes several valid ones, a’ number of which have already been described by VAN DEN BOoscH, we have tried, but without success, to identify the Juan Fernandez form with one or other of the forms from the American continent. Our new species differs greatly from the true West Indian 7. pyazdiferum by its much less divided lamina and by its dark colour; in these characters it approaches some South Brazilian forms, especially 7. emarginatum Pr., but it differs from these as well as from all other forms known to us by its cellular structure and its uneven margins. The cells are proportionally very small, very unequal in size and seldom longer than broad; by this structure the leaf becomes much firmer than in all other forms without spurious veins. Seen TRE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 5 through a strong lens, the margins of the outer parts of the segments are irregularly toothed or crenate by outwards vaulted marginal cells, the external wall being thick and hyaline. A similar feature was not found in a large number of specimens of Z. pyaidiferum sens. lat. examined. The numerous specimens collected have all the same characteristic habit, with the outer seg- ments of the pinnae very close. The diagnosis above was based on the larger fronds; most of them are considerably smaller. Area of distribution: Endemic in Masatierra. 3. T. exsectum Kunze, Anal. 47 tab, 29 f. 2 (1837); Hemsl. 69; Johow 1893: 44 f. 32, 1896: 173. — Syn. Aymenophyllum fuciforme Colla 1836: 33 tab. 63. In dense humid forests, epiphytic on trees and trunks of arboreous ferns, and in damp recesses of the narrow gorges, among moss, hanging down from the rocks. (Fertile Jan.—April.) Masatierra: C. Chifladores, above the innermost part of the Frances Valley, c. 500 m; ridge between Q. Laura and Q. Piedra agujeriada, on Dick- sonia 650 m (no. 595); C. Centinela, on fern trees 560 m; E] Pangal, wet forest in the gorge (no. 1130); B. Cumberland, in Cave VI, forming a beautiful pure carpet on the vertical walls (no. 598; also JoHow); Q. Juanango, under fallen trunks (no. 608); Q. Villagra, mountain spur W. of El Yunque, c. 500 m. Masafuera: Q. del Mono, c. 475 m; Q. de las Casas, under overhanging rocks in the canyon (no. 450, also JOHoW); Q. del Blindado, 440 m; Q. de las Vacas (JOHOW). The nearest ally of this pretty species, that reaches 4o cm in length, is T. tenerum Spr., widely spread through tropical America. Area of distribution: South Chile, Valdivia to Chiloé; Juan Fernandez. Serpyllopsis v. d. Bosch 4. S. caespitosa (Gaud.) C. Chr. Arkiv for Bot. 10: 29 c. fig. (1910). — Syn. Hymenophyllum caespitosum Gaud. Ann. sc. nat. 5: 99 (1825); Freyc. Voy. Bot. 374 tab. 5 f. 2 (mala); 77échomanes caespitosum Hook. Sp. Fil. 1: 132 tab. we B- C..’Chr. Ind. 637; Serpyllopszs antarctica v. d. Bosch Ned. Kr. Arch. 4: 377 (1859). Var. fernandeziana C. Chr. et Skottsb. nov. var. This interesting moss-like fern was first detected in the islands by SKOTTS- BERG, 1908. In 1910 C. CHRISTENSEN referred the specimens of the 1908 collection to var. densifolia (Phil.) C. Chr. 1. c. 30 (Aym. densifolium Phil. Linn. 29: 108, 1857), which was collected in the Guaitecas Islands and presum- ably is not rare in the humid forests of the Chilean Archipelago. The specimens from Juan Fernandez were sterile; now we have a beautiful lot of specimens, some of which are fertile, and an examination of the sori has proved that we 6 CARL CHRISTENSEN AND CARL SKOTTISBERG have to do with a form different from densifolia. It may be described as follows. Lamina luteo-viridis, pinnis plerumque ovatis secus costam parcissime pilosis, ut in var densifolia; differt soris profunde immersis utrinque late foliaceo- alatis, fere ad medium bipartitis, labiis rotundatis, distincte dentatis, extus 2—4 cristis longitudinalibus dentatis instructis, receptaculo ad maturitatem exserto. In the moss carpets on trunks and branches of trees, in the elevated forests of Masatierra, where it has not been found below 450 m; on Masafuera also in the heath of the “highland, among rocks. (Fertile in Jan.) Masatierra: C. Chifladores, in the innermost part of the Frances valley, on Drimys Wintert c. 500 m (no. 1128); ridge between Q. Laura and Q. Piedra agujeriada c. 650 m (no. 1129); on the main crest of the island, above Pangal, on Drimys c. 800 m (no. 1127); V. Colonial, along the road to Portezuelo, on Drimys c. 450 m (no. 136); C. Salsipuedes, brushwood on the ridge, 615 m, and in ©. Helechos, on Drimys, 660 m (no. 284). Masafuera: Heath above Q. del Mono, c. 850 m (no. 421); rocks at Las Torres, 1350-1370 m (no. 421 b—see below); S. slope of Los Inocentes, in Dicksonia-forest on the trunks, c. 950 m; also near the highest summit, 1450 m in moss mats (no. 381); Q. Loberia, between blocks in the stream. Var. densifolia has 1—8 sori terminal on the pinnae, the lips obscurely toothed and faintly crested, the wings on both sides much narrower. The typical form of the species, from the Falkland Islands, has terminal sori with very short lips; it turns brown when dried and is much more hairy, with more oblong pinnae. Var. fernandeziana is generally small with only 2—5 pairs of pinnae which are close, often imbricated and clasped along the midrib. There is, however, also a form with more distant and more oblong pinnae (no. 421 b), a form corresponding to var. elongata Hook. of the type; the special characters are due to external conditions (moisture and shade). Area of distribution: South Chile to Fuegia, Falkland Islands; the var. Jernandeziana endemic in Juan Fernandez. Hymenoglossum Pres! 5. H. cruentum (Cav.) Pr., Hymenoph. 35 (1843); v. d. Bosch, Ned. Kr. Arch. 4: 385. — Syn. Hymenophyllum cruentum Cav. Descr. 275 (1802) et auctt., Hemsl. 67; Johow 1893: 4o f. 26, 1896: 169; C. Chr. Ind. 359. Plate 1 Fig, 1. In the montane forest belt, with moss on the soil or on trunks of trees, scarce. (Fertile Dec.—Jan.) Masatierra, not rare in the central parts: V. Anson, NE. slope of La Damajuana c. 430 m (no. 353); slopes of El Yunque, 480 m; V. Colonial, C. Central 535 m; along the road to Portezuelo, c. 500 m (no. 67); north wall of Co Piramide, a narrow ledge 600 m; Q. Villagra, mountain spur W. of El Yunque, c. 500 m and SW, face of Co Pirdmide, among brushwood, 590 m. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 7 *Masafuera: Q. de las Casas, humid moss mats in the inner canyon, scarce (no. 473); C. del Barril, in moss on rocks, c. 750 m (no. 407). — New for this island. The specimens are small and sterile. In restoring the genus Hymenoglossum we follow VAN DEN BOSCH, the monographer of the family, contrary to all other pteridologists. Our reasons for doing so are not only the unique habit of the plant, but are based upon a very essential character, overlooked by PRESL and all later authors. The margins of the frond, not the sterile parts alone, but also the valves of the indusium, are bordered by a thick ebeneous line. The veins terminate in the parenchyma within this line. Area of distribution: South Chile, Valdivia to Chiloé and Western Patagonia to 49°; Juan Fernandez. Hymenophyllum Sm. 6. H. cuneatum Kze, Anal. 50 (1837); C. Chr. Ind. 359, Arkiv for Bot. 10: 21 (1910). — Syn. A. rarum auctt. quoad plant. amer.; Hemsl. 68; Johow 1893: 42 f. 30, 1896: 171. H. polyanthos auctt. quoad plant. fernand.; Hemsl. 67; Johow 1893: 42 (pro pa te? non f. 29), 1896: 172 (pro parte?). A. dichoto- mum, Johow 1893: 44 (pro parte?) f. 27, 1896: 173 (pro parte?). Fig. 3. : Probably fertile at any time of the year. Under this name whe unite all the smaller glabrous Hymenophylla of the islands, being convinced that they all belong to one species, although the ex- treme forms look very different. All forms agree in cellular structure and in shape of the indusium, which is broad, flat with a broad subtruncate or broadly cuneate base and with rounded or sometimes subacute valves. The indusium varies, as our figures show, in the same degree in the largest and smallest forms. KUNZE described the stipe as unwinged. We find among the very ex- tensive material several individuals which correspond perfectly to his description (no. 396); they have ovate or subdeltoid, middle-sized, yellowish-green fronds with wingless stipe and broad, semiorbicular indusia. Other plants agree with these in all respects but have the stipe winged to the base (no. 1145). We suppose ‘that these are younger plants, while fronds without wings are older, having lost the wings. Large or small fronds occur with or without a winged stipe. If no. 396 is considered to be a typical, middle-sized form, the other specimens may be arranged as follows. 1) Large plants, over 15 cm long incl. stipes, more decompound, with narrower, ovate-lanceolate lamina, often of a deeper green (no. 285). Such plants have been referred to H. polyanthos Sw. by all authors, and probably also the more typical form was brought to this species by some. It is very difficult to find a single constant character by which the larger forms may be known from the variable H. polyanthos; the lobes are generally broader, the 8 ; CARL CHRISTENSEN AND CARL SKOTTSBERG valves of the indusium more rounded and flat, but certain forms of HY. foly- anthos show essentially the same characters. The best mark is, perhaps, the glabrous base of the stipes in /7Z. cuneatum, while most forms of //. polyanthos have the stipes pilose below. 2) Smaller plants, often much reduced and approaching //. rarum R. Br. (nos. 146, 1143), and referred to that species by all authors. The leaves are yellowish green, I—3 cm long, the pinnae often more or less imbricate (no. 146). It must be admitted that some of these dwarf forms closely resemble H. rvarum in most characters, but we find leaves with and without a winged stipe which otherwise are identical, and //. varum has the stipe unwinged; further, all possible transitions between these dwarf forms and the larger ones are found. None of the specimens show the brownish colour of the typical ff. rarum. Summarizing what has been said above, we find it convenient to distin- guish two types. I. typicum. (//. cuneatum Kze, H. polyanthos auctt.) — Fig. 3 a—d. One of the commonest ferns, found in all the forests in moss carpets on the soil, on stones or on trunks of trees, in the eastern and central parts of Masatierra, where it ranges from about 200 m above sea level to the highest parts reached, viz. the main ridge above Pangal, 795 m. It is also common among the brushwood on the rocky ridges and, of course, on Dcksonia. — Probably fertile at all seasons (Dec.—April, Aug.). Masatierra: numerous localities from Pto Frances to Pto Ingles (nos. 46, 66, 70, 285, 300, 591, 1144, 1177; nos. 66 c and 7o form a transition to the following; both collected on rocks). Masafuera: Q. del Mono, in forest 570 m (no. 1146); Q. del Bindado in forest (no. 440); subalpine and alpine heath, 850—1370 m (nos, 396, 429, 1146, 1147); Q. Loberfa, in moss on boulders in the woods, c. 300 m (no. 1145). 2. var. rariforme C. Chr. et Skottsb. — Fig. 3 e—g. This is exactly the A. ravum of JoHOW(!), often found on old trunks of Blechnum cycadifolium, but also forming dense, almost moss-like patches on exposed ground. Leaves often 1 cm long only, generally deltoid in outline. Masatierra: At the road to Portezuelo, c. 450 m (no. 1143); on the sharp crest separating V. Anson from Q. Damajuana, on the peak, 365 m (no. 146). — f. imbricata; Fig. 3 h: Pto Ingles, the crest of the central ridge (no. 1178). Area of distribution: South Chile, Valdivia to Western Patagonia; Juan Fernandez. Tasmania? New Zealand? South Africa? 7. H. caudiculatum Mart. Ic. Crypt. Bras. 102 tab. 67 (1834); Johow Hoga-n4ierooo: 170; C. Chr. Arkiv for’ Bot. 107 23°(19rTo): In the the elevated forests of the foggy region of Masatierra, very rare; apparently more frequent in Masafuera, found both in the woods and in the bottom of the deepest canyons. — (Fertile Dec.—Febr.) THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 9 Masatierra: Just below the gap between El Yunque and La Dama- juana, c. 5cCO m (no. 160); North face of Co Pirdmide, narrow ledge, c. Fig. 3. Hymenophyllum cuneatum (Cav.) Pr. a@ typical cuneatum (no. 396), with enlarged sori, ’ indusium acute or rounded; in one are two sori; 6 large polyanthos-like frond with enlarged tip, showing acute indusia (no. 285); ¢ blunt sorus from a plant like the former (no. 66); @ poly- anthos-like frond with winged stipe (no. 1145); e—f vrariforme with stipe and indusiam of dif- ferent types (no. 1143); g small rarzforme (no. 146); h f. cmbricata (no. 1178). All plants nat. size. Io CARL CHRISTENSEN AND CARL SKOTTSBERG 600 m (no. 360). — Only collected once before in Masatierra: REED(!) without locality. *Masafuera: (. del Mono, dense forest, 475 m (no. 388); Q. de las Chozas, in the forest; Q. de las Casas, innermost part of the canyon, under stones below the cascade, 215 m (no. 465); QO. Loberia, under stones and in moss between huge blocks, 180—300 m (no. 481). — New for Masafuera, where fine, richly fertile specimens were collected in Q. de las Casas. Jhe specimens belong to the typical form. Area of distribution: South Chile, Valdivia to Western Patagonia (49° S. L.); Juan Fernandez. South Brazil. 5. H. fuciforme Sw. Syn. 148 (1806) et auctt.; C. Chr. Ind. 361 (where for »China austr.» read »Chile austr.»); Hemsl. 67; Johow 1893: 41, f. 28, 1896: 170; C. Chr. Arkiv for Bot. 10: 24 (1910). — Syn. H. semzteres Colla 1836: 22°) tab: voll: Humid montane forests in rich soil or on decaying trunks, not observed below 480 m. (Fertile Dec.—Jan.) 3 Masatierra: Slope of C. Chifladores, above the Frances valley, c. 500 m; V. Anson, slopes of El Yunque, 480 m;-near the road to Portezuelo, c. 500 m (no. 138); north face of Co Piramide, with the preceding; Q. Villagra, mountain spur W. of El Yunque, c. 500 m and at the road-side in the highest part of the Villagra valley below the pass, c. 500 m (also JoHow). This species, the largest of the genus, varies in cutting from tri- to quadri- pinnatifid. The less divided form is not, as we supposed in 1910, the only one found in the island. Specimens in the present collection are quite as much divided as specimens from Chile. Area of distribution: South Chile to Western Patagonia; Masatierra. 9. H. ferrugineum Colla 1836: 30; C. Chr. Ind. 361. — Syn. Asya tlisstmum Kze. Anal. 49 (1837); Hemsl. 68; Johow 1893: 43 f. 31. 1896: 172; FH. Berterot Hook. Sp. fil. 1: 93 tab. 33 C (1844); Gay 6: 534. Damp, shady forests of the highest mountain slopes, not found below 480 m; characteristic of the darkest corners in the fern groves. (Fertile in Jan.) Masatierra: C. Chifladores, woods in the highest part of the Frances valley, on Dicksonta c. 500 m; the ridge between Q. Laura and Q, Piedra agujeriada, c. 600 m and high up in this valley, 650 m; C. Centinela, under old- stumps, 530—s560 m; V. Anson, slopes of El Yunque, 480 m (no. 159, also P. ARREDONDO in Herb. JoHow!); V. Colonial, north face of Co Piramide, with the preceding; C. Salsiquedes, Q. Helechos, on Dicksonza 660 m (no. 76); Q. Villagra, high up near the pass, rare. *Masafuera: Q. de las Chozas, dark fern grove under Décksonia, c. 700m (no. 535). — New for this island. With this characteristic species most authors unite 17. /vankliniae Colenso from New Zealand, to which it is so closely related that it is not unnatural to THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS I! unite them into one species. Specimens from Chile are scarcely to be distin- guished from others from New Zealand, often having distant pinnae, the lower ones being much abbreviated. All specimens from Juan Fernandez, the type locality, have closer segments and an ovate lanceolate lamina, with the lower pinnae not or very slightly shortened. //. aeruginosum (Thouars) Carm. from Tristan d’Acunha is another near relative, but much smaller than either (see Hemsl. Chall. Rep. Bot. 1. 2, tab. 38). A fourth related form grows in the caves of Amsterdam Island in the Southern Indian Ocean. It was, we believe without sufficient reason, referred to /7/. obtusum Hook. et Arn. by HEMSLEY (I. c. 271), a Hawaiian species; more properly it might be reduced to A. aeruginosum as a variety. It is an interesting fact that four so closely related forms are scattered in remote localities over such a large area of the Southern Hemisphere. They are, we should guess, daughter forms of a single species that in old times was widely dispersed over the Antarctic continent. It must not be forgotten, however, that some of the forms may be derivates of the common tropical //. ciltatum. Area of distribution: South Chile, Valdivia to Fuegia; Juan Fernandez. New Zealand. *1o. H. falklandicum Baker, Syn. fil. 68 (1867); C. Chr. Arkiv for Bot. Bees fF. 3- (1910). Fig. 4 g—h. A new addition to the flora, growing both in the moss carpets of the forest floor and on trees, as well as in the alpine heath. (Fertile Feb.— March.) Masafuera: Q. del Mono, in forest, 570 m (no. 389); Q. del Blindado, in forest, 440 m (no. 398); Q. de las Casas, far into the canyon, with Hepatics, c. 200 m (no. 474); alpine heath near the Correspondencia Camp and at Las Torres, 1100—1I370 m (nos. 1142, 1176). Widely distributed through Subantarctic America and very likely not dif- ferent from H. Menzies? Pr. Hym. 51 from Staaten Island. Still PRESL de- scribed his species as having obovate, subglobose, pedicellate sori, while 7. _falklandicum has ovate-lanceolate valves of the indusium. The specimens from Masafuera are, as could be expected, larger than those from the heaths of Fuegia and the Falkland Islands and also of a more vivid colour, but in essential characters they do not differ. The largest fronds measure nearly 10 cm in length and are of the narrow lanceolate shape shown in fig. 3 a—b quoted above. The distinctly unilateral pinnae bear from 1 to 4 toothed segments on the upper side. The sori are single, or, in larger fronds 2 to each pinna, distinctly stipitate, with ovate-lanceolate, obtuse and quite - entire valves, which turn blackish brown with age. By its narrow lanceolate fronds and especially by the long dark brown valves very different from all forms of 17. peltatum. A specimen of the 1908 collection from Masafuera was 1. c. referred to Hi. peltatum. This is incorrect; probably the bad specimen should be referred to the next species rather than to H. falklandicum. 12 CARL CHRISTENSEN AND CARL SKOTTSBERG Area of distribution: Southern West Patagonia, Fuegia, Falkland Islands; South Georgia; Masafuera. #117. Hl. rugosum C. Chr. et Skottsb. nov. spec. Syn. HH. tunbridgense of Gay and Phil.? Hemsl. 69; Johow 1893: 44, 1896: 172. Fig. 4 a—/f. Leptocionium rhizomate filiformi, longe repente; foliis remotis, stipitibus erectis, filiformibus, glabris, omnino exalatis, saepe flexuosis, 3—4 cm longis. Fig. 4. a—f Hymenophyllum rugosum C. Chr. et Skottsb. @ well developed frond (no. 630); 6 leaf recalling HY. tunbridgense (no. 423); ¢ f. lanceolata (no. 1175); @ fragment of a, X 4; e sorus, X 10; f finna of c, X 4; g—h H. falklandicum Bak.; g two pinnae (no. 389), X 4; A sorus, X 10. All entire fronds in nat. size. Lamina formae typicae ovato-deltoidea vel ovato-lanceolata, 5—6 cm longa, 3—4 cm lata, brunnescente, perfecte plana, saepe curvata, subtus papillis bre- vissimis ad rachidem venasque rugosa, tripinnatifida; rhachide flexuosa, alata, alis denticulatis. Pinnis alternis, circ. 10-jugis, infimis plus minusve abbreviatis vel subaequalibus, aequaliter pinnatifidis vel basi inferiore paulo excisis, supre- mis minoribus distinctius inaequilateralibus. Pinnulis obliquis, inaequilateralibus, in segmenta 3—4 divisis, marginibus ubique acute denticulatis. Venis distinc- tissimis, nigriscentibus. Soris 1—8 pro pinna, praesertim in segmentis lateris superioris pinnarum insidentibus. Indusiis liberis nec immersis nec stipitatis, THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 13 fere ad basin in valvas duas ovato-elongatas, brunneas vel nigrescentes, apice distincte et acute dentatas divisis; receptaculo incluso. In the higher woods and humid heaths of the highlands, apparently not very rare. (Fertile Feb.—April.) Masatierra: Elevated slopes of Q. Piedra agujeriada, on fern-trees, 650 m (no. 592); C. Centinela, dark humid forests, c. 560 m (no. 1140); the longitud. ridge, above Pangal, on fern trees, c. 800 m; Q. Villagra, mountain spur W. of El Yunque, c. 500 m (no. 630). Masafuera: Woods on the Sanchez plain, 515 m (no. 1137); heath of the Correspondencia Camp, in the ravines, 1100—1200 m (nos. 423, 1139); C. del Barril, c. 750 m; Q, del Blindado, in the forest, on ferns, 440 m (no. 397); near the summit of Los Inocentes, c. 1400 m (no. 1138). f. /anceolata: minor, lamina lanceolata, 3—5 cm longa, raro supra I cm lata, alis rachidis vix denticulatis; habitu, magnitudine et pinnis omnibus inae- quilateralibus /7. /alklandico sat similis, sed differt et cum f. typica congruit soris non stipitatis nec non valvis distincte denticulatis. — Fig. 4 c¢. Masafuera, at the Correspondencia Camp, in the ravine, with the type, 1160 m (no. 1175). This new species is to HY. tunbridgense what H. falklandicum is to H. peltatum. It is very variable in size and shape of the lamina, but all forms have the same characteristic brown or dark brown indusia with elongated, toothed valves, similar in outline to those of 1. fa/klandicum, but never stipitate and always dentate (comp. Fig. ¢ and #). The larger, ovate-deltoid fronds recall 7. adichotomum, but they are quite plane and the stipe is wingless; the sori are very different, the mature sporangia never exserted. Middle-sized leaves with short, ovate-oblong lamina with all pinnae unequal-sided (Fig. 4) can scarcely be distinguished from // tunbridgense when sterile, but leaves with mature sori at once show the difference; true H. tunbridgense never has so elongated valves, which never are brown. There is little doubt of such forms having been referred to HY. tunbridgense, a species scarcely occurring in the islands. The f. /anceolata is very similar to H. falklandicum in general habit, but is easily known by its sessile indusia and toothed valves. The collection contains many intermediate forms between the extremes. Area of distribution: Endemic. 12. H. pectinatum Cav. Descr. 275 (1802); Johow 1893: 42, 1896: 170. (Fertile, March.) Masatierra: REED (!Herb. Kew) without exact locality. *Masafuera: In the highest part of the island, under rocks etc. in the alpine heath; the Correspondencia Camp, in the ravine, 1160 m (no. 1141); Las Torres, in wet moss under rocks, rare, 1370 m (no. 476); S. slope of Los Ino- centes, in a fern grove, 950 m, on Dicksonza, and in the heath near the summit, c. 1400 m (no. 378). — New for this island. The specimens are very poorly fructiferous; the largest ones have most of the secondary segments deeply cleft, thus somewhat resembling 7. secundum. 14 CARL CHRISTENSEN AND CARL SKOTTSBERG Area of distribution; South Chile, Valdivia to Fuegia; Juan Fernandez. “13. H.secundum Hook et Grev. Ic. Fil. tab. 133 (1829); C. Chr. Ind. 367. (Poorly fertile, Feb.) Masafuera: Las Torres, edge of the western precipice, 1370 m (no. 475). New to the islands. The specimens are small but otherwise typical. By the entire valves of the indusium surely to be distinguished from /7/. rugosum. Area of distribution: South Chile, Valdivia to Fuegia; Masafuera. 14. H. dichotomum Cav. Descr. 276 (1802); Hook. Sp. Fil. 1: 98 tab. 36 A et auctt.; Hemsl. 67; Johow 1893: 40 (pro parte?, certe non f, 27), 1896: 169 (pro parte); C. Chr. Arkiv for Bot. 10: 27 (1910). — Syn. A. nigricans Colla 1836: 32 tab. 62. H. polyanthos Johow 1893: 42 (pro parte?) f. 29, 1896; 171 (pro parte?) Probably fertile at almost any season. One of the commonest species, found on the low, dry slopes of western Masatierra as well as in the central and eastern parts, where it is frequently niet with in the wooded region from its lower limit to the highest ridges; also common in Masafuera. It grows on the soil, on stones or on trees, both fallen and living ones. Masatierra: eastern and central parts, common (nos. 45, 218, 571, 593, 1133, 1134); the dry western part: ridge SW of Co Tres Puntas, 300-350 m, under stones (no. 1132), and in the bottom of dried-up ravines (no. 1132 b). Masafuera: Q. del Mono, on trees, 475 m (no. 390); Q, de las Casas, in the canyon (no. 451); Q. del Blindado, common in the forest 440 m; Dicksonia- grove on the lower slope of Los Inocentes, above Q. del Varadero, c. 950 m; Q. Loberia, between blocks in the streem, 180 m (no. 480). This common species varies considerably in size and shape of the frond, measuring from 2—3 to 20 cm in length, in shape broadly ovate to lanceo- late, but all forms have the very characteristic plicate frond of a dark green colour. Young leaves have distinct brown, crisped wings along the stipe, much narrower, however, than those of H7. tortuosum; these wings are shed early so that older leaves have an unwinged stipe. The marginal teeth are much shorter than in //. ¢ortuosum. The sori are confined to the upper third of the frond (which is often more or less contracted), where they are numerous. The ovate-subacute, frequently apiculate valves of the indusium are entire and faintly or not at all spinose on their back, as described and figured by HOOKER. The large brown sporangia are exserted in the mature sorus. The smaller forms have the leaves very much contorted and look peculiar when compared with the larger ones. We think that 7. xzgrzcans Colla is this species; at least BERTERO’Ss spe- cimens in Herb. Kew(!) under this name clearly belong here and not to 7. tortuosum to which GAY referred COLLA’s species. Evidently this species has been largely confounded with others. It is perhaps the most common of the genus in the islands, but, nevertheless, JOoHOW only quotes other collectors and not himself, this because he misunderstood the species, for his fig. 29, 1893, is not »H. polyanthos», but 7. dichotomum, THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 15 while, on the other hand, his fig. 27 »H. dichotomum» illustrates a form of FH. cuneatum. Area of distribution: South Chile, Valdivia to Western Patagonia; Juan Fernandez. 15. H. tortuosum Hook. et Grev. Ic. Fil. tab. 129 (1829); Hemsl. 69; Johow 1893: 43, 1896: 172. Apparently rare, but easily confounded with the former. (Fertile Feb.) Masatierra: REED(!) in Herb. Kew. *Masafuera: Q. del Blindado, in woods, 440 m (no. 1136); heath at the upper course of Q. del Mono, c. 850 m (no. 1135). — New for this island. To this species we refer two sets of specimens from Masafuera. They differ from the common //. dichotomum in the rather persistent, broad, crisp, brown wings of the stipe, the non-plicate lamina, which is more or less undulate with much larger marginal teeth. The segments are closer and the larger pinnae have the pendent tips found in most specimens from the continent. The valves of the indusium are rather acute and nearly entire, thus approaching //. dichotomum. HA. tortuosum was also collected in Masatierra by PHILIPP1 and DOWNTON, t. HEMSLEY. We have not seen the former; the latter (Herb. Kew!) belongs to aichotomum. Area of distribution: South Chile, Valdivia to Fuegia, Falkland Islands; Juan Fernandez. [(H. reniforme Hook. is mentioned by PHiLippr (Anal. Univ. 1856: 160, Bot. Zeit. 1856: 631) who undoubtedly based his statement on GAY 6: 535. It has not been found by any of the known collectors, and GAY does not tell where he got his information as to its occurrence in Juan Fernandez. As far as we know there are no specimens in any herbarium from this place, so that it must be excluded from the list. FT. dicranotrichum (Pr.) Sadeb. (Syn. H. chiloense Hook,). is listed by JoHOwW 1893: 44 and 1896: 172, as: gathered by REED in Masatierra. The determination is probably incorrect. In Herb. Kew.(!) are specimens of A. dentatum Cav. (Syn. H. Bridges Hook.) labelled (by PHILippl) Juan Fernandez oct. 1872 REED. As no author lists this species, the locality remains doubtful.| Cyatheaceae. Thyrsopteris Kunze 16. T. elegans Kze Linnaea 9: 507 (1834), Farrnkr. tab. 1, et auctt.; Hemsl. 66; Johow 1893: 22 f. 4, 1896: 155. — Syn. Panicularia Berteri Colla 1836: 35 tab. 64. Characteristic of the upper montane forests in Masatierra, generally not descending below 400 m; belonging to the highland heath in Masafuera, where 16 CARL CHRISTENSEN AND CARL SKOTTSBERG it is rather poorly developed. The fertile pinnae are developed during the summer; in the beginning of December we collected fronds with very young sori and still in the end of January no ripe ones were observed; in the later half of March the sporangia were ripe and the spores disseminated. Masatierra, not rare in the central part: C. Centinela c. 530 m; V. Anson, slopes of El Yunque, 480 m; C. Central, c. 500 m; along the road to Portezuelo in several places (lowest station observed 208 m, at a small stream), especially near the pass, c. 500 m (no. 14); C. Salsipuedes, open brushwood on the narrow ridge, 615, 625 m; Q. Salsipuedes, light woods, c. 500 m, numerous; ©. Helechos, not rare in Dzcksonza-groves; Pto Ingles, on the central ridge, 470 m; Q. Villagra, frequent in the humid woods below the pass (no. 277), but also seen on a low ridge just below the lower limit of trees, c. 200 m. Masafuera: ©. de las Casas (JoHOW); subalpine heath, in the shallow fern-clad beds of the streams, not rare from the Sanchez plain to the Corres- pondencia ravine, IOOO—II00 m (no. 542); C. del Barril, g00—1000 m, not rare; on the ridge forming the south wall of QO. de las Vacas, not rare above 1000 m, Lcksonza-groves on the lower S. slope of Los Inocentes, etc. etc. The specimens from the two islands are fully identical. Area of distribution: Endemic. Lophosoria Pres] 17. L. quadripinnata (Gmel.) C. Chr. — Syn. Alsophila quadripinnata (Gmel.) C. Chr. Ind. 47 cum syn.; A. prummata (Sw.) KIf. et auctt.; Hemsl. 66; Johow 1893: 20 f. 2, 1896: 154. One of the most common ferns, growing in widely different stations, and ranging from the ravines of the treeless basal region to the highest woods and heaths. Found with ripe sori in the autumn and winter. Masatierra: In all the quebradas between Pto Frances and Pto Ingles, common both in forested and in treeless tracts; numerous ravines near the colony, representing the last trace of indigenous vegetation; also on the elevated ridges separating V. Colonial from Villagra, 500—625 m (no. 590). Masafuera: Forming extensive beds in the subalpine and alpine region, ranging nearly to the highest summits (1350—1400 m), also on the highest parts of the great western precipice; not uncommon in the forests and along the streams of the canyons etc., e. g. in Q. de las Casas (no. 458). This interesting species which, as BOWER has proved, is generically distinct from Johow 1893: 33 i. 17,1500. Polystichum orbiculatum C. Chr. Arkiv for Bot. 10: 18 (1910). Fairly common in well-watered ravines and in the woods, from a few hundred m to the elevated ridges, in Masafuera also in the alpine region. Fertile specimens may be seen at Bpeny time, but are probably commoner during the summer and autumn. Masatierra: C. Chifladores, slopes above Pto Frances, c. 500 m, common in the forest; C. Centinela, from 530 m upwards; Q. Damajuana, waterfall at 250 m (no. 53); C. Central, in clearings, c. 400 m; quebradas in the bottom of V. Colonial (nos. 48, 152, 334), also in thickets of Avzstotelza maqgui near the road to Portezuelo (no. 1174); C. Salsipuedes, brushwood on the ridge, THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 21 615 m and in Q. Helechos, 660 m; Q. Juanango, c. 200 m; Q. Villagra; Q. de la Choza, in forest 250 m. Masafuera: Q. del Mono, in the woods, 4—500 m; Q. del Blindado, forest, 440 m; Q. de las Casas, in the canyon (no. 555); Q. Angosta, c. 1 km from the entrance, at the waterfall; the Correspondencia Camp in the ravine, 1160 m; edge of the western precipice, at Las Torres, 1370 m; summit of Co Correspondencia, 1420 m. It may be doubted whether the acu/eatum-form of Juan Fernandez really is identical with P. vesttum. In most characters it certainly agrees with speci- mens from Australia while, on the other hand, it differs much from the more common form of New Zealand (P. vexustum Hombr. et Jacq.). It is dark green above, rather soft in texture, the basal scales black with pale borders, the scales of the rhachis reddish, thin, tapering from the broad lacerate base into a long, contorted, hairlike apex but always lacking the black scales of P. venustum (= P. vestitum Christ Arkiv for Bot. 4: 5). The upper basal pinnules are often much lengthened and fully pinnate. The reference in 1910 to P. orbiculatum was due to the misinterpretation of the latter by CHRIST. Area .of distribution: Taken in a wide sense, the species inhabits Sub- antarctic America, Juan Fernandez, Australia and New Zealand. Arthropteris J. Sm. 22s. altescandens, (Colla) J. Sm: Hist, Bil. 225;.C.Chi. Ind..62. — Syn. Polypodium altescandens Colla 1836: 48; Nephrolepis altescandens Baker, Hemsl. 76; Johow 1893: 35 f. 20, 1896: 165. Polypodium procurrens Kze Anal. 17: ; Only found in the lower parts of the forest region, trailing over stones and fallen stems or climbing to a considerable height in the trees. (Fertile Aug., Dec.—April.) . Masatierra: Pto Frances (JoMNOW); Pangal, common from one hundred m upwards; Q. Damajuana, c. 250 m, very common on the trees (no. 60); V. Anson, in the woods surrounding the Plazoleta del Yunque; C. Central, c. 400 m, numerous; V. Colonial, quebradas in the interior, not uncommon 250— 350 m (no. 44); Q. Juanango. Masafuera, rare: Q. de las Casas (detected 1908), in the canyon; in 1917 only seen on a tree fallen down from the steep wall (no. 507). This endemic fern is a pronounced Polynesian element in the flora. No American species is known to which it might be compared, while, on the other hand, some Polynesian forms, especially from Samoa, which have been referred to the collective species A. obliterata (R. Br.) J. Sm. are intimately related to our species. Other very close relatives grow in Queensland (see DoMIN, Bibl. Bot. Heft 85: 60—64 c. figs.). The African A. ramosa, generally but incorrectly united with the eastern Polynesian forms, and the Australian A. feve//a are both more remotely related to it. The main difference between dA. a/fescandens and the Samoan form mentioned, which are both green when dried and appar- 22 CARL CHRISTENSEN AND CARL SKOTTISBERG ently exindusiate, lies in the base of the lamina. The Samoan plant has a gradually narrowed base leaving only a very short stipe free of pinnae; A. altescandens has the stipe 3—5 cm long and the lamina scarcely narrowed downwards, although frequently a single shortened basal pinna is present. The midribs of the pinnae are softly villous with articulate hairs in A. al/tescandens ; such hairs occur only sparingly in the Polynesian species. Area of distribution: Endemic. Asplenium L. 24. A. obliquum Forst. 80 (1786) var. chondrophylium (Bert.) Mett. Aspl. 137 no. 21. — Syn. A. chondrophyllum Bertero ex Colla 1836: 40 tab. 68; A. obtusatum Hemsl. 74; Johow 1893: 31 f. 14, 1896: 162; C. Chr. Ind. 124 pro parte. Along the sea shore, in crevices and under overhanging rocks. Probably fertile the year round. . Masatierra: B. Cumberland (JoHOw); Puerto Ingles, not uncommon (no. 313, also listed by JoHOW). | Santa Clara: JoHOW. Masafuera: Not uncommon on the cliffs from Q. del Ovalo along the coast to Pta Negra and probably also in other places (no. 511); also in the interior of Q. de las Casas. The differences between the two species of FORSTER, A. obtusatum and A. obliquum, were clearly explained by METTENIUS, and we do not doubt that they are both good species, distinguished besides by the number and shape of pinnae especially by the different scales on the lower side of the lamina. In A. odtw- satum these scales are minute, dilacerate, in A. ob/zquum ovate, entire, acuminate. The Juan Fernandez form agrees with A. ob/zquum in its scales, but the more common, smaller form, figured by COLLA, has much the aspect of A. obtusa- tum; still forms have been collected that scarcely differ in size from A. ob/z- quum, so that it seems quite natural to refer it to that species as a variety. From both the Australian species the variety differs by the colour of the basal scales, which are reddish brown and not dirty ferrugineous as in the others. Similar forms from Chile were described as A. sphenozdes Kze and A. consimile Gay; they have the dirty brown basal scales of typical od/¢guum, and their rather numerous pinnae are shortly acuminate, the margins duplicato- serrate with rather acute teeth, while var. chondrophylum sometimes has nearly entire, sometimes obtusely serrate, obtuse pinnae. Area of distribution: South Chile; Juan Fernandez; Australia, New Zea- land. Typical chondrophyllum only known from Juan Fernandez. 25. A. macrosorum Bertero ex Colla 1836: 39 tab. 67 (bona); Kunze, Anal. 21 (1837); Hemsl. 74; Johow 1893: 32 f. 16, 1896: 162. pllate i; igk.2: THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 23 Apparently not at all so rare as JonowW believed; characteristic of rich humous soil in the montane woods within the foggy region. (Iertile Dec.— March.) Masatierra: C. Chifladores, woods in the highest part of the Frances valley, c. 500 m, rare; dense forest on the ridge between Q. Laura and Q. Piedra agujeriada, c. 600 m, and in the last-mentioned quebrada, 650 m; Ra- banal, in the upper part; C. Centinela, 560 m (no. 1179); V. Anson just below the gap between La Damajuana and El Yunque, in rich woods, c. 500 m (no, 157; also P. ARREDONDO in herb Jonow!); C. Central, 390 m, small and very rare, scattered at 500—550 m (no. 302); C. Salsipuedes, Q. Helechos, c. 650 m, one tuft under a stone in the Décksonza-forest (no. 94); Q. Villagra, mountain ridge W. of El Yunque, c. 500 m, rare. *Masafuera: Q. de las Chozas, rich woods above 500 m (no, 364); C. del Barril, fissures in the rocks, c. 750 m (no. 406); SE. slope of Los Inocentes, Dicksonia-forest, 840 m (no. 533). — New for this island. A very distinct species with doubtful relationship: several species partly resembling it are found in many regions of both hemispheres, still it is proeb- able that it might be placed next to A. obtusifolium L. and semzcordatum Raddi, both widely dispersed through tropical America. It differs from both by its shining, dark castaneous stipe and rhachis, its few pinnae and large sori. - The texture cannot be called subcoriaceous as stated in Syn. Fil., but is rather thin although firm, pellucid. It varies in the number of pinnae; generally there are I—2 to each side of the terminal large one; they are equally cuneate or sometimes unequal with the lower side more arcuate, quite as in the two spe- cies mentioned above. In the larger specimens from Masatierra the base of the terminal pinna often has one side arcuately excised. The indusium, which is greyish, entire and often a little arcuate, covers the young sorus; in the mature ones, '/2:—2!/2 cm long and 21/2 mm wide, the indusium is quite hidden by the reddish brown sporangia. Among the specimens from Masatierra there is a luxuriant form, with 3—4 lateral pinnae on each side, the basal ones with a nearly free secondary pinnule on the lower side. The plants from the two islands are as a whole similar, still those from Masafuera are smaller than those from Masatierra with narrower pinnae, the terminal one equal-sided. Area of distribution: Endemic. 26. A. stellatum Colla 1836: 41 tab. 69; C. Chr. Arkiv for Bot. 10: 5 (1910). — Syn. A. fernandesianum Kunze, Anal. 22 (1837); A. dunulatum Hemsl. 73; Johow 1893: 32 f. 15, 1896: 162; C. Chr. Ind. 119 pro parte (zon Sw.). Not uncommon in the montane forests from Pto Frances to Vaqueria, also found in more open places at a lower altitude; less common in Masafuera,. (Fertile Aug., Dec.—April.) Masatierra: Pto Frances, brushwood on Loma Incienso, c. 350 m, for- ming a carpet on earth-covered stones (no. 128); on the slope of C. Chifla- dores, rocky soil in the forest, c 500 m; La Damajuana, c. 450 m,; NE. slope 24 CARL CHRISTENSEN AND CARL SKOTISBERG of El Yunque (no. 1149); C. Central, c. 470 m (no. 309); just below the Porte- zuelo pass, c. 500 m (no. 133); north face of Co Piramide, c. 580 m; Vaqueria, c; 250m. Masafuera: Q. de las Chozas, in dense forests, 5—600 m (no. 545); C. del Barril, among rocks, 985 m (no. 532); QO. de las Casas, under overhanging rocks in the bottom of the canyon (no. 509). A considerable number of specimens show that this is a variable species, varying in size, and in shape and cutting of the pinnae. COLLA’s figure is not good as it does not show the shortened lower pinnae, which are more or less fanshaped with flabellate venation. The leaves often produce new plants at their tips, whose leaves are again proliferous. The stipe is some times winged with the wing broadest below, sometimes unwinged. Lamina more or less thin, dark green (yellowish in open places), the rhachis narrowly alate throughout. In some specimens the pinnae resemble COLLA’s plate, but they are often shorter and broader, with a rounded or obtuse apex, commonly with a lobed auricle at the upper base, more or less excised at the lower, the large lobes entire or cleft. A. stellatum belong to the A. lunulatum-assemblage, spread through the tropics in many varieties. Recently HIERONYMUS (Hedwigia 60, 1919) de- scribed or restored a series of species of this group. Under the new species A. tabinense he mentions A. fernandeztanum Kze, a younger name for A. séel- latum. A. stellatum seems to combine several characters, of which one or more are found now in one, now in another of the allied species. In general habit it mostly resembles the South African A. crectuim Bory, in its winged rhachis and stipe d. pleropus KIf., common in tropical America. From the latter, which may be considered as its nearest relative, it differs in the coal black, rigid basal scales. Area of distribution: Endemic. 27. A. magellanicum KIf. Enum. 175 (1824); Hemsl. 74; Johow 1893: 32, 1896: 163. Found in several distant stations, but not common; chiefly in the woods or in wet moss near running water. Probably fertile at any season. Masatierra: C. Chifladores, dense forest on the slope towards Pto Fran- ces, c. 500 m; Pangal, at the cascade, in wet moss; Q. Damajuana, at the foot of a small waterfall (no. 61); in the woods on both sides of Portezuelo: along the stream, on the ground and on trees. Masafuera: Forest on the Sanchez plain, 515 m; QO, de las Chozas, dark ravine under Dicksonza; Q. del Mono, dry bed of the stream, 570 m (no. 461); QO. de las Casas, in the canyon (no. 452); Q. del Blindado, on the ground and on trees and arboreous ferns, 440 m; Q. Angosta, about 1 km from the en- trance, at the waterfall; ©. Loberia, along the stream and in the forest patch at 200 m; the Correspondencia Camp, in the Lophosorza-beds, and in moss mats at the foot of Las Torres, 1370 m. The island specimens do not quite agree with continental ones. They are THE PLERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 25 a little larger with an ovate-oblong lamina (sometimes even broadly lanceolate), rarely deltoid as the common form from southern Chile; the lobes are broader and more distinctly toothed at the outer edge and the texture thinner. In general outline, the type recalls A. rufa muraria, the island form rather A. adiantum nigrum, Some specimens are seen on Plate 1 Fig. 1. Area of distribution: Southern Chile to Fuegia and Falkland Islands; Juan Fernandez. Blechnum L. 28. B. auriculatum Cav. Descr. 262 (1802); C. Chr. Ind. 151. Syn. B. hastatum K\f. Enum. 161; Lomarta hastata Kze Farrnkr. tab. 55 f. 2; B. australe Hemsl. 73; Johow 1893: 26 f. 9, 1896: 158. Fertile at any time of the year. Masatierra: Never wanting on the low, treeless slopes, sometimes very common on the loose tufas, where few other species grow, so in V. Colonial etc. etc. In inhabits the dry regions between Villagra and Pta Larga. It is also found in all the forests (no. 55) and on the high ridges among shrubs. It grows in almost all the caves near the landing-place in B. Cumberland (no. 586, from cave II is a large luxuriant form). Santa Clara: JOHOW. Masafuera: Very common (no. 493). According to our opinion this species is not conspecific with the South African 2B. australe L. It varies considerably in the islands and JoHOw di- stinguished three varietis. The present collection contains the following forms. 1) A large form, like the continental one, fronds up to 70 cm long (stipe 20 cm) and 8 cm broad, lower reduced pinnae distant, sori often interrupted. This is JoHow’s var. 1, fig. 9 a (well illustrated by KUNZE, |. c.). JOHOW’s var. 2, fig. 9 b is only a slender form of the same. 2) remotum (Presl). — Syn. B. remotum Pr. Tent. 103 (1836, nomen); Mesothema remotum Pr. Enum. 111 (1849); B. pubescens Hook. Ic. plant. tab. 97 (1837), Gay 6: 478 (non Desv.); Lomaria pubescens Kze Farrnkr. 1: 122 tab. 55 fig. 3 (1843). Prominent pteridologists have regarded this as a distinct species. A com- parative examination of a large material shows that no definite limit can be drawn between this and typical B. auriculatum. B. remotum is smaller and has a much shorter stipe, further it is densely pubescent on stipe and rhachis, characters which undergo great changes according to the external conditions. B. remotum is the form of more open situations. On the arid slopes and higher, stony ridges another reduced form is found, which might be called f. parvala (Phil.). — Syn. B. parviulum Phil. Anal. Univ. 1873: 580 (t. spec. Mus. Nac. Santiago!); C. Chr. Ind. 158. This has been mistaken for 4. penua marina (Poir.) Kuhn (= Lomaria alpina Spr.), by Hemsl. 72; Johow 1893: 31, 1896: 161 doubted its existance. In the present collection there are specimens of typical parvala from Masatierra, C. Salsipuedes, 750 m (no. 92); it was 26 CARL CHRISTENSEN AND CARL SKOTTSBERG also observed on arid slopes in B. Chupones and on Masafuera; C. del Bar- ril, 740 m. Area of distribution: Temperate South America. Juan Fernandez. 29. B. valdiviense C. Chr. Ind. 160 (1905). —— Syn. Lomaria blechnoides Bory 1825 et auctt., Hemsl. 71; Johow 1893: 30 f. 13, 1896: 160; L. lanceolata Spr.?, Colla 1836: 45; L. 2’ Herminier?, Hemsl. 72, Johow 1893: 30, 1896: 161 (pro parte); B. arcuatum Phil. Anal Univ. 94: 352 (1896), xox Remy; B. australe var. parvulum, Johow (spec. e Masafuera). Generally confined to the forest and preferring the shadiest parts, seldom seen in more open situations. Fertile fronds develop during the summer and ripen Feb.-—April. Masatierra: C. Chifladores, slope above the innermost part of Pto Fran- ces, c. 500 m; ridge between Q. Laura and Q. Piedra agujeriada, 650 m; Pangal (JoHOw); N. slope of La Damajuana, c. 450 m; the crest separating the ©. Damajuana from V. Anson, on rocks, 150 m; along the road to Porte- zuelo, in the woods (nos. 153, 589); Q. Gutierrez (no. 49); C. Salsipuedes, Q. Helechos, 660 m, rather common and also epiphytic; Pto Ingles (joHow); Q. Juanango, very large plants (no. 611); Q. Villagra, here and there near the path; Q. de la Choza, dense forest, 250—300 m, very numerous. Masafuera: Q. de las Chozas, dry bed of a stream; Q. del Mono, dark ravine, 570 m;, Q. de las Casas, humid sheltered corners (no. 510); Q. del Varadero, dark fissures below the waterfall; O. Angosta, at the waterfall; Q. Loberia, forest patch, 280 m. In the alpine region: C. del Barril (no. 531); edge of western precipice at the foot of Las Torres, 1370 m (no. 422). A fairly constant species, varying mainly in size. The rhizome is com- monly described as creeping. This is not quite correct. It is erect from the first, but soon produces underground runners, from which new plants arise; with age these give rise to new runners with fasciculate leaves. The scales of the rhizome are entire. * — -&B. valdiviense is certainly related to B. danceolatum (R. Br.) from Australia and New Zealand but well characterized. Small plants resemble B. pexna maria, and young plants of B. Schottit may be confounded with 4. valdzvi- ense, but will be recognized by the dentate scales of the rhizome, which is long and slender, by the scattered leaves and the prominent hydathodes on their upper surface just within the margin. We have little doubt that Z. danceolata Colla and L. 7 Herminieri Hemsl. (not Bory) quoted above must be referred to 4. valdiviense. Hemsley quoted specimens collected by BERTERO and by MOSELEY: the former is beyond doubt the L. lanceolata, which according to COLLA’s few remarks seems to be ordi- nary B. valdiviense, and the latter is perhaps a large form of the same. JoHOW reprinted HEMSLEY’s remarks with the addition that he had collected it in Masatierra and that it might be united with L. attenuata (= B. Schottii). This is no wonder for JOHOW's specimens (!) are B. Schottii. The genuine B. l'Herminieri Bory, a West Indian species, does not occur in Juan Fernandez. Area of distribution: South Chile, Valdivia and Chiloé; Juan Fernandez. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 27 30. B. Schottii (Colla) C. Chr. Arkiv for Bot. 10: 7 (1910); SKOTTSBERG K. Sv. Vet. Akad. Handl. Bd. 51 no. 9 tab. 7 f. 3. Syn. Lomaria Schottii Colla 1836: 44 tab. 72; ZL. attenuata Hemsl. 71; Johow 1893: 29 f. 12, 1896: 160; L. 2’ Herminiert, Johow |. c. pro parte; L. fernandeziana Phil. Anal. Univ. 43: 105 (1873); L. della Phil. Linnaea 29: 105 (1857). Not uncommon in the forests, often climbing on trees and arboreous ferns. The fertile fronds are developed in late summer and ripen in March—Appril. Masatierra: From Pto Frances to Juanango, ranging from about 200 m to 6 or 700 m (nos. 286, 564, 1180); also on the south side, in the quebradas of Villagra. Masafuera: Q. de las Chozas, in forest 475 m (no. 517); Q. de las Casas, not rare in the canyon. In tg10 the reasons were given why this should be considered a distinct species, and the differences between it and the closely allied 4. attenuatum (Willd.) and &. meridense (KI.) C. Chr. were explained. We can add that the basal scales are more distinctly toothed and frequently concolorous in &. Schottiz, subentire and black in the centre in 4. merzdense. No. 1180 is an anomalous form with the sterile pinnae deeply pinnatifid in the external half, with subfalcate, acute segments. Area of distribution: Endemic. Pres. chilense (KIf) MettiFil)'‘Lechl! 12 14,(1856); C.*Chr. Ind. 15 — Syn. Lomaria chilensis Kaulf. Enum. 154 (1824); ZL. capenszs Hemsl. 7 Johow 1893: 28 f. 11, 1896: 159; L. Reediz Phil. Anal. Univ. 43: 581 (1873). Fairly common in the woods, preferring open glades and stony ground near running water; also in the drier tracts. Fertile fronds develop during the summer and ripen Feb.—March. Masatierra: Pto Frances, scattered in the forest, c. 500 m; C. Centinela, in the woods; Pangal, west branch on dry open ground; main gorge, at the waterfall; O. Damajuana, small waterfall, c. 250 m; V. Colonial, in all the que- bradas, common along the road to Portezuelo (no. 565, no. 331 f. anomala); C. Salsipuedes, up to 670 m; Q. Juanango, along the brook, c. 200 m; Villagra common in the large quebrada; S. slope of Co Chumacera, in the Boehmerta- grove; B. Chupones, 2—300 m. Masafuera: Q. de las Chozas, dry bed of a stream, 570 m (no. 437); Q. de las Casas, common along the course of the stream; Q. Angosta, at the waterfall; Q. del Varadero, waterfall; the Correspondencia Camp, in the ravine 1160 m; heath near Las Torres, 1350 m (dwarf specim.). 2 ae >i) B. chilense belongs to a group of closely allied species which by some authors are united under the name B. cafense (Thunb.) Schlecht. This group has never been critically revised. For the present it cannot be decided which of the numerous species described should be regarded as valid. Here we prefer to deal with B. chzlense as a separate species, although we are not prepared to give good distinctive characters. According to Jonow L. Reedi is a form with fronds partly fertile, partly sterile. Such anomalous forms are occasionally met with in other members of the group. No. 331 of the present collection is such a form, with all the pinnae 28 CARL CHRISTENSEN AND CARL SKOTTSBERG on one side fertile, and sterile on the other, corresponding to &. procerum figured by LABILLARDIERE, Nov. Holl. Pl. Spec. tab. 247. Area of distribution: Central Chile to West Patagonia, Falkland Islands; Juan Fernandez. 32. B. cycadifolium (Colla) Sturm, Enum. pl. cr. vase. chil. 23 (1858); C. Chr. Ind. 152. — Syn. Lomaria cycadifolia Colla 1836: 43 tab. 71; Johow 1896: 158; L. danuginosa Kze Anal. 19 (1837), Gay 6: 479; L. Boryana Hemsl. 2; Johow 1893: 27 f. 10; B. magellanicum var. cycadifolia C. Chr. Arkiv for Bot. 10: 9 (1910). Plate 4. Fertile in March—April. Masatierra: Characteristic of all the high ridges in the eastern and central parts, forming very dense growths above 4—500 m (no. 580) and reach- ing the summit of El Yunque (g40 m) where stout specimens may be seen with field glasses. In the dense forest it is scarce, but on open, drier mountain ridges with scattered trees it is again common. It is also encountered in the treeless basal region, from Pto Frances to Pto Ingles, descending to 30 m above sea level. On the south side of the island, small specimens were found as far west as in B. Chupones. Masafuera: On the ridges, in the alpine heath etc., smaller than in Masa- tierra buth otherwise typical; Q. del Mono, one small plant at the roadside; C. del Barril, 700-1200 m; alpine heath near the Correspondencia Camp, 1130 —1350 m (no. 405); Western precipice of C. Atravesado, well developed; slopes of Los Inocentes, scattered in the Dicksonéa-belt, important in the typical alpine heath 1350—1450 m. This, the »Cycas-fern», is certainly closely related to B. magellanicum (Desv.) Mett., to which it has been referred as a variety. From examination of dried material alone it is difficult to find good distinctive characters, but collectors apologize for its validity as a species (see Jouow |. c.). It isa true tree-fern with a stout, often slightly curved stem and very rigid, shining leaves. - Area of distribution: Endemic. 33. B. longicauda C. Chr. Arkiv for Bot. 10: 10 tab. 1 (1910). — Syn. Lomarta sp.?, Johow in sched. herb!; Asplencum longissimum, Johow 1893: 33, 1896: 163, xox Bl. Fig. 5. — Plate 5. In the quebradas, on stony or rocky soil, along running water, at the waterfalls etc., often pending from the rocks and forming fine carpets on the walls of the canyons. Not uncommon. Almost ripe fertile fronds were gathered in February. Masafuera: Q. de las Chozas, 450—500 m, dry bed of a stream (no. 365); Q. del Mono, 570 m; Q. de las Casas, in the canyon, luxuriant (no. 420); Q. del Blindado, 440 m, in the bottom of a wooded ravine; Q. Angosta, 1 km from the entrance, at the waterfall; Q. del Varadero, at the waterfall; Q. Lo- beria, waterfall in the gorge, 190 m. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 20 This highly remarkable fern was discovered by JOHOW in Q. de las Casas, and again found in the same locality by SKOTTSBERG, 1908. A _ beautiful lot NIUNTIIANTTNNR y NY ab anid ied Mann iN ANN NY Fig 5. Blechnum longicauda C. Chr. a sterile pinna, nat. size, and enlarged fragment; 4 two fertile pinnae, nat. size, and enlarged fragment; c palea form the stipe, X 5. of specimens, including fertile fronds, enable us to complete the original de- scription. The rhizome is short, more or less erect, the fronds densely fasciculate. When growing between stones along the water, the species has much the habit 30 CARL CHRISTENSEN AND CARL SKOTTSBERG of B. chilense. Rhizome and stipe are covered with thick, entire, lanceolate, castaneous scales, I,5 cm long by 3 mm broad. The upper part of the stipe and the rhachis have similar but smaller scales. Stipe short, rarely 10 cm long, angular as is also the rhachis, compressed, over */2 cm broad, deeply furrowed above. The numerous almost sessile, horizontal pinnae of the sterile frond are more or less subcordate at base with lateral lobes overlapping the rhachis. The lower pinnae are dwindling downwards, the lowest nearly semicircular. Not all the sterile leaves are proliferous, the fertile never so. Fertile fronds have very numerous, very long and narrow linear pinnae, 12—16 cm long by 2—3 mm wide only. Indusium coriaceous, nearly black, very closely imbricated and first opening at full maturity, the edges entire or faintly and repandly toothed. Lower reduced pinnae sterile and similar to those of the sterile frond, fertile pinnae in some specimens with a basal leafy obovate segment, corresponding to the basal lobes of the sterile pinnae; in others these sterile lobes are totally missing. B. longicauda differs widely from all other species except from B. Sprucez C. Chr. (Lomaria caudata Bak.) from Ecuador, to which it is closely related. We have seen the original specimens of the latter in Herb. Kew; if they are of an average size, what we do not know, this is a much smaller species. It differs mainly in the shape of the sterile pinnae, which are much shorter and suddenly contracted into a subobtuse apex, rather unlike the long narrow one in our species. The fertile pinnae are also shorter and narrower. The rhachis is perhaps more densely pubescent. Area of distribution: Endemic in Masafuera. Pellaea Link. 34. .P. chilensisFeée, Gen; 129 (1852); C. Chr. Ind. 4790. — Syan aaa cinnalis chilensis Fée, Gay 6: 447; Nothochlaena chilensis Sturm 1858; Hemsl. 77; Johow 1893: 40 f. 25, 1896: 169. Dry sunny cliffs, generally near the sea, very local. Probably fertile at any time. Masatierra: near Pto Frances (SCHREIBER t. JOHOW); Q. Piedra aguje- riada, 30—40 m (BACKSTROM); Pangal, far into the gorge, rocky wall 210 m; near the road from the colony to Pangal, on steep rocks above the sea, 45 m (no. ICO). Masafuera: GERMAIN! without locality; cliffs near the landing-place at Q. de las Casas, c. 20 m (no. 410). Very near, or too near the Andine P. mzvea (Poir.) Prantl, differing from it mainly in its sessile pinnules and generally more oblong segments. Area of distribution: Endemic. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 2 Hypolepis Bernh. 35. H. rugosula (Labill.) J. Sm. Bot. Mag. 72, Comp. 8, 1846 (»rugu- losa»). — Syn. Polypodium rugosulum Labill. Nov. Holl. Pl. Spec. 2: 92 tab. 241 (1806); Dryopterts punctata subsp. rugosula C, Chr. Ind. 287. var. Poeppigii (Kze). —- Syn. Polypodium Poeppigit Kunze, Linnaea g: 50 (1834); Phegopteris Poeppigiu Fée, Gay 6: 506; Mett. Pheg. no. 13; Polypodium rugulosum KIf. Enum. 122 et auctt.; P. punctatum Hemsl. 76 (excl. syn. Phego- pleris spectabilis); Johow 1893: 36 pro parte (zon fig. 21), 1896: 166 pro parte; Nephrodium villosum Johow 1893: 35 pro parte, fig. 19, 1896: 165 (t. spec. in Herb JoHOW! xox auctt.); Hypolepis Poeppigiana Mett. Fil. Lechl. 1: 18 (1856); Dryopteris punctata C. Chr. Ind. 555. Seems to thrive under very different conditions, in dense or open forests, on barren slopes at the roadsides etc. Probably fertile at all seasons. Masatierra: Q. Piedra agujeriada, c. 650 m (no. 596); V. Colonial, barren slopes (no. 120) and in the caves no. 1V and V (common, no. 587); Q. Vil- lagra, just below the Portezuelo pass (no. 276). Masafuera: Near the abandoned Chozas village, roadside, 400 m (no. 439); Q. del Blindado, forest, c. 440 m; C. del Barril, 985 m;Q. de las Vacas; Q. Angosta, at the waterfall. The many specimens collected clearly belong to the Chilean Pol. Poep- pigzz, and this cannot be separated from H/. rugosula as a distinct species. KUNZE found that the essential differences between his Poeppigii and rugosu- lum of LABILLARDIERE (the specific name erroneously written »rugulosum» by all authors) were the lax texture and the obtuse, approximate secondary pinnules of the former. It is true that most specimens show these characters, which, however, vary with difference in shade and humidity. The shade form (e. g. no. 587) belongs to typical Poepfpzgz7, but specimens from open soil are scarcely distinguishable from typical P. rugosulum from Australia, where the texture in the less humid climate becomes more thick, and the pinnules are more distant and more acuminate. In all other characters specimens from the two regions closely agree, and it is, therefore, impossible to deal with Poep- pigii as a species, perhaps even as a distinct variety. The specimens from Juan Fernandez are also very similar to Polypodium visctdum Roxb. (= Hypo- lepis helenensis Fée) from St. Helena, and there is little doubt that P. w2//oso- viscidum Thouars from Tristan d’Acunha is the same. All these names are quoted as synonyms of P. rugosulum Labill.; the name of THOUARS was pub- lished in 1804 and consequently has priority, but having seen no specimens, we shall not venture to create a new combination. H{. rugosula is frequently referred, as a variety or subspecies, to Po/. punctatum Yhunb., the type of which is from Japan. A comparative study of specimens from different regions will show, we believe, that 77. rugosu/la is a distinct species of more southern distribution. Both certainly belong to //yfo- lepis, not to Dryopteris. In 1856 METTENIUS described some specimens from Chile, previously referred to Pol. Poeppigit, as Hypolepis Poeppigiana, while he retained the spe- 32 CARL CHRISTENSEN AND CARL SKOTTSBERG cies of KUNZE under Phegofteris. Having examined specimens determined by METTENIUS himself we must confess that we do not understand how this keen observer has arrived at such an unnatural conclusion. The only difference between the two species accepted by METTENIUS (and under two different genera!) is that //. Poefpigiana has the fertile lobe somewhat protruding and finally recurved, covering the sori, while, in P. Poeppig7, the fertile and sterile lobes are alike, the fertile ones recurved with age. A study of more compre- hensive material will show whether this difference is sufficient for the maintaining of two species; as a character of genera it is absolutely artificial. In this place it is sufficient to state that all specimens from Juan Fernandez belong to PA. Poeppigit of METTENIUS. The present species was greatly misunderstood by JoHow, which is proved by his figures and remarks as well as by his specimens. His Polypodium punc- fatum is for the greater part Dyyopterts inaequalfolia, while Hypolepis was figured as Nefhrodium villosum; but his remarks on the latter clearly refer not to Fypoleprs, by but to D. maegualifolia. While this has an erect, some- times subarborescent caudex, //yfolepis has a widely creeping, slender rhizome, which character alone, not to speak of the numerous others, will at one serve to distinguish the two plants. The specimens collected look rather different, diverging in size, cutting, and texture. The largest ones, which could be named /7. rugosula typica, have a firm, broadly ovate lamina, 35 cm long by 25 wide, fully tripinnate-quadri- pinnatifid, the tertiary pinnules being deeply pinnatifid. Others, corresponding to Nephrodium villosum of JOHOW, are of the same length and texture, but with a narrower lamina (12—16 cm), the basal pinnae considerably reduced, bi- pinnate-tripinnatifid, with obtuse secondary pinnules. This form corresponds very well to Pol. Poeppigitt of KUNZE. There is a thinner shade-form of the same. Finally we have the fine form growing in the caves (no. 587). This has a thin, bright green lamina, lanceolate in outline (25 by 10 cm), barely pinnatifid with lobed secondary pinnulae. A specimen from Masafuera is smaller still (no. 439) the fully fertile leaves, of a rather firm texture, are only 12 by 4 cm, lanceolate, bipinnatifid. Such specimens approach very much /#. obtusata (Pr.) Kuhn (syn. Polypodium fulvescens Hook et Grev., t. spec. orig, in Herb. Kew), a high Andine species, referred to A. vugosula by HOOKER, Spalkilia::272. Area of distribution: Taken in a wider sense, the species inhabits Australia, Chile, Juan Fernandez and probably also Tristan d’Acunha. Adiantum L. 36. A. chilense’ KIf. Enum. 207 (1824); C. Chr. Ind. 24; Arkiv for Bot. 10: 4 (1910). — Syn. A. aethiopicum, Hemsl. 70; Johow 1893: 23 f. 5 a—e, 1896: 156 zon L. A very common plant, ranging from the sea coast to the highest ridges, found both in the forest region and on the open, barren slopes near the sea. With ripe sporangia in the autumn and winter. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 33 Masatierra: In all the valleys from Pto Frances to Juanango (nos. 59, 59 b, 572). Santa Clara: JOHOW. Masafuera: Common along the shore from B. Toltén round the E. and S. coasts to Q. Loberia, especially under overhanging rocks (no. 446); also in the forest (no. 518) and in the alpine heath, to 1400 m. All specimens examined belong to the genuine A. chz/ense, characterized by perfectly glabrous fronds. JOHOW described and figured three forms. Of these, the first (fig. 5 a) is a robust one, with a strong stipe, and the third a smaller one (fig. 5 c) with a slender stipe; surely they represent two local forms differing from humidity and shade. Both are represented in our collection. The second variety of JoHOwW (fig. 5 b) is remarkable by the pinnae being more or less striped with white, forming white sections of tissue without chloro- phyll. Suggesting that they were due to some pathologic cause, Dr. H. E. PETERSEN on our request made microtome sections: through such leaves and found that the cells of the white parenchyma were filled with an entangled mass of hyphae of a parasitic fungus that destroys the chlorophyll. Area of distribution: Peri to Patagonia, Falkland Islands; Juan Fer- nandez. Pteris L. 37. P. chilensis Desv. Berl. Mag. 5: 325 (1811); Hook. Sp. Fil. 2: 176 tab. 120 A; Hemsl. 70; C. Chr. Ind. 594; zon Johow 1893 fig. 7 nec 1806: 156, saltem pro parte. — Syn. P. éenera KIf. 1824; Gay 6: 488 (xox Bertero); P. fernandestana Phil. Linnaea 29: 105 (1857), t. spec. orig. Herb. Mus. Nac. Santiago! In humid forests, on soil rich in humus; not observed at a greater altitude than 500 m in Masatierra. (Fertile Dec.—April.) Masatierra, not uncommon in the central part: C. Centinela, c. 400 m (no. 609); V. Anson, lower slopes of La Damajuana, 345 m (no. 216); V. Co- lonial, at the road to Portezuelo, in the maqui-thickets, 200 m (no. 566); Q. del Monte Maderugo, c. 500 m (no. 337); C. Salsipuedes, forest patch c. 400 m; Pto Ingles, on the central ridge (no. 324); south side of the island, slope of Co Chumacera (no. 252). Masafuera: in the forest of the Sanchez plain, 515 m (no. 401); Q. de las Chozas, in the forest (no. 544); Q. de las Casas, under a large stone in the outer section of the gorge, rare. The present species is nearly related to the South Brazilian P. /eptophylla Sw., but still very distinct. Its rhizome is short creeping, rather stout, densely covered with dark brown, rigid, lanceolate scales, as is the larger part of the stipes, which rise close together. P. fernandeszana Phil. cannot stand even as a variety; the type of P. chzlenszs is quadripinnatifid, but some specimens, growing in deep shade, are barely tripinnate at base, and the lamina more open and thin, and they represent the true P. fernandeztana. JoHow’s P. chilensis, as proved by his figures, description and specimens, 3— 20208. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. I, 34 CARL CHRISTENSEN AND CARL SKOTTSBERG is = Histiopteris. Thus it is impossible to understand why he identified P. Sernandesiana with his chilenszs. Area of distribution: South Chile, Prov. Maule to Valdivia; Juan Fer- nandez. 38. P. semiadnata Phil. Linnaea 29: 106 (1857); C. Chr. Ind. 607. — Syn. P. flexuosa Mett. Fil. Lechl. 1: 12 (1856) on KIf. 1830; P. marattiaefolia Hook. Sp. Fil. 2:.177 tab. 122 B (1858); Hemsl. 71; Johow 1893: 25, 1896: 197. Characteristic of the foggy region above 500 m, in Masafuera also in the alpine heath, very local. (Fertile Jan.—Appril.) Masatierra: Densely wooded, damp quebradas on the highest part of C. Centinela, c. 660 m (no. 613); C. Salsipuedes, Q. Helechos in Drimys- Dicksonia-forest, c 600 m, a beautiful stand (no. 280). — GERMAIN without locality. *Masafuera: Q. del Mono, c. 400 m (no. 543), forest ravine; Q. del Blin- dado, forest, 440 m; C. del Barril, rock crevices above 800 m; Q. Inocentes, under Dcksonia near the stream; the Correspondencia Camp, in the ravine 1160 m; edge of western precipice, at the foot of Las Torres among rocks, 1370 m (no. 438). — New for Masafuera. The occurrence in the islands of this pretty, distinct species, recorded by HEMSLEY from specimens collected by GERMAIN, was doubted by JoHow, who suspected the locality to be false. The question is now settled by the redisco- very of the species. The specimens are identical with those previously collected by SKOTTSBERG in the island of Huafo, perhaps with the sterile tips of the pinnules more grossly serrate, at least much more deeply and broadly serrate than shown in HOOKER’s figure. The fertile edges are regularly crenate, often prolonged beyond the brownish green, continuous indusium, which generally is crenate with the crenae covering the nearly globose, rarely confluent sori, that arise from the tips of the veins. In the sori the species resembles Casse- beera, but the whole habit is that of Pterzs. It is a very large plant. The base of the stipe is clothed with dark brown, opaque, lanceolate and entire scales. We do not know any species of Eupferis nearly related to P. semiadnata. We cannot agree with METTENIUS in considering it »ex affinitate P. chilensis», but it resembles some large Andine species of Lztobrochia, especially the Mexican P. fulchra Schlecht. We are of the opinion that the subgenera of Pteris, based solely upon the venation, are very unnatural, and that it is quite natural to associate the free-veined P. semzadnata with P. pulchra and its allies, which have anastomosing veins. Area of distribution: South Chile, Valdivia to Isla Huafo; Juan Fer- nandez. 39. P. Berteroana Agardh, Rec. Gen. Pteridis 66 (1839); C. Chr. Arkiv for Bot. 10: 20 (1910). — Syn. P. éenera Bertero ex Colla 1836: 38 (t. Ag. xon Kaulf.); Letobrochia decurrens Gay 6: 491 (ex. descr., xox Presl); Pterts comans Hemsl. 70; Johow 1893: 25 f. 8, 1896: 257; P. zuczsa Johow 1893: 24 f. 6 (¢t. spec. orig.!). THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS uw 3 Common in the forest belt from Pto Frances to Juanango, and ranging vertically from 200 to 640 m or more, in dense groves or in open glades along streams and cascades. (lertile Dec.—March.) Masatierra: C. Chifladores, quebrada in the interior of Pto Frances, c. 500 m, not uncommon; Q. Piedra agujeriada, c. 650 m; Pangal; Q. de la Da- majuana, frequent at the small waterfalls, 250—350 m (no. 64, 217); near Pla- zoleta del Yunque, ca. 250 m (no. 64B); V. Colonial, Q. del Monte Maderugo, 235 m; Cave no. V, rare; C. Salsipuedes, fern grove, 600 m; Q. Villagra, below the pass, c..500 m, scarce; Q. de la Choza, in forest c. 250 m; S. slope of Co Chumacera, in a Loehmeria-patch. Masafuera: Forest on the Sanchez plain, 515 m; fissures in the coast cliffs N. of Q. de las Casas, rare; more common in the canyon (no. 464); Q. del Blindado, in the forest, 440 m; Q. de las Vacas, in the gorge; Q. Loberia, at the stream and in the forest, c. 290 m. This species cannot be united with P. comans Forst., but it is undoubtedly closely related to P. Lndlicheriana Ag. from Norfolk Island, with which the older P. Zahlbrucknertana Yndl. is probably identical. We know this species from HOOKER’s Icon. Plant. tab. 973 only, and the specimens from Juan Fer- nandez are different in habit. The venation is as described by AGARDH; above the row of regular costal areoles the veins are free or form an irregular second row of smaller areoles, while, in P. Axdlicheriana, there are four rows of areoles. The ultimate segments are also different in shape, longer and more lanceolate in P. Lerteroana. In any case, this belongs to the Polynesian element in the flora. Area of distribution: Endemic. Histiopteris (Ag.) J. Sm. foo. incisa (Uhbg), J. Sm... Hist: Fil, 295; C. Chr. Ind: 352..— Syn. Pieris incisa Thbg 1800; Hemsl. 70; Johow 1893: 24 (pro parte, xox f. 6), 1896: 156 pro parte?; P. flavescens Colla 1836: 37 tab. 67; P. patens Kze Anal. 28 (1837); Letobrochia patens Gay 6: 490; P. vespertilion’s Lab. var. flavescens Ag. Rec. 80; //. zczsa var. flavescens C. Chr. Arkiv for Bot. 10: 14 (1910); P. che- lensis Johow 1893: 24 f. 7, 1896: 156 (excl. syn.). A characteristic fern of the open forests and clearings, also occurring in the barren tracts and forming extensive mats in burnt places. (Fertile Jan.— April.) Masatierra: Ridge between the Laura and Piedra valleys, 500—600 m, along the edge of the forest; Pangal, in the dry outer sections of the gorges; clearing near Plazoleta del Yunque, copious; V. Colonial, lower slopes, in the small ravines (nos. 187, 583); QO. del Monte Maderugo (no. 333); Caves nos. I and X; Villagra, along the road from Portezuelo, very numerous in cleared spaces; S. slope of Co Chumacera; B. Chupones, rocky places, ravines, scarce. Masafuera: Q. del Mono, c. 475 m (no. 519); Q. del Blindado, 440 m; Q. Loberfa, near the stream, c. Ig0 m; highlands north of Q. Casas, seaward 36 CARL CHRISTENSEN AND CARL SKOTTSBERG face of Las Torres, among rocks, 1370 m; slopes of Los Inocentes, along the edge of the Dzcksonza-forest, 950—1200 m; on the summit, 1350—1400 m. A detailed comparison of our specimens with others from different regions has proved that the plant from Juan Fernandez cannot be distinguished from the South African type or from the Australian P. vespertilionts Labill. even as a variety. GAY distinguished two species, but it seems that his descriptions are abbreviated reprints from older works and not written after autopsy of specimens. AGARDH referred P. flavescens to P. vespertilionis WLabill. as a variety, with the following differential characters: »fronde subtus glaucescente, stipite flavo» but specimens from Tasmania are also glaucous and the stipe of P. flavescens is not always yellow but frequently light castaneous and very glossy; nor do the differences between this and fatens or zuczsa hold good. The species varies in colour of stipe and rhachis, in the degree of cutting, bipinnate-tripinnatifid to tripinnate-quadripinnatifid, the more divided form well figured by CoLLA. The lower side is pale green or subglaucous, the sori reach- ing from the base to near the obtuse tips of the segments or interrupted. The veins generally form a single row of costal areoles, but in one specimen we found them nearly free. As all these variations are also found in the South African or Australian plant we conclude that it is the typical form of P. zuczsa that has found its way to this isolated locality. The South American variety (P. elegans Sw.) looks rather different, and the Polynesian form (ZL. szuuata Brack.) is still more remotely related to it. In a single character some speci- mens recall ZL. szzuata, for the lower side is finely glandulose, quite as in spe- cimens from New Caledonia, but this character is not a constant one, and, moreover, some few glands may be found also in typical zuczsa. JoHow has misunderstood the present species, as well as the species of Pterts, which is proved from his figures and also from the specimens in his herbarium. Litobrochia appendiculata Gay 6: 491 war referred to P. zvczsa by HEMSLEY, and JOHOW states that a specimen so named in the Herb. Mus. Nac. Santiago no doubt belongs to P. chilensis. Now, P. chilensis Joh. is Histiopteris. We believe that this opinion is right. Area of distribution: South Africa, Australia, Tasmania; Juan Fernandez. Polypodium L. #41. P. Billardieri (Willd.) C. Chr. Ind. 513. var. magellanicum (Desv.) C. Chr. Arkiv for Bot. 10: 15 (1910). — Syn. Grammutis magellanica Desv. Berl. Mag. 5: 313 (1811); G. ~ana Brack. U. S. Expl. Exp. 16: 1 (1854). Rare. Fertile fronds occur at all seasons. Masatierra, only in the humid elevated parts and only found growing on Drimys Winteri: high ridge between Q. Laura and Q. Piedra agujeriada, in dense forest 650 m (no. 597); on the main crest of the island, above Pangal, c. 800 m; mountain spur W. of el Yunque, c. 500 m, very scarce (no. 632). Masafuera: C. del Barril, in moss-carpets on rocks, 985 m (no. 537). THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 37 The specimens of this species, for the first time collected in Juan Fer- nandez, agree perfectly well with the small form from Fuegia described as G. mana Brack., which is nothing but a dwarf form of G. magellanica Desy. In 1910 (1. c.) were pointed out the differences between the South American form and the typical G. Lellardiert (Pol. australe R. Br.) from Australia, New Zea land etc., and it is a matter of convenience whether the two forms ought to be regarded as two separate species or as forms of a single one. We prefer the latter as the American plant is intimately related to the typical form, while, on the other hand, it has no near relative in South America. In any case it is an Antarctic element in the South American flora. Area of distribution: Australia, Tasmania, New Zealand, Lord Howe’s Island, Norfolk Island, Auckland and Campbell’s Islands. South Chile; Valdivia to Fuegia; Juan Fernandez (var. magellanicum), P. intermedium Colla 1836: 51 tab. 74. — Syn. P. tvanslucens Kunze, Anal. pteridol. 16 (1837); Hemsl. 76 tab. 13; Johow 1893: 39 f. 23, 1896: 166 tab. 18; Gonzophlebium translucens Fée, Gay 6: 500. COLLA’s name must be preferred to KUNZE's, being one year older; it is not invalidated by an earlier P. ¢ztermedium Willd. as erroneously stated in Ind. Fil. 570. In Willd. Spec. 5: 262 P. zztermedium Mihlb. is quoted as a M.S. name only for Aspzdzum tntermedium Willd. This species is endemic, although GAy etc. also quoted Chile as its native country, for the specimens in Chilean herbaria belong to P. syxammza (Fée). As mentioned by HEMSLEY and JOHOow it is very variable. The collection includes a series of varieties, of which three correspond to those figured by HEMSLEY. The extreme forms are so different that they can be mistaken for separate species, but they all have the essential characters in common, and a similar degree of variation is known in several species, e. g. in P. vulgare, the forms of which correspond in a remarkable way to those of the present species. All forms have the same thick rhizome, densely covered with imbricate, peltate, ovate-lanceolate scales, which are dark reddish brown and thicker in the centre, paler and thinner at the margins and apex. The stipe is rather pale, glabrous, and distinctly articulated just above the base. Lamina deltoid, the pinnae connected by a narrow wing to the pale rhachis, the basal ones a little decurrent; surface glabrous, texture generally thin, margins hyaline or cartilagineous, veins very visible in their lower course, where they are dark- coloured, but becoming nearly invisible towards their tips. Costal areoles di- stinct, outside each areole 2—4 free veins, that do not reach the margin; for exceptions, see below. The areole includes a free veinlet, which bears a sorus at its apex; sometimes it is furcate. The receptacles are distinctiy linear but short, the young sori oblong, a little immersed, the mature globose and superficial. ; All forms show these characters: they differe in size, texture and degree of division. Smallest leaves are 6 cm long and wide (stipe excluded), largest 24 cm long and wide; sometimes the lamina is broader than long. Typically it is 38 CARL CHRISTENSEN AND CARL SKOTTSBERG thinly herbaceous, but specimens growing on rocks (nos. 318, 1148) have a nearly coriaceous leaf with thick cartilagineous edges and the sori more di- stinctely immersed. As stated above the species varies extraordinarily in the degree of cutting. The type, first described under the name iztermedium (or translucens), is con- fined to Masatierra. Only once, an aberrant form was found growing on rocks (no. 318); this corresponds rather to var. daszcompositum from Masafuera. When JOHOW states that in Masatierra the species shows »lamina pinada-tripinada», he must have drawn his conclusions from HEMSLEy’s plate, for he never col- lected a much divided form in Masatierra. The plants figured by HEMSLEY fig. 2, 3 were collected by Mr. DOWNTON in »Juan Fernandez». They represent two Masafueran varieties and were undoubtedly collected in Masafuera, this island also having been visited by DOWNTON, and not in Masatierra, where they do not occur. Masafuera is inhabited by a number of forms connected by transitions. As they grow side by side, sometimes two widely distant ones on the same branch, it is impossible to regard them as merely results of local influences. Either, the species is composed of a number of units, elementary species, giving the impression of a more or less continuous series, or we have a few subspecies and their hybridogenous offspring. From taxonomic point of view we have found it advisable to distinguish the following phaenotypes. I. typicuim. Endemic in Masatierra, not uncommon as an epiphyte on trees, generally high up above the ground. Fertile during the summer and autumn. Masatierra: generally found on Drimus Winter, but rarely below 400 m, e. g. on the ridge between Q. Laura and Q. Piedra agujeriada, c. 650 m; Rabanal, c. 450 m; Q. Damajuana, c. 250 m (rare), above 460 m fairly com- mon; NE. slope of El Yunque (no. 154); C. Central, above 400 m; V. Colonial, the slope below the Portezuelo pass (no. 33); Q. Seca, c. 435 m; C. Salsipuedes, 670 m, and in Q. Helechos, 660 m; Pto Ingles (JoHOW). Confined to Masatierra and well illustrated by HEMSLEY, tab. 63 fig. 1. Simply pinnate with pinnae connected by a narrow wing to the rhachis, their margins subentire or obscurely serrulate. Veins always forming regular costal areoles, often dark. In general habit resembling large forms of the Mediter- ranean P. vulgare subsp. serratum. An aberrant rock form corresponds to var. basicompositum described below: minor, coriacea, pinnis basalibus pinnatis. —- Fig. 6 d. Masatierra: Pto Ingles, two small specimens grew on the outside of the »Robinson grotto» (no. 318). We are not sure if this form is to be considered as a variety of ¢yfzcum, growing on rocks, but we find it more probably that it ought to be referred to var. bastcompositum of the Masafueran subspecies; in Masafuera, also another variety was found growing on rocks. It very likely offers an example of occasional dispersal of spores from Masafuera to Masa- tierra, where no forms of subsp. masafueranum occur. 2. subsp. masafueranum C, Chr. et Skottsb. nov. subsp. Fig. 6. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 39 Fig. 6. Polyfodium intermedium Colla subsp. masafueranum C. Chr. et Skottsb, a@ var. obtu- seserratum (no. 431); & var. basicompositum forma, with regular areoles (no. 1181); ¢ var. dast- compositum forma, with free veins (no. 1186); d form from Masatierra, corresponding to dasv- compositum (no. 318); e var. cambricoides (no. 432) with areoles or free veins. — All plants half nat. size. 40 CARL CHRISTENSEN AND CARL SKOTTSBERG A typo differt pinnis basi magis angustatis, margine distincte dentato- serratis vel grosse obtuse-serratis, interdum pinnis plus minus pinnatifidis, basa- libus compositis, rhachide angustissime alata. Fig. 6. “Endemic in Masafuera. All forms seem to differ from the type by their pinnae being more narrowed towards the base, and by the nearly unwinged rhachis. We distinguish the following varieties. All of them are epiphytic on trees, mostly on Myrceugenta Schulze, only once a form has been observed growing on rocks. All forms were seen with ripe sporangia (Feb.—March). var. obtuseserratum C. Chr. et Skottsb. nov. var. Frons plerumque haud magis quam in typo divisa, sed pinnae multo distinctius serratae vel lobulatae, serraturis vel lobis sat irregulariter dentatis; pinnae basales breviores, rarius basipinnatae. — Fig. 6 a; HEMSLEY Fig. 2. Masafuera: Forest on the Sanchez plain, 515 m, growing in company with the following forms (no. 527). Q. del Mono, 450—570 m (no. 1182); Q. del Blindado; Q. Loberia, 280 m, in forest (no. 431) and below the forest, in the gorge, on rocks (f. corzacea, no. 1148). Generally not more divided than the type, but the pinnae, often consi- derably narrowed towards the base, are much more deeply serrate, the ser- ratures again irregularly dentate, or cut more than halfway down to the midrib into oblique, repand or serrate lobes, the basal ones shorter or obsolete; some- times the basal pair of pinnae produces one or two longer lobes on the lower side, or, more rarely, on the upper side, thus forming a transition to the follow- ing variety. Veins of the lobes often free but not seldom forming one or two areoles, var. basicompositum C. Chr. et Skottsb. nov. var. Pinnae basales regulariter et profunde pinnatifidae segmentis elongatis; pinnae superiores ut in praecedenti vel interdum suprabasales segmentis 2—3 longis instructae. Fig. 6 0, c; HEMSLEY Fig. 3. Masafuera: Forest on the Sanchez plain, 515 m (no. 1181); Q. del Mono, 570 m (no. 1186); Q. de las Casas, on a tree fallen down from the wall of the canyon (no. 529). Basal pinnae regularly and deeply pinnatifid with lengthened segments on both sides; upper pinnae as in the preceding form, but sometimes the second pair of pinnae produces two or three long segments. The basal upper segment of basal pinnae short or often quite rudimentary. var. cambricoides C. Chr. et Skottsb. nov. spec. Plerumque maxima, tripinnatifida. Frons flaccida textura tenuis. Pinnae usque ad costam alatam incisae, pinnulis basalibus superioribus ad dentem axillarem reductis. Pinnulae pinnarum superiorum profunde et distanter obtuse- serratae, illae inferiorum pinnatifidae segmentis tertiis repandis vel obtuse serratis. Te) 02. THE PTERICOPHYTA OF THE JUAN FERNANDEZ ISLANDS 41 Masafuera: Forest on the Sanchez plain, 515 m (no. 540); Q. del Mono, 450—570 m (no. 432); Q. del Blindado. Generally large, bipinnatifid. All pinnae incised to a narrow wing into long secondary pinnules, the upper basal ones being reduced to a tooth in the axil of the costa. Pinnules of upper pinnae deeply and distinctly serrate, resembling the upper pinnae of var. odtuseserratum, those of the lower pinnae again pinnatifid with repand or obtusely serrate tertiary segments. Veins in ultimate lobes free or also forming a single areole, rarely dark. Texture mostly very thin, the lamina flaccid. This form corresponds to the var. cambricum of P. vulgare, although somewhat more divided. Area of distribution of P. zxtermedium: Endemic. 43. P. pycnocarpum C. Chr. Ind. 557. — Syn. P. macrocarpum Presl, Rel. Haenck. 1: 23 tab. 1 f. 24 (1825), xox Bory; Johow 1896: 167. P. masa- Juerae Phil. Linnaea 29: 107 (1857); Hemsl. 77; Johow 1893: 38; P. sguamatum Phil., P. atacamense Bak. Masafuera: GERMAIN! Only a single, badly dried specimen is known, now in the Mus. Nac. in Santiago; it was collected in Masafuera by GERMAIN and was described by PHILIPPI as a new species. The impossibility of refinding it made it rather improbable that it should be an endemic species. After an examina- tion of the original plant we have arrived at the conclusion that JOHOW was right in naming it P. macrocarpum, an Andine species. The single plant found perhaps grew from a spore brought from the Andes with a strong northerly wind, just as the curious occurrence of Elaphoglossum, mentioned below, may be explained. JOHOW suggested that P. atacamense Bak. (= squa- matum Phil.) from the Atacama desert would be a more xerophilous form of the same species; thanks to Dr. O. STAPF, who sent us a leaf of the type, we are able to confirm JOHOW’s opinion. It is not nearly related to P. ple- bejum Ch. et Schl., as thought HEMSLEY. Area of distribution: Mexico to Chile and Argentina; Masafuera. 44. P. lanceolatum L. Spec. Pl. 2: 1082 (1753); Hemsl. 76; Johow 1893: 36; C. Chr. Ind. 537. — Syn. P. peltatum? Colla 1836: 52 non Cav.; Drynaria elongata Fée, Gay 6: 512; Gymnogramme elongata (Hemsl. 77?), Johow 1893: 38 f. 24, 1896: 168. A rather common epiphyte, also in the thin woods, generally found on Drimys Waintert and often in company with P. ntermedium; rarely terrestrial. (Fertile Dec.—April.) Masatierra: Pto Frances, c. 250 m (no. 129); ridge between Q. Laura and Q. Piedra agujeriada, c. 650 m; Rabanal, frequent in the interior; C. Cen- tinela, on the ground in a shallow depression of the ridge not far south of the Radio station; higher up a common epiphyte (no. 1150); Pangal (JOHOW); QO. de la Damajuana, c. 250 m (no. 62); V. Anson, on the slopes of La Damajuana (no. 215) and of El Yunque, common above 400 m; V. Colonial, along the road to Portezuelo (no. 32); Q. Seca, c. 435 m; C. Salsipuedes, Q. Helechos, 660 m. Masafuera: Forest on the Sanchez plain, 515 m; Q. del Mono, 475 m 42 CARL CHRISTENSEN AND CARL SKOTTSBERG (no. 1183); Q. de las Casas, growing on the rocky wall not far from the en: trance; and also found in the canyon on a fallen tree (no. 546). All the numerous specimens collected belong to the genuine P. /anceola- tum and no form seen can with any right be placed under Gymnogramme elongata (Sw.) Hook. HEMSLEY says that the subspecies had been collected once but that he had not seen specimens. JOHOW had two species in 1893; in 1896 he merges P. lanceolatum into Gr. elongata, evidently unaware of the differences between the two types. It seems certain that only one form occurs in the islands, and, therefore, G. e/ongata must be excluded form the list. Some of the specimens are very large (leaves 25 cm long by 2?/2 cm wide) and several have the edges of the upper part sinuate. No. 32 includes several frond with the tips once or twice furcate or cleft into three lobes. Area of distribution: Pantropical; also in Chile. Elaphoglossum Schott *45. E. Lindenii (Bory) Moore. C. Chr. Ind. 310. —- Syn. Acrostichum Lindentt Bory; Fée Acrost. 48 tab. 18 f. 3. Masatierra: north precipice of Co Piramide, c. 600 m, among mosses on a narrow ledge; very few specimens seen (no. 193). A most interesting addition to the flora of the islands. The few plants found agree in all details with the Mexican’ form described as Acrostichum ve- nustum Liebm. A possible explanation of the occurrence of this species in the islands is given under Polypodium pycnocarpum above; the mountain wall where it was found faces the North and will be struck by the rare but strong north- erly storms. Area of distribution: Tropical America; Masatierra. Not in Chile. Gleicheniaceae. Gleichenia Smith While HEMSLEY listed two species of Gleichenia, G. cryptocarpa Hook. and G. revoluta H. B. K., evidently overlooking that HOOKER in Syn. Fil. 13 quoted the locality »Juan Fernandez, Bertero», for G. pedalis {KIf.) Spr., JoHOW’s list contains only G. fedalis, as he believed that both species listed by HEMs- LEY were incorrectly named and only represented G. pedalis. In this he is quite right: G. cryptocarfa, quoted on the authority of PHILIPPI (Bot. Zeit. . 1856: 631) is shown by the specimens in Mus. Nac. Santiago to be G. pedalis (they were thus labelled by PHILIPPI himself), and Mr. WRiGTH kindly informs us that DOWNTON’s specimen of G. revoluta in the Kew Herbarium is G. feda- “is and that MOSELEY’s fragmentary plant also belongs to the same species. Thus, only one species of the genus was hitherto known from the islands. THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 43 *40. G. quadripartita (Poir.) Moore, Ind. 382 (1862); C. Chr. Ind. 324. Only in the Alpine re March.) - Masafuera, in the highest parts: The Correspondencia Camp, in the ravine, 1160 m (no. 409); Las Torres, 1370 m; C. del Barril, 1290 m; Los Inocentes, 1375 m. — New for Juan Fernandez. gion, in sheltered places. Rare. (Fertile Feb. z The specimens agree closely with an authentical specimen leg. COMMER- SON, Herb. Haun. The primary branches are sometimes once, sometimes twice furcate. Younger leaves decidedly glandulose, especially beneath. This species is often confounded with G. cryptocarpa Hook., but it is not proliferous although having a scaly bud in the primary furcation; the costae and costules are rather densely chaffy with rufous scales and flaccid hairs, the edges revolute but not covering the sporangia, and the slender, branched rhi- zome is very sparingly paleaceous with ovate, peltate, appressed, very deciduous scales, while the rhizome of G. crypfocarpa is stronger and more densely and permanently paleaceous. Area of distribution: Llanquihue to Fuegia; Masafuera. 47. G. pedalis (Kaulf.) Spr. Syst. 4: 26 (1827); Johow 1893: 20 f. 1, 1896: 153. — Syn. Mertensia pedalis Kaulf. Enum. 39 (1824). Rocky ridges in dry, sunny places, very local. (Fertile in March.) Masatierra: V. Colonial, near the road to Portezuelo, 3—400 m, forming small beds in places (no. 563); C. Central, on the narrow crest, 570 m; C. Sal- sipuedes, narrow ridge between Q. Salsipuedes and Q. Helechos, c. 500 m, Q. Villagra, low ridge near the camping place, c. 200 m. The specimens are rather typical, still the costa and costules are per- fectly destitute of scales; both surfaces are finely glandulose by minute, glist- ening glands. Area of distribution: Cordilleras of Southern Chile; maritime provinces from Valdivia to the Chonos Islands; Masatierra. : eo Gr, cl. litoralis: (Phil) © Chr. lnd..322) Masafuera: The Correspondencia Camp, in the ravine, very scarce (no. 403.) Sterile specimens from Masafuera that resemble G. peda/7s in colour, veins, structure of scales and growth, but differ in its palaceous costae and costules; the scale are rufous, paler with age, ciliate, those of the costules lacerate, small, all leaving small red spots when falling; it also differs in the shape of the segments, which are broader below, gradually tapering toward the apex and finally in the primary branches being twice furcate, carrying segments above quite to the base, naked outwards from the first to the second furcation. Typical G. pedalis, as a rule, has the primary branches prolonged without seg- ments on both sides save for a larger one close to the furcation. After inspection, Prof. HIERONYMUS informs as that it stands near /zforalis of Phi- lippi, which has narrower scales on the midrib. 44 CARL CHRISTENSEN AND CARL SKOTTSBERG Ophioglossaceae. Ophioglossum L. . *ag. O. fernandezianum C. Chr. nov. spec. Rhizoma parvum, 3—4 mm latum, hypogaeum, radiculas numerosas brun- neas emittens. Folia saepe bina. Petiolus communis hypogaeus, tenuis, palli- dus, compressus, rarius ad 2 cm longus. Lamina sterilis petiolo hypogaeo O,5—I,s5 cm longo, pallido, complanato, ovata, 1,5—2,; cm longa, paulo supra basin rotundatam 10—14 mm lata, apiculata, carnosula. Vena mediana in- Fig. 7. Ophioglossum fernandezianum C. Chr. a, 6 two plants, x 2; ... marks the surface of soil; ¢ leat, X 4, showing venation. distincta, quam laterales vix crassior, interdum subnulla; areolae magnae. Vitta deest. Pedunculus spicae lamina sterili brevior vel subaequans, fere ad apicem hypogaeus et pallidus. Sporangia 7—10-juga, apice sterili spicae ovato-acuto. Masatierra: In the western part, slopes towards B. Tierras Blancas, c. 50 m (BACKSTROM; no. 1126). — Fertile. In the summer, no trace of this plant was seen; in the spring, when Mr. BACKSTROM returned to Masatierra, he found this species (Aug., 1917). This new little species certainly falls within the aggregation of forms called O. mudicaule by BAKER, but it cannot be referred to any species de- scribed by PRANTL in his monograph of the genus. It mostly resembles O. ypanense Mart. from Brazil and Columbia, but it lacks the »vitta», the pale ‘ THE PTERIDOPHYTA OF THE JUAN FERNANDEZ ISLANDS 45 stripe along the middle of the sterile lamina. [From all species known to us it seems very well marked by the long shaft of the sterile lamina and by the short peduncle, which scarcely and rarely overtops the sterile lamina. The whole common stalk as well as the stipe of the sterile lamina and nearly the whole peduncle of the spike are, at least in all specimens collected, hypogaeus and without chlorophyll. The venation is seen in fig. 7 c, drawn by Mag. MATTHIESEN. Area of distribution: Endemic in Masatierra, Lycopodiaceae. Lycopodium L. *so. L. magellanicum Sw. Syn. Fil. 180 (1806). Only in the alpine heath, rare. Fertile in the autumn. Masafuera: In the higher parts of the Alpine region, Las Torres, c. 1350 m, scarce (no. 482). — New for Juan Fernandez. A very robust form, with deeply subterraneous rhizome, larger than any form we have seen from. Subantarctic districts. Area of distribution: South Chile, Valdivia to Fuegia and Patagonia; Atl. subantarctic islands; Masafuera. New Zealand and subantarctic islands. 51. L. Gayanum Remy; Gay 6: 545. Subalpine and alpine heaths. Fertile in the autumn and winter, Masafuera: Stony heath in the highlands above the Mono and Pasto val- leys, I100—1300 m, not uncommon (no. 416); C. del Barril, 740 m (no. 534). — Discovered by SKOTTSBERG 1908, at the Correspondencia Camp. BAKER, Fern Allies 29 regards L. Gayanum as intermediate between the typical ZL. scartosum Forst. from New Zealand and its var. Fusszaec Desv. from tropical America. We prefer to regard it as specifically distinct from both. Area of distribution: South Chile, Valdivia to the Guaitecas Islands; Masafuera. Before the investigations of SKOTTSBERG, the known vascular cryptogams of Juan Fernandez numbered 39 species; in 1908 3 species were added, and in I916—17 9 more, giving 51 species as a total. Of these 36 species are com- mon to Masatierra and Masafuera, 6 only found in Masatierra and 9 only in Masafuera. Of the 51 species 17 are endemic in the islands (3 of these confined to Masatierra and 1 to Masafuera). Two species are known elsewhere, but represented by endemic varieties. Of the genera only one, 7yrsopieris, in endemic. The percentage of endemic species is high, amounting to 33,3 %. 46 Plate Plate Plate Plate Plate ie CARL CHRISTENSEN AND CARL SKOTTSBERG Explanation of the Plates. All figures reproduced after photographs taken by C. SKOTTSBERG. 1. Hymenoglossum cruentum (Cav.) Pres] on the slope of La Damajuana, Masatierra. Asplenium magellanicum K\f. in the centre and on the left side. 2. Asplenium macro- sorum Bert. in Q. de las Chozas, Masafuera. Dicksonia Berteroana (Colla) Hook. var. virgata C. Chr. et Skottsb. in Q. del Blin- dado, Masafuera. Mrs. S. at the foot of the tree. Dryopteris inaegualifolia (Colla) C. Chr. near Portezuelo, Masatierra. Blechnum cycadifolium (Colla) Sturm near the road to Portezuelo, Masatierra. Mrs. S. in the foreground, Blechnum longicauda C. Chr. in Q. de las Casas, Masafuera. ia fe ( ] Fernandez and Faster T[s/. Juan Hist. F. Vat. PLAT! Isl. Vol. 1. and faster Fernandes Juan ry Nat. First. Nat. Hist. Fuan Fernandes and keaster /s/. l i, /. 4 = ee eae - Nat. Hist. Fuan Fernandez and Easter 1 Vol. 1 Wr/ ti. iy dg Aytier" “ age os hy ‘ ‘ Mt Meg as da - N See a Cee aX reek “Bie f &% ! EMR las Ney ie 4, De A es q rune he 2. The Ferns of Easter Island. By CARL CHRISTENSEN and CARL SKOTTSBERG. With 3 text figures. The present collection was made by the undersigned, assisted by Mrs. SKOTTSBERG, during the Swedish Expedition to the Pacific, 1916—17. We spent the later half of June, 1917, on Easter Island (Isla de Pascua). Before our visit, only one representative collection’ seems to have been made on the island, by Mr. F. FUENTES of the Museo Nacional in Santiago, 1911; this was kindly put to my disposal in the said Museum. A list of Mr. FUENTES’ ferns was published by C. M. HICKEN: Contribucion al Estudio de las Pteriddfitas de la Isla de Pascua i descripcidn de dos nuevas especies (Bol. Mus. Nac. Chile V: 1, 1913); also in Rev. Chil. Hist. Nat. XVII (1913). The ferns were also included in FUENTES’: Resena botanica sobre la Isla de Pascua (Inst. Centr. Meteor. y Geofis. de Chile no. 4, 1913 and Bol. Mus. Nac. Chile V: 2, 1913). FUENTES’ list contains three species, not mentioned by HICKEN, Asfplenium obtusatum, Dryopteris parasitica and Polystichum aculeatum; Mr. HICKEN, con- sequently, did not receive specimens of these.. We did not find P. aculeatum or any allied species, nor were there any specimens of it in the collection shown to me by Mr. FUENTES. In the short list of plants communicated by HEMSLEY (Report Challeng. Bot. I: 15) no fern figures. Dr. CHRISTENSEN kindly undertook to revise our collection and the new forms described are due to him. A map, of the island, showing the localities quoted, is published in Vol. I, No. 1. — Altitudes in meters above sea level. An* before a name signifies an addition to the flora. The numbers in brackets refer to our collection. ; C. Skottsberg. Polypodiaceae. Asplenium L. #1. A. adiantoides (L.) C. Chr. Ind. 99 var. squamulosum noy. var. A forma typica differt: foliis utrinque ad rachidem venasque dense paleaceis; paleis rhachidis fuscis, venarum minutis, e basi lata fimbriata subito in apicem subulatum angustissimum contractis. 1 Nothing has been published on the »Albatross» collection (AGAssiIz a. 0.). 48 CARL CHRISTENSEN AND CARL SKOTTSBERG Hab. in the crater lake of Rano Kao, c. 110 m., growing on the firm moss peat (no. 677; also AGAssiIz!). — Fertile. — Not recorded before. Area of distribution: Asia, trop. Australia, New Zealand, Polynesia, Afr. orient. and adj. islands. 2. A. obtusatum Forst. — FUENTES, Bol. 327. Rocks near the sea shore, probably not uncommon. Hab. Estern headland, humid fissures on the steep slopes of Vaintu Rova, c. 275 m; Hanga Ho Orno, coast cliffs (no. 633); between Hanga Piko and Pta Baquedano, not uncommon, well developed. — Fertile. Area of distribution: West coast of South America to Chiloé; Austra- lia, New Zealand, Oceanic Islands. Doodia R. Br. *2. D. paschalis nov. spec. Fig. 1. Rhizomate obliquo, brevi, paleis nigris, filiformibus cum parte basali stipitis subdense onusto. Stipitibus fasciculatis, strictis, parce scabris vel sublaevibus, a basi ad medium atris vel nigrescentibus, sursum pallide stramineis vel fuscis, basi paleis filiformibus crinitis, sursum glabris, superne late sulcatis instructis, ad 15 cm longis. Lamina pinnatifida, oblanceolata vel rarius subgradatim attenuata, versus basin subito contracta, ima basi saepe hastata, versus apicem breve caudata, usque ad 30 cm. longa, supra medium 6—10 cm lata, papyracea vel subcoriacea, obscure viridi; rhachide et costa mediana straminea, minuto elanduloso-pubescente, superne compressa, late canaliculata; pagina superiore glabra, inferiore minuto glanduloso-pubescente, postea glabra. Segmentis usque ad 30-jugis, approximatis, basi confluentibus, oblongis, obtusis, maximis 3,5;—5 cm longis, 6—8 mm latis, marginibus inaequaliter argute serratis, inferioribus 3—5-jugis subito contractis, semiorbicularibus vel triangularibus sub I cm. longis, infimis frondium perplurimum elongatis, 2—2.5 cm longis. Venulis extus seriem areolarum costalium plerumque furcatis, in dentes abeuntibus. Soris nervo mediano valde approximatis, brevissimis (I—15 mm longis), plus minusve lunatis, utroque latere costz uniseriatis; rarius secunda series interrupta prime approximata adest. Indusiis integris, angustis. Hab. Eastern headland, humid fissures on the steep slopes of Vaintu Rova and Tea-tea (no. 655); Rano Aroi, rocks inside the crater, c. 420 m, rare. — Fertile. — A new addition to the flora. This new species is closely allied to D. dblechnoides A. Cunn. from Austra- lia, at least the plant cultivated under that name and accurately figured by METTENIUS (Fil. Lips. tab. 6 fig. 3—4), but BAKER (Syn. Fil.) says that the genuine 2. blechnoides does not show the reduced lower pinnz of our species. METTENIUS’ description covers our plant very well with the exception that young D. paschalis is rather densely glandulose beneath, and that, in larger leaves, there are traces of a secord series of sori. Remarkable is the pair of leng- thened basal pinnae which make the whole lamina hastate; sometimes only one of the two pinne is long, rarely both are abbreviated. From the common Polynesian DY. media R. Br. and its numerous forms 2). paschalis differs in its pinne being contiguous at base, in the sori being placed close to the midrib, and in the furcate veins outside the costular areoles. Area of distribution: Endemic. Fig. 1. 49 “> ei - = a a_i CF 7 — i yseouoc « re. Stes ttr ss, he —- ed rr) Co re , ee SSRI ORI AR RES Stee, : Fig., 2: Fig. 1. Doodia paschalis C. Chr.; entire frond, half natural size; @ 2 pinnae from different spe- cimens, nat, size; 6 pinna with partly double rows of sori, nat. size; ¢ fragment of pinna, x 2. Elaphoglossum tahitense Brack.; entire plant, half natural size; @ fragment of frond, c from the stipe, d from lower side of leaf, all Fig. 2. nat. size; d—d paleae, 6 from the rhizome, enlarged. Voki. The Nat, Hist. of Juan Fernandez and Easter Isl. 4 — 20206. 50 CARL CHRISTENSEN AND CARL SKOTTSBERG Dryopteris Adans. 4. D. Espinosai Hicken, Bol. 135, tab. XXVII, Rev. 94, tab. 9. — Fig. 3. Hab. Only found in the crater of Rano Aroi, c. 420 m, between large boulders (no. 690). — Fertile. This species, recently described and figured by HICKEN after specimens collected by FUENTES, is a very remarkable one, belonging to the subgenus Ludryopterts which is very weakly represented in Central Polynesia but rather rich in species in the Hawaiian Islands. YD. sfznosaz, however, differs greatly from all Polynesian and Asiatic species of that subgenus known to us, while, on the other hand, it is closely related to a group of Tropical American spe- cies (D. patula, cinnamomea and mexicana). In most features it agres very well with D. mexicana (Pr.) C. Chr. (syn. Aspidium flaccidum), being, however, much larger, and with certain forms of the Brazilian D. patula (Sw.) Underw. It differs from these especially in the glandulose-ciliate basal scales and in the very glandulose lamina. As the original description was published in two perio- dicals rather rare in Europe, a full description is given below. Rhizomate obliquo cum parte basali stipitum paleis lutescentibus vel pallide rubro-brunneis tenuissimis, lanceolatis, apice longe filiformibus, marginibus glan- dulosis, dense obtecto. Stipitibus fasciculatis, stramineis, ad 15 cm longis. Lamina textura tenuissima, laete virescente, ubique dense glandulosa, deltoidea — ovata vel ovato-lanceolata, circiter 20 cm longa, maxima 20 cm. lata, tri- quadripinnatifida. Pinnis saepe fere oppositis, distantibus, ovatis vel ovato- lanceolatis, maximis 10 cm longis, breve stipitatis, acuminatis vel subobtusis, infimis parum inaequilateralibus, latere basiscopico vix: aucto. Pinnulis secundi ordinis inferioribus liberis, superioribus decurrenti-confluentibus, fere ad basin pinnatifidis, segmentis tertiariis decurrentibus, integris, crenato-dentatis vel maximis subpinnatifidis. Venis medianis basi decurrentibus, ultimis indivisis vel interdum furcatis. Soris majusculis, saepe uno pro Jobo ultimo, nervo mediano approximatis. Indusiis reniformibus, subpersistentibus. Area of distribution: Endemic. #5. D. gongylodes (Schkuhr) OK. Hab. Rano Aroi, c. 400 m, edge of the crater lake with Polygonum acuminatum, very scarce; Rano Kao, bottom of crater, c. 110 m, in the Scir- petum, and among the large boulders a little higher up (no. 678). — Fertile. — Not recorded before. Area of distribution: Pantropical. 6. D. parasitica (L.) OK. — FUENTES, Bol. 327. Hab. Crater of Rano Aroi, c. 420 m, and also crater of Rano Kao, c. 125 m (no. 666), among boulders. — Fertile. This is taken in a wide sense; probably DY. favasztica includes several distinct species. Area of distribution; Tropical and subtropical regions. 7. Elaphoglossum tahitense Brack. — Syn. £. Gayanum, Hicken Bol. 137, Rev. 96. zon Moore. — Fig. 2. Hab. Fissures in the rocks in Rano Aroi, c. 420 m (no. 691). — Fertile. THE FERNS OF EASTER ISLAND 51 The specimens collected agree in all essential characters with the descrip- tion of BRACKENRIDGE; previously, it was known only from Tahiti. It is a much larger species than /£. Gayanum (Fée) Moore, to which the specimens in the FUENTES collection were referred by HICKEN; moreover, it is well marked by the scattered minute scales on the underside. Area of distribution: Tahiti; Easter Island. Fig. 3. Dryopieris Esfinosat Hicken; basal pinna, natural size; a pinnule of 1. order, X 2, & af 23 order! 5<"4. Microlepia Pres 8. M. strigosa (Thbg) Presl. — HICKEN Bol. 132, Rev. 90. The only common fern, probably growing all over the island in scattered specimens, generally among rocks and stones, Hab. Eastern headland, Vaintu Rova and Tea-tea; plains near Hanga Ho Orno, not rare (no. 647); Rano Aroi, fairly common; between Hanga Roa village and Mataveri, scattered; Rano Kao, in the crater, not rare (no. 1125). — Fertile. Area of distribution: SE. Asia, Polynesia. Polypodium L. 9. P. phymatodes L. — Syn. P. phymaiodes et P. Fuentesi Hicken, Boly.132 tab. XXVI, Rev. 91 tab. 8. 52 CARL CHRISTENSEN AND CARL SKOTTSBERG Hab. Eastern headland, Vaintu Rova and ‘Tea-tea, humid fissures, not rare (no. 654); Rano Aroi, fairly common; Rano Kao, stone-heaps near the lake, c. 125 m (no. 673), but also seen higher up. — Fertile. This common Polynesian fern varies in the island as in most other loca- lities greatly in size and division, the leaves now being entire, now trifid or pinnatifid with several pair of segments, which are longer or shorter, acute or obtuse. PP. Fuentesizz is one of the larger forms, but certainly not ae able even as variety. Typical /wenteszz is in the coliection. Area of distribution: The tropics of the Old World, including Austra- lia and Polynesia. Vittaria J. Sm. 10. V. elongata Sw. — Syn. V. costata Hicken Bol. 137, Rev. 96, non Kunze. Hab. In fissures between blocks, with moss, in the craters of Rano Aroi, c. 420 m and Rano Kao, c. 125 m. (no. 672), in the last mentioned also in the moss peat of the crater lake. — Fertile. This very common Polynesian species is totally different from the Ame- rican V. costata Kze, to which HICKEN referred specimens from the island. Area of distribution: Tropical Asia, Polynesia, N. S. Wales. Ophioglossaceae. Ophioglossum L. #11. O. reticulatum L. Probably not uncommon in short grass on the plains and slopes. Hab. West slope of Mt Katiki (no. 1184); near the landing-place in Hanga Ho Orno (no. 648); near Mataveri; also seen in the centre of the island. — Fertile. — Not recorded before. Area of distribution: Tropical Africa and America; Polynesia. * ‘12. O. coriaceum A. Cunn. Hab. With the preceding, slope of Mt Katiki (no. 651), and at Hanga Ho Orno. — Fertile. — Not recorded before. We have some doubt of this being the true O. corzaceum A. Cunn.; it certainly belongs to O. coréaceum Prantl, Mon. Ophiogl. 314 tab. 7 fig. 9, 10, 15; on the other hand, it falls under O. valgatum var. australasicum Luetss. Mus. -Godeffr. 3: 246, corresponding closely to his fig. 73. PRANTL refers LUERSSEN’s variety to O. pedunculosum; to this species our specimens certainly do not belong. Area of distribution: Australia, Tasmania, New Zealand; Easter Island; Bolivia. THE FERNS OF EASTER ISLAND 53 If we exclude the dubious record of Polystichum aculeatum, Mr. FUENTES’ list comprises 7 species (Polypodium [uentesii was reduced to P. phyma- todes), which were all found by us with addition of 5 species not recorded before, making a total of 12 ferns known in Easter Island. Two of these are endemic. ak is a ieee Pale & \ Pie ne, rae yee eT eee Fae oe: thy B) manic is sé 3. Die Myxomyceten der Juan Fernandez-Inseln. Von ROB. E.. FRIES. Die von Dr. C. SKOTTSBERG 1916--17 auf den Juan Fernandez-Inseln zusammengebrachte Myxomycetensammlung, die mir giitigst zu Bearbeitung anvertraut worden ist, umfasst insgesammt 16 Arten. Alle diese sind fiir die Inselgruppe neu, mit Ausnahme von zwei Arten, die laut JoHOwW (Flora de las Islas de Juan Fernandez (1896) S. 195) schon dort angetroffen sind. Noch zwei, von SKOTTSBERG nicht gesammelt, werden daselbst angefiihrt. Insge- sammt sind also 18 Arten von der Inselgruppe bekannt, eine (Lzcea Schoenleiniz; siehe unten) jedoch sehr unsicher. Alle diese sind von der Insel Masatierra bekannt, nur drei derselben ausserdem von Masafuera. Diese von Juan Fernandez bisher bekannten Myxomyceten, iiber welche ein Verzeichnis hier mitgeteilt wird, bilden sicherlich nur einen kleineren Teil der dort vorkommenden Arten. Sie bieten allerdings ein grosses Interesse dar, weil sie deutlich zeigen, dass die Myxomycetenflora gar nicht denselben Reich- tum endemischer Arten wie die der héheren Pflanzen aufweist. Im Gegensatz sind, wenn man von der sehr unsicheren, JOHOWschen Lzcea-Art absieht, alle ausgepragte Kosmopoliten. Um dies zu beleuchten diirfte es geniigen zu er- wahnen, dass von den 17 guten Arten nicht weniger als 13 sowohl aus Europa, Asien, Afrika, Australien als Nord- und Siidamerika schon bekannt sind, die ubrigen 4 aus wenigstens Europa, Asien und Nordamerika. Ceratiomyxa fruticulosa (Muell.) Macbr. Von JOHOW unter dem Namen Ceratium hydnoides Alb. & Schw. fiir Masatierra angegeben. Physarum cinereum (Batsch) Pers. Masatierra: Siidwestseite von Portezuelo, im Wald ungef. 500 m ii. M. (1orJan. TI 7): ROB. E. FRIES Kesh -~ 9. Die Exemplare gehoren der Form an, die von G. LISTER (Monograph of the Mycetozoa. Ed. Il. S. 73. Ig11) als cevereum in beschranktem Umfang aufgefasst wird. Die Sporen sind namlich hell braun, glatt, grosstenteils 8—9 p, Diam. Sie variieren jedoch sehr und erreichen sogar Dimensionen bis zu 18 u.; diese Riesensporen sind oft ellipsoidisch, ei- oder nierenformig und diirf- ten wohl von ungiinstigen Entwicklungsverhaltnissen bedingt sein. Physarum sinuosum (Bull.) Weinm. ex Fr. Masatierra: Quebrada de la Damajuana (30. Dez. 1916); Wald bei Pla- zoleta del Yunque (28. Dez. 1916), — Gut entwickeltes und typisches Material. Fuligo septica (L.) Gmel. Masatierra: Cordén del Centinela (11. Apr. 1917). — Die Exemplare ge- hoéren der var. flava (Pers.) an, die tibrigens schon von BERTERO und JoHOW auf Masatierra gesammelt wurde. Diachza leucopoda (Bull.) Rost. Masatierra: Wald in der Nahe von Plazoleta del Yunque (28. Dez. 1916); Quebrada de la Damajuana (30. Dez. 1916). — Typisch, auf trockenen Zweigen und Blattern (von Arzstotelza, Farnen u. a.) reichlich wachsend, Didymium melanospermum (Pers.) Macbr. var. manus List. Masatierra: Quebrada de -la Damajuana (30. Dez. 1916); Cordén del Centinela (11. Apr. 1917). Didymium nigripes (Link) Fr. Masatierra; Quebrada de la Damajuana (30. Dez. 1916); Pangal (1. Jan. 1917); Pico Central (18. Jan. 1917). Alle sind zur Hauptart mit dunkel orange- farbigen Stielen zu rechnen, die Exemplare von Pangal sind jedoch auffallend grob gebaut. Stemonites splendens Rost. Masatierra: La Damajuana, im Wald ungef. 500 m ii. M. (29. Jan. 1917). — Gehort zur Hauptart. Das oberflachliche Netzwerk des Kapillitiums hat kleine Maschen (15—40 ». Diam.); die Sporen sind glatt und zeigen sogar bei 1200 mal Vergrosserung keine deutlichen Warzen. DIE MYXOMYCETEN DER JUAN FERNANDEZ-INSELN Comatricha nigra (Pers.) Schroet. Masatierra: La Damajuana, ca. 500 m ii. M. (29. Jan. 1917); Wald bei Yunque (Aug. 1917). — Alle Exemplare haben zylindrische (ca. 2 mm lange) Sporangien, die im Bau des Kapillitiums mit der vay. dava (Rost.) R. E. Fr. (Syn. Comatricha laxa Rost.| ibereinstimmen, Cribraria aurantiaca Schrad. Masatierra; Quebrada de la Damajuana (9. Apr. 1917). Licea flexuosa Pers.? Die von JoHow auf Masatierra gesammelte und unter dem Namen Licea Schoenleinit beschriebene Art ist der kurzen Beschreibung nach zu urteilen wahr- scheinlichst mit ffexwosa zu identifizieren (vgl. LisrER, Monogr, ed. II. S. 189). Lycogala Epidendrum (L.) Fr. Masatierra: Juanango (9. Apr. 1917). Trichia verrucosa Berk. Masatierra: ohne naher angegebenen Fundort. Masafuera: Las Chozas (26. Nebr. 1917). Die Elateren beider Kollektionen sind auffallend dick, ungef. 7,5 . in Diameter. Die Exemplare von Masafuera sind im iibrigen ganz typisch, die von Masatierra bieten ein besonderes Interesse dar, weil sie teilsweise Uber- eangsformen zu 7richia affinis De Bary aufweisen. Die Sporangien sind nam- lich teils fast ganz stiellos, teils kurz, teils typisch lang gestielt. Sie sind jedoch im allgemeinen, wie bei der echten verrucosa, gruppenweise vereinigt. Trichia decipiens (Pers.) Macbr. Masatierra: Quebrada de la Damajuana, ungef. 500 m ii. M. (29. Jan. 1917). — Nur ein paar kleine Sporangiumbruchstiicke, dank dem charakterist- ischen Bau des Kapillitiums und der Sporen allerdings gut erkennbar. Trichia Botrytis Pers. Masatierra: Valle Colonial, Pico Central 350 m ii. M. (18. Jan. 1917). Die Exemplare gehéren der var. /ateritia (Lév.) List. an, die zuerst aus Chile — als Art — beschrieben worden ist. Nach LisTER (Monogr. ed. IL. 4* — 20206. The Nat, Hist. of Juan Fernandez and Easter Isl. Vol. II. 58 ROB. E. FRIES S. 218) haben diese Exemplare »nearly black sporangia». Das Material von Juan Fernandez hat auch schwarze Sporangien, die einzeln sitzen und von langen, helleren Stielen getragen werden. Das Kapillitium ist typisch. Hemitrichia clavata (Pers.) Rost. Masatierra; im Wald auf dem Gebirgsriicken zwischen den Piedra agu- jeriada und Laura-Talern, ca. 500 m ii. M. (4. Apr. 1917). —- Wohl entwickelte, typische Exemplare mit gestielten Sporangien; das Kapillitium mit wenigen Verzweigungen. Von JOHOW ist diese Art sowohl fiir Masatierra als fir Masa- fuera angegeben worden. Arcyria denudata (L.) Macbr. Masatierra; Plazoleta del Yunque (28. Dez. 1916); Valle Colonial, Pico Central, ca. 350 m wu. M. (18. Jan. 1917); Quebrada del Juanango (3 Apr. 1917). Masafuera: Quebrada de las Casas (11. Febr. 1917). Arcyria insignis Kalchbr. & Cooke. Masatierra: Quebrada de la Damajuana (30. Dez. 1917). , 54. 4 a ao, haw: § 4 a 2 plete “a : THE NATURAL HISTORY OF JUAN FERNANDEZ A. Bee AND EASTER ISLAND BDITED BY DR. CARL SKOTTSBERG a ~ o #2 oF a Se et yr ay > VOL; TI . BOTANY PART: bi 4. THORE C. E. FRIES: Die Gasteromyceten der Juan Fernandez- und Osterinseln. Wee CARL SKOTTSBERG: The Phanerogams of Easter Island. 6. K. MUNSTER STROM: Freshwater Alge from Juan Fernandez and Hastie Tend: s 7. CARL SKOTTSBERG: The Phanerogams of the Juan Fernandez Bs Islands. ; oe ‘ine UPPSALA ne 2:2 ss ALMQVIST & WIKSELLS BOKTRYCKERL-A--B. 4. Die Gasteromyceten der Juan Fernandez- und Osterinseln. Von THOKE. CE; FRIES, Dr. C. SKOTTSBERG hatte die Giite, mir das von ihm wahrend seiner Reise 1916—1917 gesammelte Gasteromycetenmaterial zur Bestimmung zu iiber- geben. Die Sammlung ist unbedeutend und enthalt nur drei Arten. Doch sind samtliche neu fiir die in Frage kommenden Inseln. Vorher hat JoHOW (Flora de las Islas de Juan Fernandez (1896) p. 188) zwei Gasteromycetenarten von Juan Fernandez angegeben. Mit Riicksicht dar- auf, dass die Kenntnis der Gasteromycetenflora sowohl auf Juan Fernandez als auf der Osterinsel noch immer als sehr mangelhaft angesehen werden muss, bietet die fragliche Pilzgruppe wenig pflanzengeographisches Interesse. Samtliche nach Art bestimmbare Kollektionen gehdren doch bekannten Arten aus Australien und Amerika an. — Es diirfte deshalb mit grosser Wahrschein- lichkeit angenommen werden k6nnen, dass die Gasteromycetenflora der Juan Fernandez- und Osterinseln eine grossere Anzahl endemischer Arten nicht auf- zuweisen haben. Tleodictyon gracile Berk. Von JOHOW angegeben aus Masatierra. Mit Riicksicht auf die ziemlich schwebende Artbegrenzung innerhalb der Genus //eodictyon diirfte [. cebarium Tul. als mehr Kollektiv vorzuziehen sein. — Ausbreitung: Australien, Neu- seeland, Siidamerika und Siidafrika. Calvatia sp. JOoHOW nimmt in seine Liste ein Lycoperdon sp. (an nova?) aus Masa- tierra und S:ta Clara auf. Aus der Beschreibung geht mit voller Deutlich- keit hervor, dass eine Art von der Gattung Ca/vatia (Fr.) Morg. gemeint wird. Ohne Zuginglichkeit des JoHow’schen Materiales ist indessen eine Artbestim- - mung unmoglich. 4** — 20100. The nat. Hist. of Juan Fernandez and East. Isl. Vol. II. 60 THORE ¢. E. FRIES Lycoperdon (?) sp. Ein einziges unreifes Exemplar. Artbestimmung unmoglich; ist indessen nicht identisch mit JoHOW’s Lycoperdon sp. (an novar). — Masafuera: Los Inocentes, ca. 1100 m u. M. °/s 1917. C. SKOTTSBERG. Bovistella pusilla Lloyd (in Myc. Not. N:o 34. 1910). Drei gute Exemplare. — Osterinsel: Cerro Katiki *°/6 1917. C. SKOTTS- BERG. Ausbreitung: Australien. Geaster Ohiensis (Cooke) Lloyd. (Syn. Cycloderma Ofiensis Cooke und Geaster velutinus Morg.). Zwei ungedffnete Exemplare. — Masatiera: Siidabhang des Yunque 7*/4 1917. C. SKOTTSBERG. — Ausbreitung: Nordamerika, Samoa-Inseln. 3. The Phanerogams of Easter Island. By CARL SKOTTSBERG. With Plates 6—9 and 2 text-figures. In no. 2 of this volume I have published, together with Dr. CHRISTENSEN, a list of the ferns reported from Easter Island. While reading the proofs I received from Professor B. L. ROBINSON of Harvard University, Cambridge, Mass., a complete set of the collection made by AGAssIz and his comrades during the Albatross Expedition (1904). It was then impossible to take into due consideration the ferns contained in this collection. It does not extend our knowledge of the fern flora; still I find it better to enumerate the species found by AGASSIZ with the remarks written on his labels as an appendix to this paper, especially as his collection was made before that of FUENTES. The latest and most complete enumeration of the flowering plants is to be found in FUENTES’ »Resefia botanica sobre la Isla de Pascua» (Inst. Centr. Meteor. y Geofis. de Chile no. 4, 1913 and Bol. Mus. Nac. de Chile V: 2, 1913). The earliest collection seems to have been made by G. FORSTER du- ring COOK’s second voyage: most of the plants found by FORSTER were enumerated in his »Florulae insularum australium prodromus», Gottingen 1786. Some species mentioned by FORSTER in his narrative (A voyage round the world, London 1771 vol. 1) do not figure in the Prodromus, as for example such an important plant as Sophora (the Mimosa of FORSTER). A. v. CHAMISSO, during the voyage of KOTZEBUE (see »Reise um die Welt, 4. Aufl., Berlin 1856) also collected plants on Easter Island, and has been quoted as collector in a few cases. In 1836, S. ENDLICHER included all statements related to Easter Island plants in his »Bemerkungen tiber die Flora der Siidseeinseln (Ann. des Wiener Museums der Naturgeschichte I). HEMSLEY’s list in »Report on the present state of knowledge of various Insular Floras» (Report Chall- enger, Bot. I p. 15, 1885) is based on ENDLICHER, but Sophora tetrapitera and Sesuvium portulacastrum are added. FUENTES confined himself to an enumeration of what he had collected on the island in 1911; his list contains 40 indigenous or naturalized species. Many of these had been found by the Albatross expedition seven years earlier, 4*** — 29100. The nat. Hist. of Juan Fernandez and East. Isl. Vol. II. 62 CARL SKOTTSBERG together with several species not found by FUENTES. On the request of Pro- fessor ROBINSON I have determined the »Albatross» collection. All the deter- minations, together with the notes on the labels’, as well as all previous re- cords known to me, are included in my list. Finally, it should not be forgotten that the cultivated plants were treated by FUENTES (I. c.) and lately by W. KNOCHE in »Ueber die Kulturpflanzen der Osterinsel» (Zeitschr. des Deutsch. wiss. Ver. zur Kultur- und Landeskunde Argentiniens, Buenos Aires 1919). In some cases we cannot tell whether a plant used by the natives as a food plant or for other practical purpose was a member of the flora when the first settlers arrived or if it was introduced by the natives on that occasion or later (but in any case before the discovery ot the island in 1722). When I made my remarks on the cultivated plants (Notes on a visit to Easter Island, Vol. I of this work), the news of KNOCHE’s paper had not yet reached this country. According to him, the sweet potato existed in Easter Island already before the visit of ROGGEVEEN, the discoverer; for BEHRENS, who accompanied ROGGEVEEN, enumerates »Bataten, die wie Brot schmeckten». ROGGEVEEN himself did not mention this plant, unless it was included under »aardgewasch». However, as BEHRENS expressly states that sweet potatoes were plentiful, they must have existed long before the Europ- eans landed on the island. In the determination of my plants I was assisted by Prof. Dr. R. PILGER (Gramineae), Superintendent Dr. G. KUKENTHAL (Cyperaceae, see FEDDE’s Repertorium XVI, 1920), Mr. J. HUTCHINSSON (Sisyrinchium, Polycarpon), Mr. O. E. SCHULZ (a Nasturtium) and Prof. Dr. G. BITTER (a Solanum). I hereby cordially thank them for their kind assistance. My thanks are also due to Dr. O. STapF and other members of the staff at the Kew Herbarium for great help in various matters during my visit to that establishment. An* before a latin name signifies that the species has not been recorded before. The specimens collected by the Albatross Expedition are quoted thus: Albatross no. —-! All other numbers refer to the collection made by the writer and Mrs. SKOTTSBERG. The zoologist, Mr. BACKSTROM, brought me a few plants from Motu Nui, the small rock near the southwestern promontory of Easter Island. A map, showing the localities mentioned below, was pub- lished in Vol. I, No. 1. 1 »Rana Koa» and »Rana Roka» on the labels have been corrected to Rano Kao and R. Raraku. THE PHANEROGAMS OF EASTER ISLAND 03 Catalogue of Easter Island Phanerogams. 1. Species undoubtedly or probably indigenous. Gramineae. Paspalum L. *7, P. Forsterianum Fliigge. — Plate 6, fig. 1. Albatross no. 7! Hanga Ho Orno, near the sea (no. 641); rocky beach forth of Hanga Piko (no. 665); Islet Motu Nui. Dr. PILGER remarks: »Mir lag ein ziemlich unvollkommenes Exemplar aus dem Hb. SPRENGEL vor (Nova Caledonia: FORSTER): die unteren Halmteile fehlen an dem Exemplar. Doch erscheint die Zugehorigkeit der Exemplare von der Oster-Insel zur selben Art sicher.» Area of distribution: New Caledonia, Easter Island. 2. P. scrobiculatum L. var. ordiculare (Forst.) Domin. FORSTER; FUENTES. — Between Hanga Ho Orno and Puhi; not uncom- mon round Mataveri; Rano Kao (no. 694). Area of distribution: Austral., N. Caled., N. Guinea, Polyn. Axonopus Beauv. 3. A. paschalis Pilger nov. spec. — Syn. P. scoparmum Fluigge var. nov. oligostachyum WHACK. ex FUENTES, |. c., P. paschale Stapf, nomen, ibid. — Plate 6, fig. 2. Culmus erectus, vaginatus, apice breviter cum panicula exsertus, vel usque ad basin paniculae vaginis vestitus, vaginis ad basin culmi distiche sese tegenti- bus compressis, vagina suprema sub panicula elongata, ad 20 cm longa; culmus cum panicula 30—50 cm altus, e vagina suprema nonnumquam florifero-ramosus, nodis dense albido-pilosus, innovationes extravaginales erectae; foliorum lamina linearis, secus medianum plicata vel demum subplana, apice ipso brevissime contracta et acuta, apicem versus scaberula, ceterum + hirsuto-pubescens vel fere glabrescens, basi longe hirsuta, ligula margo brevissimus ciliolulatus; pani- cula spicis paucis erectis composita, rhachis communis brevis; spicae ad g—10 cm longae strictae, rhachis angulata, triquetra, marginibus scaberula; spiculae singulae, biseriatae, ovales, apice parum attenuatae, compressae, 3 mm longae, gluma infera et gluma florifera anticae; gluma vacua infera dorso applanata, marginibus inflexa, membranacea, ovata, subobtusa, nervus medianus nullus, laterales + conspicui utroque latere 3—4; gluma supera ovata, magis angus tata, dorso haud applanata, medianus tenuis, nervi laterales utroque latere 2—3; glumae dorso ad nervos pubescentes; gluma florifera indurata straminea, elumis 64 CARL SKOTISBERG vacuis paulo brevior, extus visa ovalis, expansa rotundato-elliptica, apice minute puberula. Apparently widely dispersed over the island. — FUENTES; »common throughout the island», summit of Rano Kao (Albatross no. 49!); Mt. Katiki, in the crater, 412 m (no. 1191); Hanga Ho Orno (no. 644); between Mataveri and Rano Raraku; slopes of Rano Kao, scarce; crater lake of Rano Kao, on the moss peat, 110 m (no. 1189). Dr. PILGER remarks: »Die Gattung Axonxopus ist hier im Sinne der ameri- kanischen Agrostologen gefasst, mit 4. compressus (Sw.) Beauv. (Paspalum platycaulon Poir.) als typische Art (vergl. iiber die Gattung A. CHASE, Notes on Genera of Panicum IV, Proc. Biol. Soc. Washington XXIV (1911) 129 ff.). Die Art ist wohl dieselbe, die in der Arbeit von FUENTES als Paspalum sco- parium Fligge var. oligostachyum Hack. oder P. paschale Stapf nov. spec. ohne Beschreibung aufgefiihrt wird. Eine Beschreibung ist mir nicht bekannt ge- worden. Axonopus paschalis ist mit Paspalum scoparium, welche Art auch zu Axonopus zu stellen ist, verwandt, aber gut unterschieden, neben anderen Merk- malen schon durch den Bliitenstand.» Area of distribution: Endemic. Stipa L. *4. S. horridula Pilger nov. spec. — Plate 6, fig. 3; text fig. 1a. Perennis caespitosa, innovationibus intravaginalibus dense aggregatis; folio- rum innovationum lamina stricta, angustissima, filiformis, subtereti-convoluta, brevissime hirto-scabra, ad 30 cm longa, ligula brevis bilabiata, vagina brevis striata, margine superne breviter ciliolata, ad 4—6 cm longa; culmus caespitem innovationum valde superans, erectus, cum panicula 80—go cm altus, 3—4-nodus, vaginatus, vaginis internodia parum longitudine superantibus, nodis glabris, la- mina foliorum culmi ad 2 cm longa; internodium supremum elongatum, panicula basi vagina circumdata; panicula 30—40 cm longa angusta laxiflora, rami a basi pluries divisi, ramuli igitur ad nodos subfasciculati breves erecti, infimi ad 7 cm longi, internodia inferiora paniculae 3—5 cm longa; spiculae angustae; glumae vacuae tenues angustae nitidulae, longe tenuiter angustatae acutissimae, subaequales, 13—14 mm longae, scaberulae; gluma florifera angusta, 4 mm longa, callo elongato acutissimo, albido- vel flavidulo-hirsuto, circ. 2 mm longo instructa, satis parce albido-hirsuta, coronula nulla, arista tenuis 5—6,5 mm longa, parte infra geniculum parum torta circ. I cm tantum longa, parce hir- tula, parte supra geniculum scaberula. Mt Katiki, stray patches on the western slope (no. 660). Dr. PILGER remarks: »Die neue Art gehdrt zur Gruppe der Stipen mit unbefiederter Granne und fehlendem Krénchen, die Granne ist gegen die Deck- spelze nur einfach abgegliedert. Verwandte Arten kommen sowohl in Siid- amerika wie in Australien vor (etwa S. scabra Lindl., etc.). Ein naherer geo- graphischer Anschluss der schénen neuen Spezies ist somit nicht sicher an- zugeben. Area of distribution: Endemic. THE PHANEROGAMS OF EASTER ISLAND 65 Sporobolus R. Br. 5. S. indicus (L) R. Br. One of the most important grasses, covering considerable areas. — FUENTES. »Common throughout the island» (Albatross nos. g! 23!). Observed in great quantities along the north coast and on Mt. Katiki (no. 639) up to the top, 412 m (no. 1190); also common round Mataveri (no. 702) and along the west coast as well as on Rano Kao and from here to Rano Raraku, which is covered by this species (also Albatross no. 48)). Area of distribution: S. Asia to Australia and Polynesia. Agrostis L. 6. A. filiformis (Forst.) Spreng. Fl. Hal. Mant. Prima (1807) p. 32. — ext fig: 1 b, c. Dr. PILGER remarks on this species: »FORSTER beschrieb die Art zuerst in Florul. Ins. Austr. Prodr. (1786) 9 als Avena filzformis (panicula erecta tenuissima, calycibus unifloris, aristis calyce duplo longioribus. F. Nova Zee- landia et insula Paschatis). SPRENGEL, der die Art zu Agrosts brachte, giebt als Standort nur Insula Paschatis an. Die Art liegt in der Sammlung in zwei Formen vor, die beide nicht genau dem Original von FORSTER entsprechen, das mir aus dem Herb. SPRENGEL vorlag. Dieses hat eine ziemlich dichte Rispe, die Hiillspelzen sind 3*/2 bis fast 4 mm lang, die Deckspelze ist fast 2 mm lang, ihre Behaarung ist sparlich, die Granne ist verhaltnismassig kraftig, etwas gekniet, bis 6 mm lang; der Achsenfortsatz ist sehr kurz. Das Ex. no. 687, das der Originalform naher steht (Textfig. 1b) hat 4'/1—5 mm lange Hiillspelzen; die Deckspelze ist 2*/2 mm lang, ziemlich reichlich weisslich be- haart, 4-spitzig, die Granne ist 4'/2 mm lang; die lanzettlich-eiformige Vor- spelze erreicht fast die Lange der Deckspelze, der Achsenfortsatz ist 1 mm lang, lang behaart; Staubblatter 3 mit °/s mm langen Antheren; die Rispe ist sehr locker, mit stark abspreizenden Asten. Die zweite Form entspricht mit ihren kleineren Ahrchen der Form, die WILLDENOW als A. retrofracta beschrieben hat (Textfig. 1c); die Hiillspelzen sind 2'/2 bis 3 mm lang, die Deckspelze 1’/2 mm, ihre Granne 3'/2 mm. Agrostis filiformis ist unter vielerlei Namen beschrieben worden, die DOMIN in Bibl. Bot. 85 (1915) 351—52 zusammengestellt (unter Deyeuxza filiformts). Bei den neueren Autoren geht die Art gewohnlich unter Deyewxza (D. Forsteri Kunth, BENTHAM, FL Austral. VII. 579; CHEESEMAN, Manual New Zeal. FI. 868; D. filiformis Petrie in CHILTON, Subant. Islands New Zeal. II (1909) 474; DomIN |. c.). Trotz des deutlichen Achsenfortsatzes muss aber die Art zu Agrostis gestellt werden, da sie sich-von Deyeuxia (welche Gattung mit Ca/a- magrostis zu vereinigen ist) durch die kurze, breite, oben stark abgestutzte und gezihnelte Deckspelze unterscheidet; die Callushaare am Grunde der Deckspelze sind sehr kurz. Sie ist in Australien, auf Neu-Seeland und Tasmanien ver- breitet und kommt auch auf Hawaii vor. A. filiformis variiert sehr stark inbe- 5 — 20100. The nat. Hist. of Juan Fernandez and East. Isl. Vol. II. 66 CARL SKOTTSBERG Fig. 1. a Spikelet of Stipa horridula Pilger, X 21/2; 6 of Agrostis filiformis (Forst.) Spreng., X 10 and ¢ of A. vetrofracta Willd., * 10. d—h Danthonia paschalis Pilger, X 5; @ spikelet, e flowering glume, back view; f flower, front view; g, 2 paleae, outer and inner surface. — Drawn by C. SKOTTSBERG. zug auf die Grosse der Ahrchen, Behaarung der Deckspelze, Lange der Granne und des Achsenfortsatzes u. s. w.; CHEESEMAN besonders ftihrt von Neu-See- land eine Anzahl von Varietaéten an, doch ist bisher keine befriedigende weitere Gliederung der Art erreicht worden.» The more typical form with the larger spikelets (Fig. 1 b) was found once only, in the swamp of the crater lake of Rano Aroi, c. 400 m, rather abundant THE PHANEROGAMS OF EASTER ISLAND 67 (no. 687). A. retrofacta (fig. 1c) is in the Albatross collection: no. 26! 27! »summit of Rano Kao, probably abundant»: no. 53! I found it in the small crater of Mt. Katiki, 412 m (no. 661) and also in Rano Kao. Area of distribution: Austral., Tasman., N. Zeal., Hawaii, Easter I. A. retrofracta was described from Australia. In ENDLICHER 1. c. p. 158 Deyeuxia Chamissonis (Trin.) Kunth is quoted from Easter Island. To judge from the description in Enum. plant. I p. 242 this must be another form of A. fi//formzs, with a shorter awn. Further, END- LICHER lists Arundo conspicua Forst. = Agrostis conspicua (Forst.) Roem. et Schult. from New Zealand and Easter Island, leg. FORSTER. The latter locality must be a mistake, FORSTER does not mention Easter Island, where the species in question does not occur. It is endemic in New Zealand and a very con- spicuous grass. Dichelachne Endl. 7, D. sciurea (R. Br.) Hook. fil. — Syn. D. crintta and D. scturea var. nova monandra Hack., FUENTES 1. c. Not uncommon. — »Common throughout island, scattered tussocks, 6—7,5 dm» (Albatross nos. 4! 28!); top of Rano Kao (Albatross no. 46], f. ad D. cri- nitam); Mt. Katiki, on the slopes (no. 659, f. ad D. crinitam) and edge of crater, 412 m (no. 1187); Hanga Ho Orno, scattered; Rano Aroi, stony ground in the crater, c. 400 m (no. 688). Dr. PILGER remarks to this species: »Die langbegrannte Form (no. 659) steht der D. crinita (L.) Hook. fil. sehr nahe; die Hiillspelzen sind fein und lang gespitzt und fast gleich, die Granne ist tief auf dem Ricken der Deck- spelze eingefiigt, wahrend bei der typischen D. sczwrea die Hillspelzen kurzer gespitzt sind und das freie Ende der Deckspelze nur kurz ist (nos. 688, 1187). Auch CHEESEMAN (Manual New Zeal. Fl. (1906) 874) erwahnt bei D. crinita eine Uebergangsform zu D. sciurea, die var. intermedia Hack. Die Blite hat nur ein Staubblatt, wie es auch schon R. BROWN fir seine Agrostis sciurea und A. rara angiebt (Prodr. 171). DOMIN setzt fiir D. scewrea den Namen D. micrantha (Cav.) Domin wegen St#pa micrantha Cav. Icon. V (1799) 42 t. 467. Wie schon BENTHAM in FI. Austral. VII. 566 hervorhebt, ist die dem schlechten Habitusbild beigegebene Figur B. ganzlich mit Dzchelachne unvereinbar; auch die Beschreibung stimmt nicht fiir diese Gattung.» Area of distribution: Austral., N. Zeal., Easter I. Danthonia DC. #8. D. paschalis Pilger nov. spec. — Text fig. 1 d—h. Dense caespitosa, innovationibus permultis extravaginalibus, basi squamis rigidis instructis, + erectis et aggregatis, rarius parte inferiore decumbentibus, brevibus, internodis perbrevibus, vaginis magna pro parte sese tegentibus, folio- 68 CARL SKOTISBERG rum lamina angustissima, filiformi-involuta, sulcato-striata, parum vel vix hir- sutulo-inspersa, ad 7—8 cm longa, vagina dense sulcato-striata, ore albido-pilosa; culmi graciles tenues caespitem valde superantes ad 35 cm alti, erecti vel parum geniculati, laeves, nodi glabri; inflorescentia e vagina suprema longe exserta, vagina suprema 3—5 cm longa, lamina ad 3 cm longa; internodium sequens ad 10 cm longum, vagina pluries brevior; inflorescentia depauperata 3—7 cm longa simpliciter racemosa, spiculis paucis breviter pedicellatis com- posita, vel ad basin inflorescentiae ramulus bispiculatus; spiculae ad 6-florae (flore supremo parum evoluto), glumis vacuis circumdatae; glumae vacuae lan- ceolatae, apice acutiusculae vel obtusiusculae, g—10 mm longae, inferior superi- orem basi amplectens, parum vel vix longior, glumae medio tantum nervosae, margo latus enervius, nervi 5, in inferiore et 7, virides, prominentes, approxi- mati; gluma florifera expansa elliptica, cum callo (internodio rhachillae) albido- barbato, I—I,5 mm longo decidua, sine lobis lateralibus 4 mm longa, lobi laterales lanceolati, sensim sat aequaliter angustati et aristiformi producti, arista inter lobos 3 mm _ supra basin glumae orta, pars inferior brunnea, 2—3 mm longa, cum lobis lateralibus + connata et vix vel non torta, pars superior ad 6—7 mm longa, obscure viridis; nervi laterales utroque latere 3, virides, pro- minentes, in lobos laterales excurrentes, gluma basi longe albido-pilosa, praeterea ad marginem utroque latere fasciculo pilorum albidorum insertionem aristae parum superantium instructa; palea brevis, ovalis, apice truncata et leviter bilo- bata, 2,;—3 mm longa, nervi marginales; caryopsis ambitu ovalis, crassa, com- pressa, latere hili applanata ad parum concava, 2 mm longa. Only seen once, very scarce on the slope of Mt. Katiki near the summit (no. 658). There is more material in the Albatross collection; the label runs: »summit of Rano Kao»; »common throughout the island» (no. 51!). I do not think the latter statement is quite correct: even if the season was rather un- favorable during our visit, this grass is too conspicuous to be overlooked, and it was not observed by FUENTES. Dr. PILGER remarks on the affinity: »Die neue Art ist verwandt mit der chilenischen D. chilens?s WDesv., aber doch durch charakteristische Merkmale unterschieden: bei DL. chzlensts ist die Vorspelze verhaltnismassig langer und nach der Spitze zu verschmilert, die Seitenlappen der Deckspelze sind nach aussen zu breiter, fast gefliigelt, die Nerven treten nur schwach hervor und die Rander der Deckspelze sind gleichmassig behaart, wahrend bei der neuen Art 2 Haarbiischel hervortreten, die mit ihren Spitzen etwas uber die Einsatzstelle der Granne hervorragen.» Area of distribution: Endemic. Eragrostis Host. 9. E. elongata Jacq. — Syn. &. diandra Steud., FUENTES |. c. Albatross nos. 8! 24! Slope of Mt. Katiki; round Hanga Ho Orno, on lava fields (no. 645); between the north coast and Puhi, frequent on small red tufa hillocks. Area of distribution: East India, Polynesia. y THE PHANEROGAMS OF EASTER ISLAND 69 (Lepturus repens R. Br. (Australia, Polynesia) is listed by ENDLICHER I. c. (»teste Trin.»). TRINIUS must have based his statement on material collected by CHAMISSO, whose specimens probably belong to one of the Petersburg her- baria. Nobody has seen the plant later and the statement cannot be verified ) Cyperaceae. Cyperus L. 10. C. vegetus Willd. Moist places, edges of crater lakes, etc, — FUENTES. »Enclosure for banana tree (Albatross no. 11!) Mt. Vaintu Rova, humid cave; Rano Aroi, vigorous plants in the crater; crater of Rano Kao (no. 1170); »scattered patches about edge of marsh, also in wet places about the island», (Albatross no. 54!); west slope of Rano Kao; Rano Raraku (Albatross nos. 31! 32!). Area of distribution: Temp. S. Amer., Juan Fernandez, Easter I. *11. C. polystachyus Rottb. Rano Aroi, in the crater, c. 400 m, one specimen seen (no. 686). Area of distribution: Subcosmop. trop.-subtrop. 12. C. cyperoides (L.) Kiikenth. — Syn. C. (Mariscus) Szederzanus Kth.., FUENTES Il. c. N. slope of Rano Kao, scattered (no. 675); between Matave1i and the lepra colony, common in wet patches. Area of distribution: Widely dispersed trop.-subtrop. Kyllinga Rottb. 13. K. brevifolia Rottb. Scattered over the island, more frequent on higher ground. — FUENTES. Mt. Katiki, quite common, copious in the crater, 412 m; Hanga Ho Orno, in crevices between lava boulders (no. 643); »common on plains» (Albatross no. 10!); Rano Aroi, frequent along the pond, c. 400 m; Rano Kao, common (also Albatross no. 52! »abundant throughout the island»); between Mataveri and Rano Raraku; Rano Raraku (Albatross no. 34!). Area of distribution: Widely dispersed within the tropics. Scirpus L. 14. S. riparius Pres] var. paschalis Kiikenth. |. c. — Plate 7. »Culmus 3—4 metralis subteres ad 2 cm crassus, basi vaginis haud reti- culatim fissis obtectus. Anthela decomposita ampla spiculis innumerosis ornata». »S. esp. gigantesca»; »en los crdteres» (FUENTES!); Rano Aroi, smaller; 70 CARL SKOTISBERG Rano Kao, forming an extensive association along the edge of the lake (no. 676; »common rush of the marsh growing 2,5 m high», Albatross no. 55!); Rano Raraku; few, small -stands (also Albatross no. 33!). The variety differs from the ordinary form mainly in its greater size and more compound panicle. The largest plants grow in Rano Kao. The plant from Rano Aroi was called S. r7pfartus by FUENTES, who regarded the large form’ from Rano Kao, which he collected sterile, as another species. The smaller rush was seen by us in a small stream with almost stagnant water, flowing from Rano Aroi, but was not collected, as we intended to return to the same place; from this we were prevented by the sudden departure of the vessel. I do not think that more than one species grows in the island: if FUENTES 5S. rzparius is of the same size and habit as the ordinary plant, var. paschalis may be a local form of the sheltered crater lakes, best developed in the deepest crater, that of Rano Kao. KUKENTHAL does not indicate any characters in the spikelets or flowers to distinguish it from the type. Area of distribution: Calif. and Texas to Fuegia and Falkl., Hawaii, Easter I. Juncaceae. Juncus L. Soe —j- plebcius ke -Br Western edge of Rano Kao (no. 680), c. 300 m. — New for Easter Island. The specimens are typical. According to BUCHENAU in Pflanzenreich, Junc. p. 115, the flowers probably always are cleistogamous. This is also the case with the majority of the flowers in my material; they have three stamens with anthers coherent with the very small stigmas. Only in very few cases normal open flowers were seen, with six fully developed stamens and long stigmas. Area of distribution: Austral., Tasman., N. Zeal., Easter I. Polygonaceae. Polygonum L. 16. P. acuminatum H. B. K. — Plate 8, fig. 1. Common in the three large craters. — »Comun en los 3 crateres» (FUEN- TES!); Rano Aroi, few and small patches; Rano Kao, abundant along the lake (no. 679; »very abundant in marsh», Albatross no. 56!); Rano Raraku (also Albatross nos. 35! 36!, specimens with narrower leaves). FUENTES remarks |. c.: »Aunque difiere un poco de la forma tipica sud- americana, respeto la opinion de los eminentes botanicos doctores ENGLER 1 STAPF sobre su identificacion.» THE PHANEROGAMS OF EASTER ISLAND 71 I have compared my specimens with the rich material of this highly poly- morphic species in Kew; they differ a little from most other plants in the shorter internodes, often completely covered by the vaginae, whose marginal bristles or setae are very long; the nut seems to be a trifle larger than in most other specimens. Still, plants from Uruguay, collected by E. GiBERT, are very like the insular ones. On the whole, my specimens agree with « Huméoldtiz Meisn. F]. Brasil. V: I p. 14, tab. 4 fig. I: only, the leaves are more acuminate and the setae of the ochrea considerably longer. To judge from the description, 4 set#igerum Meisn. |. c. p. 15 (P. setigerum Wedd.) must be very similar. The anthers are described as included in « and exserted in 4, but this is of little consequence, as the species, to judge from my specimens, is heterostylous. Area of distribution: S. Amer., Columb.-Uruguay (not in Chile). S. Afr. (an aberrant form?). — FUENTES remarks: »Ameér. trop. S. de Australia», but Australia must be a misprint for Africa. Chenopodiaceae. Chenopodium L. 17. Ch. ambiguum R. Br. — Ch. glaucum, FORSTER |. c.; var. paschale FUENTES |. c. Comp. J. MurRR in Allg. bot. Zeitschr. 12 (1906) p. III. On the sea shore. Hanga Ho Orno (no. 662); rocky beach near Hanga Piko; Islet Motu Nui (no. 1194). My specimens were compared with the material in Herb. Kew and Upsala. Some authors regard Ch. ambiguum as identical with C. glaucum L., while others retain it as a variety. I prefer to keep it as a species, mainly on ac- count of its geographical distribution, but also because I have not seen inter- mediate forms. FUENTES probably had but a scanty material of Ch. glaucum for comparison, for he does not give any character serving to distinguish his insular form from this. I do not understand what he means by »spiculas race- mosas muy flojas», for all the specimens I have seen have short and more or less dense inflorescence. Area of distribution: Austral., Tasman., N. Zeal., Easter I. Nyctaginiaceae. Boerhaavia L. *18. B. diffusa L. forma. Tongariki, near the sea shore (no. €698)..— New for the island. The Easter Island Boerhaavia is decumbent, nearly glabrous, with linear- lanceolate or almost ovate leaves, obtuse and minutely apiculate, and whitish below. The inflorescence is mostly a simple axillary umbel of 3—4 flowers, with 72 CARL SKOTTSBERG =< a peduncle shorter or longer than the supporting leaf, rarely a poorly devel- oped panicle. It is probably the same form as understood by TRIMEN, FI. Ceylon III p. 390, DRAKE DEL CASTILLO, Flore Polyn. Franc. p. 158, etc. under B. diffusa. But according to CHorsy in DC, Prodr. XIII: 2 p. 452 this has the leaves green on both sides, and my plants agree better with his de- scription of B. repens L., but typical repens has only been reported from Africa. HEIMERL in ENGLER & PRANTL, Nat. Pflzfam., regards 3. diffusa as a col- lective species, of which B. repens is one of many forms. Typical B. dzffusa is a large plant with a well developed panicle, but there are specimens in herbaria recognized as diffusa but with single axillary umbels and also inter- mediate forms. The leaves are sometimes conspicuously glaucous beneath. The long peduncles in my form remove it from typical B. repens. Area of distribution: Tropical Asia, Austral., Polyn., Afr. Aizoaceae. Tetragonia L. *19. T. expansa Murr. On the coast near Tongariki (no. 700). — First record for Easter Island. Area of distribution: Japan, Austral., N. Zeal., Polyn., Juan Fern., extratrop. S. Amer. (Sesuvium portulacastrum L. is reported, without any indication of the source, for Easter Island by HEMSLEY ]. c. Mr. HUTCHINSON kindly informs me that there is no specimen from this place in the Kew Herbarium, nor has it been quoted by any other author. Its occurrence is, of course, quite possible, but requires further evidence.) Cruciferae. Nasturtium Rk. Br. *20. N. sarmentosum (Sol.) O. E. Schulz. Mt. Vaintu Rova, moist floor of a small cave; scanty and sterile material in winter state (no. 652). — New for Easter Island. Area of distribution: Austral., Polynesia, Hawaii. Leguminosae. Caesalpinia L. 21. C. bondue (L.) Roxb. »Encontré dos ejemplares en el centro de la isla»: FUENTES |. c. — Crater of Rano Kao, rare (no. 696). THE PHANEROGAMS OF EASTER ISLAND 73 It is of course not at all certain, that this belongs to the original flora of the island, the more as the tradition, also quoted by FUENTES, tells that it was introduced by the first settlers, long before the discovery of the island by the Dutch. However, its faculty for dispersal with the ocean currents is a well known fact, so that nothing prevents us from believing that it reached Easter Island without human aid. According to GUPPY, Observations of a naturalist in the Pacific I, the seeds of inland stations have lost their buoyancy. FUEN- TES states that C. donduc is on the verge of extinction in the island; I can neither confirm nor deny this. It is not in cultivation -—— if it ever was —, and no plants were found along the sea shore. Guppy (Il. c. p. 185) gives a good summary of the two generally re- cognized species C. donduc and bonducella; he has found intermediate forms and comes to the conclusion that all may belong to one complex taxonomic unity. The Easter Island plant is no »typical» donduc. In the size of its flowers it agrees with this and it has no trace of stipules. But the leaflets are not un- equal at the base, as has been stated for C. donduc (comp. GRISEBRACH, Flora Brit. W. Ind. Isl. p. 204), but equal as in donducella. According to HELLER, Plants of the Hawaiian Islands (Minn. Bot. Stud. 1897, p 831), C. bonduc is glabrous and has broader, blunt leaflets, while C. donducella is pubescent, with comparatively narrow and acute leaflets. The leaflets in my plant answer better to HELLER’s C. donduc, but they are clearly pubescent. Bracts are totally absent, even with the young buds. The colour of the seed is greyish as in donducella, not yellowish as in donduc, but GUPPY shows that this char- acter is not a reliable one, and TRIMEN, Fl. Ceylon II p. 98 calls the seeds of C. donduc greyish green, which answers well to the observations of the pre- sent writer. Area of distribution: Trop. Asia, Austral., Polyn. Sophora L. 22. S. (Edwardsia) toromiro (R. A. Phil.) Skottsb. — Edzwardsia toro- mivo R. A. Philippi, Bot. Zeit. 1873 p. 737; S. tetraptera auctt. non J. Mill. nec Ait. — Plate 9; Text fig. 2 a—c. Crater of Rano Kao, very rare on the steep slope, only one small tree seen by us, growing between large boulders (no. 688; also reported from this locality by FUENTES |. c.). Edwardsia toromiro of PHILIPPI is hardly more than a name, as no de- scription ever appeared; it was characterized by PHILIPPI as possessing a wood different from that of the Chilean species, which generally goes under the name S. tetraptera. FUENTES did not consider the Easter Island plant to be a separate species. A description follows. Arbor pumila (specimen unicum ab auctore visum 190 cm altum), tor- -tuosa; cortex junior saltem rufo-brunneus, longitudinaliter sulcatus. Folia ver- sus apices ramulorum conferta, plerumque 40—55 mm _ longa, 8—o- (rarius ad 12-) paria, rhachide tenui dense albosericea; foliola brevissime stipitata, stipite albo-sericeo, ovato-elliptica (sed foliolum terminale plerumque subobo- Tee CARL SKOTTSBERG vatum), basi cuneato-rotundata, apice truncata et minutissime emarginata, peni- cillo brevi albo ornata, textura tenui, supra laete viridia, parce pilosa, subtus glaucescentia, pilis albis praecipue secus medianum densius_ vestita. Legumen longe pedicellatum, ad 10 cm longum, ad 7-locellatum, pro sectione typicum quadrialatum alis angustis circ. I,5 mm altis transverse striatis, parie- tibus sat tenuibus, dilute bruaneis. Semina sat parva, 4,2—5 mm longa et 3,5—4 mm lata, ovoideo-globosa, basi et interdum apice apiculata, fulva, um- bilico obscuro notata. As far as I can understand, this ta a very good species. It differs in several respects from both Chilean and New Zealand specimens of .S. ¢etra- ptera. Yhe pubescence is whitish even in the youngest parts, the leaflets are O¢ Ys OO Q 00 Fig. 2 a—c Sophora toromiro (R. A. Phil) Skottsb. @ leaf, nat. size; 6 leaflet, lower surface, X 2; ¢ four seeds, nat. size; d Lycium sandvicense A. Gray, flower. x 5. — Drawn by G. SKOTTSBERG. of a yellowish or otherwise bright green colour above, and very thin, papy- raceous; they are also less numerous to the midrib. Finally the seeds are smaller, more globose and yellow. All the forms from Juan Fernandez have quite unwinged pods and larger, oblong-obtuse seeds of a dark castaneous colour, very different from those of the Easter Island species, which is one of the more interesting endemic plants. S. toromiro will probably soon be extinct. Already at the time of COOK and FORSTER this tree, called A/zmosa by the latter (Voyage I p. 568) was scarce, though he observed small shrubberies in some places (I. c. 578, 592). Concerning the supposed forests, see my paper in Vol. I (No. 1) of this work. Although, to judge from the famous wood-carvings of the islanders, the stem formerly attained a thickness of a couple of dm at least, only small stunted specimens were left in 1870 (see PHILIPPI, |. c. p. 744). FUENTES calls the tree »muy escasa», and I have not the slightest doubt that the interesting toromiro is on the verge of extinction, which means the disappearance of a species, important from geographical as well as from ethnological point of view. Area of distribution: Endemic. THE PHANEROGAMS OF EASTER ISLAND 75 Euphorbiaceae. Euphorbia L. OAs otave ALEY ta, Le! Slope of Mt. Katiki (no. 653). »Abundant, rocky places, near La Pérouse Bay (= Hanga Ho Orno), Albatross no, 16! Area of distribution: Trop. Amer., Polyn., etc.; trop. subcosmop. A common tropical weed, perhaps not truly indigenous in Easter I. 24. E. serpens H. B. K. — E£. ovalifolia Engelm., FUENTES |. c. Common on rocky and stony ground all over the island, near the sea (no. 1196) or inland (no. 638). »On rocks and barren places, abundant» (Alba- tross nos. 14! 15)). Area of distribution: Trop. America, Polyn., also reported from the Indo-Malayan region. A widespread weed, possibly not indigenous in Easter I. Umbelliferae. Apium L. 25. A. prostratum Labill. — Syn. A. graveolens L., FORSTER, De Plantis esculentis insularum oceani australis commentatio botanica. Berlin P7GG sp. ‘07, Rocky beach and caves near Hanga Piko (no. 697). — My specimens agree very well with A. australe Thouars, which, with the material at hand, I cannot distinguish from A. prostratum: in fact, I believe they are identical. Area of distribution: Magellan coasts, Falkl., Tristan d’Ac., Austral., N. Zeal, and outlying islands, Easter I. Primulaceae. Samolus L. 26. S. repens (Forst.) Pers. On the sea shore:. FORSTER, FUENTES, — N. of Hanga Piko, not un- common (no. 701); near Tongariki (no. 699). Area of distribution: Coasts of temp. S. Amer., Austral., Tasman., N. Zeal., Easter I. 76 CARL SKOTISBERG Gentianaceae. Erythraea Borkh. *o7, KE. australis R. Br. »Abundant, rocky plains, La Pérouse Bay (i. e. Hanga Ho Orno); Alba- tross Dec. 16 (no. 19!) — Had disappeared long before our visit, but seedlings, apparently belonging to this species, were observed. Area of distribution: Australia; Easter I. Convolvulaceae. Calystegia R. Br. #28. G.-sepram (L.) Ri .Breforma: In the crater of Rano Aroi, c. 400, among large boulders. — A rather narrow-leaved form, probably belonging to this species; only sterile specimens. Area of distribution: Widely dispersed in temperate regions, also in Australia, Tasmania, New Zealand, etc. Ipomaea L. 29. I. pes caprae (L.) Roth. Apparently rare. Near Vaihu (I. affinis pes caprae», FUENTES l. c.); Hanga Ho Orno, stony beach on the water’s edge (no. 634). Area of distribution: Widely dispersed on tropical shores. Solanaceae. Lycium L. *20, L. sandvicense A. Gray. — Plate 8 fig. 2; text fig. 2d. Rocky shore below the western slope of Rano Kao (no. 644). — Rather an interesting discovery. The specimens are exactly like Hawaiian ones in the Kew Herb., and also agree very well with the description in Proc. Amer. Acad. VI (1863) p. 44. The leaves show numerous whitish pustules, irregular vesicles in the mesophyll, filled with salt. Area of distribution: Hawaii, Easter I. THE PHANEROGAMS OF EASTER ISLAND 1h 2. Plants now occurring in a wild or semi-wild state but reported to have been intentionally introduced by the first native immigrants. There is a tradition that all the plants of economic importance were in- troduced and brought into cultivation by the party of the chief Hotu Matua, who was the first to reach Easter Island. Of the few plants mentioned below this may be true, although proofs are wanting, but, on the other hand, it is not altogether impossible that species such as Cordyline terminalis, Triumfetta semitriloba or Thespesia populnea belonged to the original flora, and that the immigrants were acquainted with their useful qualities and started to propagate them. KNOCHE (lI. c. p. 10) regards all the plants used for industrial or medicinal purposes as only naturalized and not wild, but when he includes Sophora toromiro among these he is wrong, for we have seen that it is an endemic species, and its nearest relatives, the Chilean and New Zealand species of this genus, do not belong to the widely dispersed oceanic plants, but have a very restricted geographical range; nor have they ever been in cultivation as far as I am aware. It is of course not quite impossible that the islanders, seeing that the precious ‘ovomzro was rapidly vanishing, tried to propagate it, but there is no reliable evidence that they did so. Liliaceae. Cordyline Comm. C. terminalis Kunth. In the crater of Rano Kao, oa the south side between large stones, c. 125 m, scarce (no. 693); see also FUENTES 1. c. Area of distribution: Indo-Mal. region, N. Guinea, Austral., Polyn., N. Zeal., formerly cultivated by the Maoris. Moraceae. Broussonetia Vent. B. papyrifera Vent. »Crater of Rano Kao, abundant, but mature trees not numerous»: Alba- tross no. 58! — We found a number of trees on the stony slope near the pond (no. 670); they agree with the ordinary plant as represented in the Kew Herb. F. foliis subtrilobis grosseserratis magis scabris: with the former (no. 1147). In the Kew Herb. there is the same or closely allied forms from Oahu (SEEMAN, Macrak), Fiji (HERALD), Nagasaki (OLDHAM) and Siam. _Inter- mediate forms occur also in Easter Island. FUENTES calls the mahute »B. aff. papyracea Vent.» adding: »su clasi- 78 CARL SKOTTSBERG ficacion es dudosa, difiere algo de la forma tipica», but this applies only to the aberrant form mentioned above. Area of distribution: S. E. Asia, Polynesia. DRAKE DEL CASTILLO says |. c. p. 193 that Broussonetia is »sans doute originaire dans la Chine». Tiliaceae. Triumfetta L. T. semitriloba L. Rano Kao, with Broussonetia, rare (no. 671). Listed by FUENTES with a query. His plants, which I have seen, as well as mine, are sterile, but I think the determination is correct. This is the jaujau of the islanders, used for making cordage; it was pointed out to us as the true plant for this purpose and we obtained large pieces of string freshly prepared. Area of distribution: Trop. Asia, Polyn. and Amer. Sapindaceae. Sapindus L. S. saponaria L. FORSTER and FUENTES. Malvaceae. Thespesia Carr. T. populnea (L.) Carr. FORSTER, Prodr., reports Hibiscus bacctferus, which is considered to be identical with Zhespesia populnea, from Easter Island. No later collector ever observed it there, so it may have disappeared from the island. Solanaceae. Solanum L. S. (Morella) insulae-paschalis Bitter nov. spec. (S. 2zgrum L. coll.). »Herbaceum, maleolens; partes novellae non jam satis evolutae albide subcanescentes; rami superiores diam. 3—4 mm, obtuse angulati, lineis decur- rentibus manifeste prominentibus instructi, pilis simplicibus pluricellularibus (mem- branis tenuibus in statu sicco +ve collabentibus) apice in glandulam minutam subglobosam vel ellipsoideam vel rarius in cellulam diaphanam ceteris similem subacutam exeuntibus patentibus sordide albidis 1 —1,5 mm longis densis ob- THE PHANEROGAMS OF EASTER ISLAND 79 tecti; internodia 2,;—8 cm longa; folia superiora partim false geminata, tunc inaequalia; petioli 0}6—2 cm longi, ad laminas versus sensim latius alati, eodem indumento denso quo rami vestiti; laminae oblique ovati-ellipticae vel ovatae, infra medium latissimae, basi rotundatae vel + ve late cuneatim in petiolum alatum abeuntes, ad apicem versus magis sensim angustatae, apice ipso obtusae vel obtusiusculae, cr. 4,5: 2,5, 5,5: 3,7, 8: 4,5 cm, margine leviter et obtuse repandi dentato, dentibus in utroque latere I—3 parum prominentibus obtusis, plerumque infimis tantum unicis in utroque latere distinctioribus; laminae mem- branaceae, utrinque sordide virides, utrinque pilis simplicibus pluricellularibus inaequilongis (in venis majoribus utrinque longioribus densioribusque, in meso. phyllo brevioribus sparsioribusque tamen hic quoque satis crebris) apice acutius- culis partimque subglandulosis partim eglandulosis patentibus praeditae; vena media, venae later. prim. in utroque latere 5—6 curvatim ascendentes et venae later. secund. complures subtus prominentes; inflorescentiae laterales, a foliis remotae, simplices, 4—5-florae; pedunculus cr. 8— (in statu fructifero) 17 mm longus, rhachis brevis, 2— (tandem) 4 mm longa; flores ultimi dense subum- bellatim secuti; pedicelli 6—7 mm longi, in statu florifero erecti, in statu fructifero vix longiores, deflexi; calyx campanulatus, 2—2,5 mm longus, apice 3 mm diam., profunde in lobos ovati-ellipticos 1—1,5: 0,6 mm obtusiusculos partitus, extus sicut pedunculus, rhachis et pedicelli pilis simplicibus pluricellu- laribus apice subglandulosis vel eglandulosis crebris obsitus, intus glandulis minutis breviter stipitatis crebris praeditus; corolla stellata, diam. 11 mm, pro- funde in lobos 5 late lanceolatos vel ovati-lanceolatos 4—5: (basi) 2 mm acuti- usculos extus et in margine pilis simplicibus brevibus paucicellularibus acutis crebris obsitos intus praeter apicem subcucullatum pilis brevibus densiusculis instructum glabros partita; stamina I—1,2 mm supra corollae basim inserta; filamenta pro floris parvitate longiuscula, 1— (tandem) 1,8 mm longa, intus pilis longiusculis simplicibus pluricellularibus apice acutis vel partim sub- glandulosis densis obtecta; antherae breviusculae, late ellipsoideae, cr. 1,3: 0,8—1 mm, basi cordatae, apice emarginatae, poris introrsis apicalibus obli- quis; ovarium subglobosum, 1 mm diam., glabrum; stylus stamina paulum superans vel illis fere aequilongus, 3 mm longus, rectus, gracilis, paulo supra basim glabram cr. %— } longitudinis pilis simplicibus pluricellularibus acutis vel apice partim anguste subglandulosis patentibus densis obtectus, apice glaber; stigma styli apice manifeste crassius, capitatum, obtusum; calyx fructifer paulum auctus (explanatus diam. 7 mm), ejus lobi 2: (basi) 1,5 mm; bacca globosa, diam. 7—8 mm, ut videtur pallide lutea; granula sclerotica desunt; semina oblique reniformia, lenticulariter applanata, 1,5: I: 0,3 mm, minute reticulata.» Hanga Ho Orno, in an old, dry well (no. 663); in the crater of Rano Kao, c. 125 m (no. 695). »Eine Kleinart aus der Verwandtschaft des S. xigrum L.., deren wichtigste Merkmale auf der Behaarung (ungleich lange, einfache, mehrzellige Haare, die zum grossen Teil in oft undeutliche schmale, seltener kugelige, winzige Dri- senképfchen endigen, sowie auf den schliesslich ziemlich langen Staubfaden und den kurzen Staubbeuteln beruhen. Der iible Geruch riihrt offenbar von der driisigen Behaarung her.» FORSTER tells that he found So/anum in the plantations (Voy. I p. 578); 80 CARL SKOTTSBERG according to LANGLE, a member of LA PEROUSE'’s party, it was cultivated as a medicinal herb (see KNOCHE |. c. p. 13). FUENTES lists »So/anum sp. (?)» adding »solo conozco por referencias». I have no doubt that all these state- ments refer to the species found by us and described above by Professor BirterR. This well known expert on the genus So/anum has described a num- ber of Morellae of S. xzgrum type, endemic in the southern seas; in a letter to the writer he points out that »die meisten Morellen sind Menschenbegleiter und durchgangig in der Nachbarschaft des Menschen ansassig». 3. Species unintentionally introduced with the human traffic after the discovery by the Europeans in 1722. Sorghum halepense (L.) Pers. var. effusum (Hack.) Pilger. — Common in enclosures protected from sheep and cattle: Hanga Ho Orno (no. 656); fre- quently seen between Rano Aroi and Mataveri and between the latter place and Rano Raraku. Introduced from Tahiti acc. to FUENTES I. c. p. 5; brought from Tahiti, not very long ago and possibly on purpose, all acc. to Mr, Ep- MUNDS, the manager in the island. * Digitaria horizontalis Willd. — Hanga Ho Orno, in crevices between stones (no. 649, also Albatross no. 6!); Rano Kao, not uncommon (no. 667); islet Motu Nui (no. 1200). Digitaria sanguinalis (L.) Scop. — FUENTES |. c. — Widely dispersed over the island (nos. 646, 1192). [Panicum filiforme in FORSY. Prodr. no. 38 is called Paspalum filiforme Sw. by ENDLICHER 1. c. p. 157. These two names correspond to quite different plants, acc. to Prof. PILGER; the latter is a West Indian species, the former is now called Digitaria filiformis Ell. and found in N. and S. America. Possibly the Easter Island plant is only a form of D. sanguenals.] Cenchrus echinatus LL. — FUENTES 1. c. — »On hill side, about 15—60 m alt.» (Albatross no. 3!); Hanga Ho Orno (no. 640). Bromus unioloides H. B. K. — FUENTES 1. c. — »Foot of cave just inside Rano Kao» (Albatross no. 50!); Rano Kao, here and there in the crater (no. 674); islet Motu Nui (no. 1107). Cynodon dactylon (L.) Pers. — FUENTES I. c. — »Abundant everywhere, the common grass» (Albatross no. 5!); Hanga Ho Orno (no. 650); Mataveri; Rano Kao; islet Motu Nui (no. 1188). Eleusine indica (L.) Gaertn. — FUENTES |. c. — Mataveri (no. 985). * Hordeum murinum L. — »Inside of Rano Kao, no other example seen» (Albatross no. 47!) * Sixyrinchium micranthum Cay. — »Abundant in rocky plains of low eleva- tions» (Albatross nos. 12! 30!) — A Peruvian species, also introduced into Australia and New Zealand acc. to CHEESEMAN. * Portulaca oleracea \.. — Growing near beach, not abundant» (Albatross no. 17!); Hanga Ho Orno, not uncommon (no. 636); Mataveri (no. 1198). This may equally well be regarded as indigenous; still, GUPPy thinks it has been introduced into several places in the Pacific, f. inst. to Hawaii (I. c. p. 552). THE PHANEROGAMS OF EASTER ISLAND 81 *Cerastium glomeratum Thuill. — Very minute specimens, Mt. Vaintu Rova (no. 1199). *Polycarpon tetraphyllum L. — »Growing on rocks, abundant» (Albatross no. 13!). *Coronopus didymus (L.) Sm. — Hanga Ho Orno, caves in the coast rocks (no. 635). Crotalaria Saltiana Andr. — »Abundant near village» (Albatross no. 65!); along the road from Hanga Roa to Mataveri (no. 681; also FUENTES l. c.). Trifolium spec. — Sterile. Hanga Roa. Phaseolus spec., »affinis vulgaris L.» — Rano Kao, FUENTES l. c. Oxalis corniculata L. — Dispersed over the entire island in stony places. — »Abundant about rocky walls etc.» (Albatross no. 18!); Hanga Ho Orno (no. 637); Mataveri; Rano Kao. Malvastrum coromandelianum (L.) Garcke (»coromandelicum»). — FUENTES lc. s. n. MW. tricuspidatum A. Gray. — »Crater of Rano Kao, scattered patches (Albatross no. 61!); Mataveri (no. 692). *Malva spec. — A weed in the garden of Mataveri. Sida paniculata L. — A weed in the garden of Mataveri (FUENTES, I. c.). Verbena litoralis H. B. K. — FUENTES |. c. — »On hill side, 60 m alt. La Pérouse Bay» (= Hanga Ho Orno): Albatross no. 20! Very common along the shore, but also inland in scattered specimens, in the craters and in the cultivated fields (no. 682). Plantago major L. — FUENTES 1. c. — »Common near landing, Cook Bay (= Hanga Roa, Albatross no. 64!): Plantago lanceolata L. — Mataveri (also FUENTES I. c.). *Vomordica charantia L.. — Mataveri (no. 683). Possibly escaped from old cultivations. Centaurea melitensis 1. — CHAMISSO acc. to ENDLICHER Il. c. p. 107, s. n. C. apula Lam. Not found by later collectors. *Ageratum conysoides LL. — In the crater of Rano Kao (no. 669). A common tropical weed. Erigeron linifolius Willd. — Hanga Roa, »only example seen (Albatross no. 63!) Common on stony ground, f. inst. Hanga Ho Orno (no. 642). — FUENTES s. n. E. canadensis L., which it is not; VIERHAPPER confirms the determination of my specimen. Bidens pilosa L, — FUENTES 1. c. »On hilltop, 60—90 m alt., sparse» (Albatross no. 21!); Hanga Ho Orno (no. 567); Mataveri, in the garden (no. 1193). Galinsoga parviflora L. — FUENTES |. c. — Mataveri (no. 684). *Hypochoeris radicata L. — Rano Aroi. Sonchus oleraceus 1. — FUENTES 1. c. — »Summit of Rano Kao, but scattered examples everywhere on island» (Albatross no. 62!); Mt. Teatea; Rano Aroi; Mataveri (no. 1195); Ratio Kao. t 6 — 20100. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 82 CARL SKOTTISBERG Appendix. Ferns collected by the Albatross Expedition. Asplenium adiantoides (L.) C. Chr. var sguamulosum C. Chr. — » Crater ot Rano Kao, near foot, scarce» (no. 41!). Asplenium obtusatum Forst. — »Scattering, near shore, on rocks» (no. 1!); »crater of Rano Kao, not abundant» (no. 45!) Dryopteris parasitica (L.) OK. — »About inside of crater of Rano Kao» (no. 37!); »in cave at summit of Rano Kao» (no. 38!). Microlepia strigosa (Thunb.) Presl. — »About rocks in all places where moisture is available» (no. 2!); »from crater of Rano Kao, abundant, the com- mon fern of the island» (nos. 39! 42!). Polypodium phymatodes L. — no, 22! »summit of Rano Kao» (no. 43!); »along large rocks near bottom of crater of Rano Kao» (no. 44)). Vittaria elongata Sw. — »Summit of Rano Kao, scarce» (no. 40!). The flora of Easter Island is very poor. As in similar cases, it is some- times difficult to tell if a certain plant is indigenous or not. If the old authors are trustworthy, cultivation once extended over a large part of the ground, so that several species, members of the original flora, may have become extermin- ated, and we have seen that Sopkora toromiro probably goes to share their fate. The small copses said to have existed in the time of COOK may have sheltered some herbs that disappeared later. Unfortunately, there are no high mountains, but the host of weeds is able to sweep over the entire area. Many introduced species play an important part in the plant associations of the island, which I shall describe briefly in another paper. Later, I shall also discuss the affinities and probable origin of the flora; this time I find it proper to confine myself to the following remarks. Of wild phanerogams, 30 have been reported, out of these 12 for the first time. Four are considered to be endemic (Axonopus paschalis, Stipa horrt- dula, Danthonia paschalis, Sophora toromiro); three (Cyperus vegetus, Scirpus riparius, Polygonum acuminatum) are common to Easter Island and America but (with the exception of Scirpus, reported from Hawaii) not found in other parts of the Pacific; the majority or 23 species are Australian and Polynesian, many of them wide-spread tropical or subtropical plants. If we add the 12 ferns, of which 2 are endemic and 10 Polynesian or pantropical, the total number of indigenous vascular plants amounts to 42, of which 6 are endemic, THE PHANEROGAMS OF EASTER ISLAND 83 3 American and 33 Polynesian, pantropical etc. Of the endemic species, Dryopteris Esptnosat, Axonopus paschalis and Danthonia paschalis are consid- ered to have their nearest relatives in South America. Thus, it is of interest to find that there is a small »American element» in the flora of Easter Island, distant some 2,000 miles from the American shores. 84 Plate 6. Piater7e Plate 8. Plate 9. CARL SKOTTSBERG Explanation of plates. 1. Paspalum Forsterianum Fligge, flowering shoot (right) and innovation (left). 2. Axonopus paschalis Pilger. 3. Stipa horridula Pilger. — About 1/3 nat. size. Scirpus riparius Presl var. paschalis Kiikenth., top of large plant, 7/2 nat. SIZE. 1. Polygonum acuminatum FH. B. K. 2. Lyctum sandvicense A, Gray (base upwards). — '/2 nat. size. Sophora toromiro (R. A. Phil.) Skottsb., fruiting branch, */s nat. size. Nat. Hist. Fuan Fernandes and blaster [sl. Vol. 11. PLATE 6. PU DIDGEY (Ce SMOUstere ‘1. Paspalum Forsterianum Fliigge. 2. Axonopus paschalis Pilger. 3. Stipa horridula Pilger. Nat. Hast. Fuan Fernandez and Easter Isl. Vol. Ll. PLANES Photo by C. Skottsberg Scirpus riparius Pres] var. paschalis Kiikenth. Nat. Hist. Fuan Fernandez and Faster Isl. Vol. 11. PLATES. meee ey C- Shatisbere 1. Polygonum acuminatum H. B. K. 2. Lycium sandvicence A. Gray. Ieiby MINE) (6) Nat. Hist. Fuan Fernandes and Faster Isl. Vol. I. Sophora toromiro (R. A. Phil.) Skottsb. Photo by C. Skottsberg 6. Freshwater Alge from Juan Fernandez and Easter Island. By K. MUNSTER STROM. With 1 textfigure, The material, upon which the present contribution to the flora of Juan Fernandez and Easter Island is based, was collected by Professor Dr. C. SKOTTS- BERG during the Swedish Expedition to the Pacific 1916—1917. Numerous samples were collected; partially they were preserved in bottles, partially they were dried up. On the whole, I have examined 28 samples preserved in chemicals and about 25 dried ones. Only very few came from Easter Island. The collection was very kindly submitted for my inspection by Prof. SKOTTSBERG and for- warded to Christiania. Very little is previously known concerning the freshwater flora of these is- lands, but unfortunately the present collections were so poor that only twenty- five species were observed, among them only two Desmids. I therefore very much regret being unable to give any account of the geographical relationship of the freshwater microflora to that of other coun- tries. A thorough examination of the Freshwater Algze of these Islands would no doubt be of interest, as it may help to supply the phyto- and zodgeographical results, made out from the examination of the higher plants and of the fauna. Probably, the extreme poverty of the samples is partly due to the fact that they were collected by a non-specialist, but there are reasons to believe that the Islands are not very rich in Freshwater Algee. This may be due to many causes (there does not exist a single lake or pond in Juan Fernandez); perhaps the lonely position of the islands is of some significance. I am, however, inclined to believe that at least a greater number of sub- aérial Algaee may be found by future explorers as the climate of the higher parts of Juan Fernandez is very damp. The majority of the Alga are wholly ubiquitous species, even the two Desmids are widely distributed forms. I have, however, mentioned the area of 86 K. MUNSTER STROM distribution when dealing with each species in the following account, in con- formity with the other contributions to this work. Only in a few cases I have mentioned the names of the collecting-places proper, as they are of very little interest when such scarce material is investiga- ted, but often I have made use of Prof. SKOTTSBERG’s notes on the conditions under which the Algz grew. The number of species may be apportioned as follows. Chlorophycez V2. PROTOCOCCALES I. CONJUGAT 5 Myxophycez 2. OEDOGONIALES i¢ COCCOGONE 2 CLADOPHORALES © 3. HORMOGONE oe CHATOPHORALES 2 Total 25. As to be seen from this, the number of species in the Myxophycez was relatively large. The ecological groups of Algz, which were best represented were the Algz of wet and dripping rocks, and a few strictly subaérial ones. Finally I wish to express my best thanks to Professor Dr. N. WILLE for kindly giving me his opinion on some critical species. Systematic account of the species observed. Class Chlorophycee. Order Conjugate. Family Desmidiacee. Genus Cosmarium Corda. I. Cosmarium subspeciosum Nordst. Norpstepr: Desm. Arctoe. W. & G. S. Wesr: British Desmidiacez Vol. III. After due consideration I have referred the specimens observed to this species. Many of them closely resemble Cosmarium binum, but in no case could I observe the characteristic and indispensable markings. The nature of the central tumour furnished with the vertical granulate ridges in Cosmartum binum must be regarded as the essential distinguishing feature. Moreover, the measurements and the number of crenations on each semi- cell are in accordance with the typical form of Cosmarium subspectosum. The crenations are by no means regularly binate granulated. All these characters serve to distinguish the species. However, the specimens observed at a first glance so much resemble cer- tain forms of Cosmarium binum that the two species may be easily confused, when dealing with such forms as the present. FRESHWATER ALGA FROM JUAN FERNANDEZ AND EASTER ISLAND 87 Size: long. 41—46 p; lat. 29—33 p; lat. isthm. 12—15 wp; crass. 24 yp. Masatierra: among mosses on wet rocks in Pangal. Area of distribution: probably .ubiquitous, known from Europe, Arctic Countries, Asia, Africa, North and South America. 2. Cosmarium humile var. striatum (Boldt) Schmidle. “=. ScumipLE: Beitr. Alpin. Algen; Cosmarium striatum Bo.ipr: Siber. Chloroph. ; W. & G. S. West: British Desmid. Vol. III. This Desmid is represented with a few specimens in the same sample as the preceding. The forms are typical of the variety. Srrere longap 5. 05 lat..12 tj lats,isthm. 4 ps Masatierra. Area of distribution: Europe, Asia, North and South America, but pro- bably even more widely distributed. Family Zygnemacee. Genus Zygnema Ag. 3. Zygnema sp. Sterile specimens of Zyguema are present in a couple of samples. Masatierra: in a brook. Masafuera: in a brook with slowly running water. Distribution of the genus: ubiquitous. Genus Spirogyra Link. 4 & 5. Spirogyra sp. At least two sterile species of Sfzrogyra were observed. The one was a large form 70—8o y, broad, the other a slender species. Masatierra: Colonial Valley; Masafuera; Easter Island. Distribution of the genus: world-wide. Order Oedogoniales. Family Oedogoniacee. Genus Oedogonium Link. 6. Oedogonium sp. Specimens of Oedogonium are widely distributed in all the islands, but invariabily they were sterile. Masatierra; Masafuera; Easter Island. Distribution of genus: world-wide. 88 K. MUNSTER STROM Order Cladophorales. Family Cladophoracez. Genus Cladophora Kuetz. 7. Cladophora fracta (Dillw.) Kuetz. Kuetzinc: Phycol. German.; Conferva fracta DiLLwyn: British conferve. Some of the samples examined consist of this species, which together with Cladophora glomerata is among the most widely distributed Algz in the world. Between the branches of some specimens there is a rich vegetation of Diatoms. Masatierra. Area of aistribution: ubiquitous. 8. Cladophora sp. Masatierra. Undeterminable C/ladophora-species are present in a couple of collections. Distribution of the genus: ubiquitous, both freshwater and marine. Genus Rhizoclonium Kuetz. g. Rhizoclonium hieroglyphicum (Ag.) Kuetz. Kuetzinc: Phyc. Gener.; Conferva hieroglyphica AGARDH in Flora 1827. Masatierra: This common species occurred in a few samples. Area of distribution: ubiquitous. Order Cheetophorales. Family Trentepohliacez. Genus Trentepohlia Martius. 10. Trentepohlia aurea L. LINNE: Systema Nature (Byssus aureus). Masatierra: The typical form of this species, which is among the most widely distributed Algz in the world, was common in caves. Area of distribution: world-wide. FRESHWATER ALGH FROM JUAN FERNANDEZ AND EASTER ISLAND 89 It. Trentepohlia aurea ? forma. Forma in Printz: Subaérial Algz from South Africa. Det Kgl. Norske Viden- skabers Selskabs Skrifter 1920. ‘Trondhjem 1921. The form observed very much resembled the forma recorded by PRINTZ from South Africa. The filaments were a little broader (breadth from 10 to 15 y.) but he also mentions specimens up to 17 yp. I quite agree with Dr. PRINTZ in regarding the difference between 77ent/e- pohlia aurea and Trentepohlia abietina as insufficient for specific separation. The reasons for this, put forward in his paper seem very convincing. Pre- viously a couple of 7ventepohlia species are known to occur in Juan Fernandez and Easter Island, such as Jyrentepohlia polycarpa and the doubtful 77ente- pohlia Tuckermanniana. Masatierra: in caves, Cumberland Bay. Area of distribution: previously only known from South Africa. Order Protococcales. Family Protococcacee. Genus Chlorella Beijerinck. 12. Chlorella ellipsoidea Gern. GeRNECK: Z. Kenntn. nied. Chlorophyceen. The specimens, which occurred in considerable quantities, were all very badly preserved and the cytological structure could not be observed with accur- acy. To judge from their size, however, they seem to belong to Chlorella ellipsoidea Gerneck and certainly they come nearer to this than to any other species of the genus. Masatierra: forming green, slimy masses on rocks in cave. Area of distribution: hitherto only reported from a few countries, mostly from cultures in laboratories, but undoubtedly of wide distribution. Class Myxophyceee. Order Coccogonee. Family Chroococcaceez. Genus Chroococcus NAEGELI. 13. Chroococcus minor (Kuetz.) Naeg. NaEGELI: Gatt. Einz. Alg.; Protococcus minor Kurrzinc: Species Algarum. The specimens observed were few and in no good state of preservation. I have compared them with the original ones in the herbarium of NAEGELI go K. MUNSTER STROM and there can be no doubt about the identification. They were quite similar to the forms of the species so frequently found throughout Europe. Masatterra: in caves. Area of distribution: world-wide. Genus Gloeocapsa Kuetz. em. Naegeli. 14. Gloeocapsa montana Kuetz. Kuerzinc: Phycol. generalis, incl. G/. polydermatica and Gil. quaternata. Masatierra: This common species occurred in a single sample, scattered among specimens of Wostoc sphericum. Area of distribution: ubiquitous. Order Hormogonee. Family Stigonemacee. Genus Stizonenia Agardh. 15. Stigonema turfaceum (Berk.) Cooke. Cooke: British Freshwater Algze; BERKELEY in English Botany tab. 2826. Masatierra: The typical form of this characteristic species occurred in one sample, collected on rocks and forming a nigro-pulvinate stratum. Size: Crass. 25—38 p; Crass. horm. 12 p.; Long. horm. 45 p. Area of distribution: ubiquitous. 16. Stigonema tomentosum (Kuetz.) Hieron. Hieronymus in Hedwigia 1895; Strosiphon tomentosus KuEYTZzING in Bot. Zeit. 1847; RABENHORST: Exsicc. No. 694, etc. Masafuera: on rocks. This distinct species is represented with its typical form in a single sample. Area of distribution: probably ubiquitous. Genus Diplonema Borzi. 17. Diplonema rupicolum Borzi. Borzi: Studie sulle Mixoficee II. I am in some doubt if the species observed is quite identical with BORZI's species. FRESHWATER ALGAAS FROM JUAN FERNANDEZ AND EASTER ISLAND 9! The specimens, however, very much resemble his figures of Dzplonema rupitcolum, and the false branches are exactly like those. The dimensions, moreover, are quite the same as recorded by him, and I have observed the curious chroococcoid stage which occurred interiorly in the frond. To judge from his drawings the species presents a considerable range of variation and certainly the Myxophycea observed comes nearer to Diplonema rupicolum than to any other species. Masatierra: on rocks in cave. Area of distribution: previously only known from Italy. Family Nostocacez. Genus Nostoce Vaucher. 18. Nostoc sphericum Vaucher. VaucHER: Hist. Conferves d’eau douce. The determination is not perfectly sure as neither heterocysts nor spores could be found. However, the macroscopical appearence and the characters of the vegetative cells are facts so strongly supporting the determination that there can be very little doubt concerning the identification of the specimens. Masatierra. Area of distribution: world-wide. 19. Nostoc ellipsosporum (Desmaz.) Rabh. RABENHORST: Flora Europ. Alg.; Hormosiphon ellipsosporus DESMAZIERES: PI. crypt. de France; Wirrrock & Norpstept: Exsicc. 1329. Masafuera: The typical form of this species was observed in a single sample, collected on damp soil between Hepaticae. Size: Long. cell. 7—12 ; lat. cell. 4 »; lat. heterocyst. 6—7 u; long. heterocyst. 7—13 u.; long. spor. 12—19 w,; lat. spor. 6—8 yp. Area of distribution: probably ubiquitous. Family Oscillatoriacee. Genus Lyngbya Ag. 20. Lyngbya sp. Masatierra: among other Algz. The species was undeterminable as only a couple of filaments were observed. Distribution of the genus: world-wide. g2 K. MUNSTER STROM Genus Phormidium Kuetz. 21. Phormidium Retzii (Ag.) Gom. GomontT in Journ. de Botanique IV; AGarpDH: Disp. Alg. Suec. The form observed no doubt belongs to this species, but presents some likeness to Phormidium ambiguum. Szze: lat. fil. 6—7 pw. ; Masafuera: forming a stratum in waterfall. Area of distribution: ubiquitous. 22. Phormidium uncinatum (Ag.) Gom. GomonT in Journ. de Botanique IV; Osc7llatoria uncinata AGARDH in Flora X; Wittrock & Norpsrept: Exsicc. No. 96, »Phormidium membranaceum». Masatierra: The typical form was observed in a single sample. Area of distribution: world-wide. Genus Oscillatoria Vaucher. 23. Oscillatoria tenuis Ag. AGARDH: Alg. Dec. Masaterra: This common species was present in one sample, from wet rocks in Pangal. Area of distribution: one of the most generally distributed Algz in the world. 24. Oscillatoria sp. ° Oscillatoria terebriformis forma tenuis W. & G. S. Wesr: Freshwater Alge. British Antarctic Expedition 1907—g9. In one sample there was an Oscillatoria, which presented a striking simil- arity to the forma tenuzs described by Messrs. WEST from the Antarctic. The =, a = é ae a a x20 Os aaa Fis, 1. Ca ooa Xs dimensions were just the same, and also the cytological structure as far as it could be observed (the specimens were preserved in formaline solution), but the filaments were very little, if at all, terebriform. However, if to judge from Messrs. WEST’s figures, this character is not very striking in the specimens FRESHWATER ALG FROM JUAN FERNANDEZ AND EASTER ISLAND 93 they observed from the Antarctic, and certainly my specimens come very near to this form. Szse: lat. fil. 3 UW. Masatierra: On freshwater sponges. Area of distribution: The above-mentioned form is previously known from Cape Royds, Victoria Land. Addenda. 25. Nostoc punctiforme (Kuetz.) Hariot. Hariot in Journ. de Bot. 1891; Polycoccus punctiformis Kurtzinc Phyc. gen. This ostoc occurred in the intercellular spaces of three Guznera-species. The specimens are quite like those collected by LAGERHEIM in Equador and issued in WITTROCK, NORDSTEDT & LAGERHEIM: Algz exsiccate 1338. Masatierra: in Gunnera peltata Phil. and Gunnera bracteata Steud. Masafuera: in Gunnera Masafuere Skottsb. Area of distribution: World-wide. Christiania, The Botanic Museum of the Royal Fredericks University. April 27th, 1927. $6 °° Gaerne) 04k HOME. HAT OFT NDIA BATAWHESST | NS, 2 3in09 enainissqe uct’ elntas199 bere: satin ori font hover a ene Per. eyietin ag! Kaede! i | me a \-. aly ¢ la del savas ot) Pics. Ree Cae teaqoop rolawilesi nO + Aah ee mol inrotel ‘elatiotvata al ao} bariatiiaun- ovodie oth Torii Wo 8 cele iis ee es hcl bly if | img nese AP SORA: “otight (‘sieu2!) sarvoliteasd’ soto vie a ; ; a aay Deaf tyes EDA Doane hy angers PAN i a aes beet sh. aio Te asisoe aaah) ieeel) “le AEE wast tat hav va sah, sult. nt. bartiog Bac bak qoleu pat fat MISTS rid Ped aici }y:3 vi etiony. ook oak” oh ase, . To eae ae seeds sb 4a Re ‘Hi orecD ert A » Sond LEH a i % A caTeador aft OAELI Yo Bas SE RSE EE, RN r 5 ie - om iL % if 46 * a 7 | ih 5 te: " Fy taut SO5tN TA TA ' ul Paevy : | ics vn re ee, ™ atidsic crate gif. Sever 4 eae t “ry. Pare i, co oa tiie} “FU FO VE WOU Ee a ae Ee Se SO Tees ae Te ME als ; , RT, ee ee mE ee sc Mare ie Pts sy EEG ents ey anes | ST) oN CRUISES eae ere Oe ee aeey Wade Lor ig: Sivade t * yeu OO) Sa Fa é * fart 7 ; as a or upe mis a £ TC Ww i CG ‘ sae. Adasillaborin 22. - ; . a 4 Oh = oA a ee Fe ae FOr) ry) . tN \ a & f 7 Beil Asie Repeditins 18 ORs watople thete wasn Cheillubees, which prewoated ag a - " 2% - \o J SRG tO Use, Jarre ones describes by Moet Wrev from de ANTE 7- The Phanerogams of the Juan Fernandez Islands. By CARL SKOTTSBERG. With 11 plates (nos. 10—20) and 39 text figures. The compilation of a phanerogamic flora of Juan Fernandez has required more time and labour than could be anticipated. The general belief was, I dare say, that the higher flora of the group was very well known. That the writer did not share this opinion is evident from the fact that floristic investiga- tion was one of the principal objects of the voyage in 1916—17. My short visit in 1908 made me believe that the remote island of Masafuera was a promising field of research. The results communicated below will show, I think, that I was not disappointed. Masatierra, where quite a number of botanists have been, had more in store for us than we expected. There were many places of difficult access only visited by BERTERO, eighty years ago. Our knowledge of the distribution of the species in the islands was, in many cases at least, quite vague, as the earlier collectors seldom recorded the exact localities on their herbarium labels. Several species had been found once or twice only and remained little known. The starting point for my studies was of course the well known work of F. JoHow. But I had not advanced very far when I found that a reexamina- tion of many of the types was urgently needed. The majority of these had been discovered by BERTERO, and the discrepance between the writer and JOHOW concerning several of the species is partly due to the fact that JoHOw had no occasion to study BERTERO’s specimens. In other instances his species conception is a little too wide. I found it necessary to examine practically every type of the endemic plants. The collections of Santiago (GERMAIN, PHILIPPI, REED a. o. in the herbarium of Museo Nacional, and JOHOW’s private collection) and Kew (DOUGLAS, SCOULER, BERTERO, DOWNTON, MOSELEY a. 0.) were studied during longer visits to these places; other material was kindly sent for inspection from Berlin, Kew, Paris, Lund, Stockholm and Upsala. As usual, several collegues have assisted me in the determination of my plants. Dr. R. PILGER, Berlin-Dahlem, revised the Gramineae in collaboration with the writer, Dr. C. LINDMAN, Stockholm, determined some Poae, Rev. Dr. 96 CARL SKOTTSBERG G. KUKENTHAL, Coburg, revised the Cyperaceae, Dr. G. BITTER, Bremen, made a special study of Solanum, Acaena and Margyricarpus, Mr. O. E. SCHULZ, Berlin, gave me his opinion on Cardamine, Dr. F VIERHAPPER communicated some notes on my collection of Erigeron, and Dr. H. DAHLSTEDT, Stockholm, described a new form of Taraxacum. To all these gentlemen I wish to express my heartiest thanks for their generous assistance. I also thank Mr. A. KARNELL, Gothenburg, for the help he gave me when preparing the negatives of Plates TO, Al, 15s and 07; Finally, I use this occasion to thank my wife, Mrs. INGA SKOTTSBERG, for her never failing enthusiasm and her self-sacrifice during a voyage that was a great strain on body and mind alike. She was a most skilful, active and cheerful. companion, and contributed largely to the good results of our survey. The localities in Masatierra are enumerated from E. to W. along the north and south sides of the island, respectively; in the case of Masafuera, they are enumerated from N. to S. along the east coast and thence round to the west coast, and followed by the observations in the high mountainous region. Spanish names are used, with the following abbreviations: B. = Bahia, bay; C. = Cordon, mountain ridge;-Co = Cerro, mountain, peak; Pta = Punta, cape; Pto = Puerto, cove, harbour; Q. = Quebrada, valley with steep sides, gorge, canyon; V. = Valle, valley. Maps will accompany vol. I. Where a collector's or observer's name is not mentioned, the species has been observed by us in all the places quoted. The numbers in brackets refer to our collection, of which sets are kept in Stockholm (Riksmuseum), Gothen- burg (Botanical Garden), Upsala (Museum of the University), and in several foreign herbaria. A ! after the name of a collector signifies that I have examined the speci- men referred to. An * before a latin name indicates a new addition to the flora of Juan Fernandez; if before the name of one of the islands, it signifies that the species was observed for the first time in the island in question. The altitudes are in meters above sea level. It has been found advisable to indicate the state in which the specimens were collected: fl. —in flower, fr. with fruit. Other abbreviations will be understood without further explanation. All the illustrations in the text are reproduced after drawings made by the writer. ; C. Skottsberg. I. Indigenous species. Gramineae. Stipa L. 1. S. fernandeziana Phil. Anal. Univ. XLIII.560. — Syn. S. dzcolor in Jouow, Estud. 134 non Vahl; S. Skotisberga Pilger ex SKOTTSBERG, Stud. 22. — Fig. 1 a—d. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 97 On dry treeless slopes, also in many places along the high rocky ridges; an important grass in the western half of Masatierra and in the basal region of Masafuera. Masatierra: GERMAIN! PHILIPPI! — High peak between the colony and Pangal, 365 m; V. Colonial, not uncommon (fl. °/12 16, no. 19); Portezuelo de Villagra, 590 m (fl. */12 16, no. 28); C. Salsipuedes (fl.—fr. 8/12 16, no. go, 7°/12 16, no. 170); Q. Juanango; between Villagra and Pta Larga, in many places, but not abundant (fr. °/1 17, no. 244). Masafuera: Probably in all the valleys; Q. Seca, abundant; Q. de las Casas, very common to a couple of hundred m inland; Q. de las Vacas, abun- dant in the outer parts (fr. '°/2 17, no. 447). Leaves to 4 or 5 mm broad, generally convolute, densely pubescent below, glabrous or with few hairs above. Rays of the panicle about 6-flowered. Empty glumes very acute, outer 3-, inner 5-nerved, brownish purple with translucent margins and point, 14—15 mm long in my specimens. Pedicel of flower 3—3,5 mm, densely silky with longer hairs on the ventral side. Floral glume 7 mm long, with 5 prominent nerves, dorsal line pubescent to about 7/3 of its length, thence glabrous; external face covered with acute papillae; collar 1 mm high, brownish purple, the edge with a crown of white hairs about I mm long, but shorter on the ventral side; awn to 70 mm, slightly geniculate or almost straight, laxly twisted and pilose in the lower half, thence scabrous. Lodicules about I,r mm, ovate-truncate. Palea nearly 2 mm long, ovate-lanceolate, acutate, nerveless. Stamens dimorphous, the anterior with normal anther, 1 mm long, linear, fertile, the others with more ovoid, '/3s—'/z2 mm long, almost entirely sterile anthers. Anthers sometimes aristulate. Ovary about 1 mm long, fusi- form, stigmas short pedicellate. Flowers cleistogamous. This species was listed as S. manicata Desv. by PHILIPPI, Bot. Zeit. X1V.630, but distinguished as var. /zrsuta on the label of GERMAIN’s specimen in Herb. Santiago, collected Oct. 1854 in Masatierra. Later PHILIPPI, who collected the same species in Nov. 1864, described it as S. fernandeztana Phil. A comparison with the type showed that all my material must be referred to PHILIPPI's species; his plants, however, differ from mine in the empty glumes reaching a length of 21 mm while the awn is as much as 90 mm long. It is not at all remarkable that PHILIPPI at first regarded the island Supa as a variety of S. manicata. I have not seen authentic material of the latter, but to judge from the detailed description in GAy, Bot. VI.288, it must come very near this. .S. manzcata differs by the shorter leaves, which are hairy on the inside and glabrous on the outside, the longer ligule, the shorter glumes, described as green below, the shorter pedicel of the flower, shorter awn etc. The anthers seem to be alike in both. As only one S#ifa has been collected in the islands by all visitors known to me, it seems natural to identify S. Weeséana var. fernandesiana Trin. et Rupr. with S. fernandesiana Phil., but the description »foliis planiusculis cum nodis glabris ligula 11/¢-linealix does not answer to any form seen by me in the islands or in herbaria. My specimens of S. Neeséana from Central Chile (no. 1017), determined by Prof. PILGER, are glabrous and have smaller spikelets than S. fernandeziana. Prof. P1LGER was inclined to regard the latter as identical with 7— 20100. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 98 CARL SKOTTSBERG S. longifiora Steud Syn. I. 124, of which there is a sheet in Berlin (BERTERO no. 800). But the description of STEUDEL clearly forbids the identification. The leaves are I mm wide only, the rays of the panicle bear 1—3 flowers, the glumes are 5-nerved and only 6 to 8 mm long, the »floscule» is described as pilose all over. I collected a S#fa near Valparaiso (no. 1019), that answers very well to STEUDEL’s description, and this is not S. fermandeziana. 1 have no reason to doubt that S. fernandeziana is the same as S. longiflora Herb. Berlin, for STEUDEL quotes BERTERO no. 800 ex parte for his species. Prof. PILGER thinks that S. macrathera Phil. Anal. Univ. XCIII.720 is another name for S. fernandeziana. The description is incomplete, and I have no authentic spe- cimen at hand. S. fernandesiana Steud. |. c. was shown by PILGER ex SKOTTSBERG, Stud. 22 to be = Piptochaetium bicolor. My identification of S. Skottsbergiz Pilger |. c. with S. fernandestana Phil. was approved by the author of the former. There is a certain variation in the leaves, probably due to external conditions. The same kind of variation is found in Piptochaetium bicolor. Area of distribution: Central Chile; Juan Fernandez. Piptochaetium Presl. 2. P. bicolor (Vahl) Presl. — Jonow, Estud. 125. — Syn. Oryzopsis bicolor Speg., SKOTTSBERG, Stud. 22; Stipa bicolor Vahl, non JOoHOW nec HEMSLEY; S. fernandestana Steud. non Phil. Masatierra: Open ground in the centre of the island, not common. V. Colonial, near the colony (fl.-fr. ®/12 16, no. 18) and also higher up (fr. 74/12 16, no. 1201); Pto Ingles (fr. 19/1 17, no. 315); grassy slopes of Villagra, rather abundant (fr. °/1 17, no. 243). Area of distribution: South Brazil, Argentina, Chile, Masatierra. 3. P. laevissimum Phil. — JoHOw, Estud. 135. In the lower treeless parts of the valleys, abundant in the western half of Masatierra and in the basal region of Masafuera. Masatierra: V. Colonial, east part, dry soil near the sea (fr. '’/13 16, no. 110); C. Salsipuedes, here and there up to about 600 m; Pto Ingles, com- mon (fr. 19/1 17, no. 316); Q. Juanango, very common; between Villagra and Pta Larga, abundant, quite dominating over considerable areas (fr. °/1 17, no. 240). Masafuera: JoHow. — Forming extensive mats in the basal region from Toltén to Vacas, south of this valley not so common. It covers the bottom of the outer part of the valleys, as well as the ridges between them, alternating with the forest patches (fr. 1°/2 17, no. 553). This grass was first found in Masatierra by PHILIPPI in 1864 and re- discovered in 1908 by the writer. It is, perhaps, the most common of all grasses considered to be indigenous. Area of distribution: Central Chile (quoted by JOHOW from the Cordil- lera of Santiago and Colchagua, collected by me near Valparaiso); Juan Fernandez. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 99 ‘ Fig. 1. a—d Stipa fernandeziana: a_spikelet, X 24; b flowering glume, X 5; ¢ flower with palea and two lodicules, % 20; d anthers, X 10. e—i FPolypfogon imberbis, spikelets, flowering glume and palea: e leg. REED, f SKOTTSBERG no. 298, g no. 491, h no. 471, i f. avistata no. 1109. k P. crinitus, spikelet, flowering glume and palea. 1 Agrostis masafuerana, spikelet. — e—l X 10, : Podophorus Phil. 4. P. bromoides Phil. — Jonow, Estud. 135. Masatierra: GERMAIN! Discovered by GERMAIN in the latter half of October, 1854. PHILIPPI 100 CARL SKOTTSBERG states (Bot. Zeit. X1V.649) that it is »frequens in insula Juan Fernandez», but it has never been found a second time. All the material consists of the two sheets in Santiago and a third one in Kew. I need not tell that we made a careful search after this most interesting grass, but unfortunately without result. I must believe that it is a very rare plant. As it was in flower and fruit in October, it may have remained sterile during our permanence (Dec. 1—April 30), and was perhaps confounded with Avomus (Megalachne). Still, every suspected grass tuft was examined. Area of distribution: Monotypic and endemic in Masatierra. Polypogon Desf. 5. P. imberbis (Phil.) Johow, Estud. 136. — Syn. NMowodworskya imber- ézs Phil. Anal. Univ. XLIII.562. — Fig. 1 e—i. Masatierra: REED! (fig. Ie). —- Dry soil near the sea, in the western section, rare; Pto Ingles (PHILIPPI!); B. del Padre (fl. **/1 17, no.208, hewim): *Masafuera: near the sea in several places. B. Toltén (fl. ?/2 17, no. 471, fig. 1h); Tierras Blancas, not uncommon (fl.-fr. *"/2 17, no. 4gI, fig. 1 g). — New for Masafuera. f. avistata n. f. A typo differt gluma florali longe aristata. Masafuera: Playa Ancha (fl.-fr. '?/s 17, no. 1109, fig. 1 i). A careful comparison with PHILIPPI’s material in Herb. Santiago was made. The following notes may be added. Empty glumes 2,;—3 mm long, not. counting the awn, the inner generally a little shorter, entire, very acute, scabrous on the back, awn terminal, 0,;—2 mm long; floral glume about 2 mm long, exaristate, sparingly scabrid on the back, 5-nerved with 3 stronger and 2 weaker nerves, each prolonged into a scabrid point; palea about 1 mm, hyaline, ovate, + truncate and generally 2-pointed. Area of distribution: Endemic. 6. P. chilensis (Kunth) Pilger, Uber einige Gram. 386. — Chaetotropis chilensis Kunth; JOHOW, Estud. 136. Masatierra: on open ground in some of the valleys, also in the dry western section. V. Colonial (begin. fl. ?°/12 16, no. 206); Pto Ingles (fl. 19/1 17, no. 317); Q. Juanango (fl.—fr. °/4 17, no. 605); Villagra, not uncommon (fl. °/1 17, no. 239). Area of distribution: Central provinces of Chile and Masatierra acc. to JoHow, but apparently also in other parts of South America, for PILGER quotes as synonyms V7/fa muricata Presl from Pert and Agrostis pectinata Hackel et Arechavaleta from Uruguay. It was perhaps introduced to Masatierra with the traffic of the 18th century. Agrostis L. #7, A. masafuerana Pilger, Uber einige Gram. 388. — Fig. 11. Masafuera: in the alpine fell-fields, very local; Las Torres, 1370 m (past fl. 14/2 17, no. 424); C. del Barril, 1290 m. — An addition to the magellanic element. Area of distribution: Endemic in Masafuera. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 101 Trisetum Pers. 8. T. chromostachyum Desv. — JOHOW, Estud. 137. Masatierra: GERMAIN! REED! — In my previous paper on the islands I excluded this from the list of indigenous species. There are, however, no special reasons to regard it as introduced. It must be rare, and we never found it. Area of distribution: Central and South Chile; Masatierra. Danthonia DC. 9g. D. collina Phil. — Jonow, Estud. 138. — Fig. 2 a. Masatierra: PHILIPPI! — Dry rocky ridges and sunny arid slopes; the peak between Pangal and V. Colonial, 365 m (past fl. 17/12 16, no. 148); Q. Dama- juana (fl. °/12 16, no. 52); C. Salsipuedes, dry, wind-exposed rocks, not uncom- mon; Pto Ingles (fr. dispersed, '°/1 17, no. 328); Villagra, frequent on the dry slopes (SKOTTSBERG 1908; fl.-fr. °/1 17, no. 247). First found by PHILIPPI and rediscovered (sterile) by the writer in 1908. As there are so many American species of this genus I have thought it better to add a few figures for comparison. Area of distribution: South Chile (Valdivia); Masatierra. Koeleria Pers. 10. K. micrathera (Desv.) Griseb. Symb. Fl. Argent. 292. — Syn. 77ise- tum micratherum Desv. ex Gay, Bot. VI.352; 7. /axum Phil. Anal. Univ. XLIII.568, Jonow, Estud. 137; PILGEeR, Uber einige Gram. 387. — Fig. 2 b—h. Masatierra: PHILIPPI! — Rocky places near Portezuelo de Villagra, in brushwood c. 590 m, very scarce (fl. '°/1 17, no. 279). I have not seen the type of Z7r¢setum micratherum, to which 7. /axum is referred on the authority of Prof. PILGER; according to the description the former has much shorter leaves (only 2—3 inches), pubescent sheaths, outer glume 7/3 as long as the inner, etc., but these differences disappear, it seems, at a closer look, for PILGER writes |. c. 388: »Die Skottsbergschen Exemplare stimmen mit denen von Valdivia durchaus iiberein». If the plant from Argentina belongs here is an open question. In his monograph of Koeleria, Bibl. Bot. 65, DOMIN calls K. mzcrathera a »mixtum compositum». He had seen PHILIPPI’s specimens from Valdivia and concludes that they must belong to some other genus. His reason for excluding it from Koelerta evidently is that, as it is perennial and caespitose, it belongs to the subgenus in sylvaticis petrosis montium editiorum>. Gay, Bot. H.35, quotes Co//etia from Pto Frances and says that it flowers in March. JOHOW lists as collectors only BERTERO and JOHOW, but, as seen above, the species is not quite so rare as has been supposed. Area of distribution: Endemic in Masatierra. Flacourtiaceae. Azara Ruiz et Pav. 75. A. fernandeziana Gay. — A. serrata Ruiz et Pav. var. fernandeziana - (Gay) Reiche, Fl. de Chile I. 134; JoHow, Estud. 109. — Fig. 15 a. Masatierra: In the montane forest, supposed by JOHOW to be limited to a single locality, but found in many places by the writer. Only a single or some few trees were observed in each place. — C. Chifladores, small valley facing the Frances Valley, solitary, c. 500 m; on the ridge separating Q. Laura 10 — 20100. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 146 CARL SKOTTSBERG from Q. Piedra Agujereada, c. 500 m, rare; V. Colonial, below Portezuelo, c. 450 m, one large tree in dense forest (fl. */12 16, no. 8) and 570 m, two small plants; C. Central, c. 400 m; Q. Gutierrez, 350—450 m, rare (fl. “liz 16, no. 41); Q. Seca, c. 500 m; C. Salsipuedes; mountain ridge west of Co Yunque, 530 m; Q. Villagra, 500-550 m, scattered (unr. fr. Apr. 17, no. 584); Q. dela Choza, several trees, c. 300 m. : In Stud. 18 I restored GAy’s species, saying that it differs quite as much from serrata as this from J/anceolata. This statement is not correct, for the island species is very near se7- rata. Of the latter I have seen ample material in Santiago and Kew. We learn from the original description (Bot. I. 196), that A. fernandeziana Nas membranous leaves and a fruit twice as long as the calyx, while A. serrata has coriaceous leaves and the fruit three to four times as long as the calyx. The only notable difference in STEUDEL’s descrip- tion (Flora 1856, 421) is the glabrous calyx in the former, but the calyx is hairy in both. REICHE tried to distinguish the island form by the broad, thin leaves; JoHow adds that the berries are pure white, while, in A. serrata, they are blue and glaucous. I have not seen any quite ripe fruits of A. fernandezzana, but all I have are very dark, and I doubt they would have turned white, had they been left to ripen. The species is upheld with hesitation. Certainly there are differences (see fig. 15); the leaves are thinher (but not broader, nor are the stipules larger), the pubescence of the stem and peduncles is denser, the sepals 2,;—3 mm long against I,5—2 in serrata, the anthers a trifle larger, the style shorter, 3,5 mm against 4—5 in the other. Area of distribution: Endemic in Masatierra. Fig. 15. Flower and sepals of a Azara fernandeziana, b A. serrata. X 5. Myrtaceae. Ugni Turcz. 76. U. Selkirkii (Hook. et Arn.) Berg. — Jonow, Estud. QI. Masatierra: Common on the highest ridges above the forest, associated with Pernettya rigida but much more scarce than this. C. Centinela, 500—800 m; ridges of Co Damajuana, large shrubs (fl. ?°/1 17, no. 352); V. Colonial, C. Central, 450—570 m; Portezuelo de Viilagra, the steep slope of Co Piramide, 575—600 m (buds 72/1 17, fl. 31/1 17, no. 329); C. Salsipuedes, 500—700 m. Area of distribution: Endemic in Masatierra. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 147 Myrteola Berg. *77. M. nummularia (Poir.) Berg. Masafuera, in the alpine region: near the Correspondencia Camp, 1130 m (fr. °/3 17, no. 376); C. del Barril, foot of C. Atravesado, 1350 m (past fl.-fr. Ses kR 00. 548); east slope of Los Inocentes, c. 1200 m, and below the summit, 1350—1400 m (fl.-fr. °/s 17, no. 377). Another interesting addition to the subantarctic group. The specimens from Inocentes have almost orbicular leaves, being more »typical», while, in the rest, they are more ovate. Area of distribution: Chile, Cordillera of Nahuelbuta and Valdivia to Fuegia and the Falkland Islands. Myrceugenia Berg. 78. M. fernandeziana (Hook. et Arn.) Berg. — JoHow, Estud. 94. Masatierra: The leading forest tree, common from near the sea level to the high ridges and ranging from Pto Frances to the westernmost forest patches of Q. Juanango and Co Chumacera. Fl. Dec. 1916 (nos. 39, 75); unr. fr. March—Apr. 1917 (no. 575). Area of distribution: Endemic in Masatierra. 79. M. Schulzei Johow, Estud. 96 (as Schu/sz7z, but named in honour of Mr. J. Schulze). Masafuera: The principal forest tree, forming a belt in the valleys between 200 and 800 m, more or less. Fl. Feb. 1917 (no. 505). Area of distribution: Endemic in Masafuera. Gunneraceae. Gunnera L. 80. G. peltata Phil. — Jonow, Estud. 98; SKOTTSBERG, Stud. 12, Fig. 2. Syn. G. msularis Schindl. zon Phil. — Fig. 16 a. Masatierra: Common in the forested parts, preferring humid ravines, where the stem and leaves attain a very great size. Not seen at a lower altitude than 160 m. It reaches the foot of the high ridges, where intermediate _forms between this and G. éracteata are found. To the west of Villagra, on the south side of the island, G. fe/tata descends into the ravines below the limit of the forest, reaching its farthest west at the foot of Co Tres Puntas. Fl. Q-stage Dec. 1916 (nos. 87, 139); fr. Jan. 1917 (no. 139 b). I have not much to add to my earlier account of this species. This time young plants were found in plenty; they are more or less coarsely villous. The lamina of the first leaves is reniform and not peltate, while the leaves of the 148 CARL SKOTTSBERG adult plant nearly always are peltate. Inflorescences collected Dec. 1916 had lost their petals and stamens. The stigmas were white in these specimens, not crimson as stated for the plants examined before. Area of distribution: Endemic in Masatierra. 81. G. bracteata Steud. — JOHow, Estud. 99; SKOTTSBERG, Stud. 10, Fig. 1. Syn. G. zxsularts Phil. zon Schindl. — Fig. 16 b. Masatierra: In the central parts of the island, much more rare than the former and belonging to the region of dwarf trees and shrubs along the ridges. Seldom found below 500 m. — El Rabanal (JOHOW); Pico Central (JoHow); Portezuelo de Villagra, common in the vicinity of the pass, c. 550—600 m (unr.-fr. 2/12 16, no. 1; also observed by JoHOW); C. Salsipuedes, scattered in the thickets on the crest, c. 600 m; highest part of QO. Villagra, c. 500 m, with an unusually large stem but otherwise typical. — Forma folizts plerumque pel- tatis: C. Centinela, 700—800 m, common (fr. '*/4 17, no. 610). A remarkable form, found in considerable quantities on the high ridge east of Co Yunque. The leaves generally are peltate; this is, as will be shown below, not uncommon in intermediate forms, explained as hybrids, but no. 610 is a typical bracteata as to all other vegetative characters. Still, it may not be quite pure. Area of distribution: Endemic in Masatierra. [G. msularis Phil. (see SKOTTSBERG, Stud. 14). Young plants of G. drac- teata, collected by the writer, are identical with PHILIPPI’s type of zzsa/arzs in Herb. Santiago, leg. GERMAIN 1854! G. pyramidalis Schindl. (see SKOTTSBERG, Stud. 14). 1 have tried in vain to get hold of the type; I am convinced that it is not a separate species. The type consists of a piece of an old spike of BERTERO no. 1463, without leaves or ligules. It differs in the very small bracts, considerably smaller than in the top region even, of the other species. I don’t know if the bracts in the type are intact; I have seen old spikes where they have disappeared, leaving a stump standing. In Herb. Kew I saw such a piece of BERTERO 1463 (this number embraces all the forms of Guunera from Masatierra); it belongs to G. bracteata. | G. bracteata Steud. x peltata Phil. — Plate 13, 2; text fig. 16 c—i. Near Portezuelo JOHOW found a Gunnera with rugose leaves and entire scales, regarded by him as a bastard between the two species. I have found numerous intermediate forms in several places; as hardly one is like the other, I believe they are hybrids and also that the F I-generation is fertile so that we get an F 2 and following generations. As the two species differ in many characters and as there are so many possibilities to combine these, the theoretical result will be a long series of forms uniting the two extremes, and this is exactly what we find. The main differences between pe/tata and édrvacteata may be summarized as follows. G. peltata: Leaves orbicular, reniform of generally peltate, rugose, petiole and lower surface of the blade with scattered tubercles on the larger veins, THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 149 upper surface very rugose. Scales crimson, long and narrow, with thick fleshy midrib, densely and deeply lacerate, very hirsute on back. Scape more or less pilose, bracts linear, oblanceolate-acute. Budding parts not very slimy. A ont | Fig. 16. Scales of Gunnerae: a G. peltata, b G. bracteata, e—i intermediate forms — see text! All # nat. size. G. bracteata: Adult leaves orbicular-reniform, smooth, glabrous and lustrous, petiole smooth. Scales green or brownish green, thin, entire, broad ovate, very blunt but sometimes apiculate. Scape glabrous, bracts spathulate. Budding parts embedded in thick mucilage. Intermediate forms were observed in Masatierra: Portezuelo de Villagra, near the road, c. 300 m (no. 1230) and one fine specimen c. 500 m (Plate 13, 2); close to the pass, c. 575 m (fr. 7/4 17, no. 585, *°/4 17, no. 624); C. Salsipuedes, 150 CARL SKOTTSBERG 600—650 m, common together with the supposed parents (fr. */12 16, no. 80, 13/, 17, no. 80a, b); Pto Ingles, central ridge, 575 m (fr. 19/1 17, no. 323); Q. Villagra, c. 500 m (fr. *°/1 17, no. 1218). The following notes on the leaves and scales of these specimens will give an idea of the range of variation. No. 624. Fig. 16 c. — Adult leaves all peltate, very rugose below, larger tubercles on veins and petiole few, small; blade almost smooth above. Scales linear—triangular, slightly pilose on back, reddish brown; margins short lacerate. Scape glabrous, bracts mostly gone, remaining ones + intermediate in shape. No. 1230. Fig. 16 d. — Leaves peltate, petiole aculeate and pilose, nerves with few tubercles, lower surface rugulose, upper very rough, thus leaves like those of feltata, but the lobes more blunt. Scales entire to short lacerate, ovate to linear-triangulate, reddish brown, pilose on back. No. 585. Fig. 16 e. — Leaves only slightly peltate, otherwise as in fel- tata; scales narrow linear to triangular, reddish brown, shortly lacerate, densely pilose on back. Scape pilose, bracts as in peléata. Spike less dense than in this. Budding parts more slimy than in fe/tata. No. 323. Fig. 16f. — Leaves not peltate, with acute, serrate lobes, pil- ose on the nerves below, with few and low tubercles. Scales ovate—triangulate, acute, entire or slightly dentate, brownish green with reddish centre, a little pilose on back. Scape glabrous; no intact bracts left. No. 80a. Fig. 16 g. — Leaves slightly or incompletely peltate, slightly rugose below and almost smooth above, petiole strongly aculeate. Scales ovate— triangulate, acute, entire or dentate, greenish brown with reddish centre, glabrous. Scape glabrous, bracts as in éracteata. No. 80 6. — Like the former, but leaves peltate and nearly smooth on bothsides. Scales narrower, rarely dentate, slightly pilose on back. Bracts a little narrower. No. 80. Fig. 16h. — Leaves peltate, smooth, slightly pilose on the nerves; petiole with scattered low emergences. Scales greenish brown with reddish centre, broad linear—triangular, acute, entire or lacerate in the upper half, almost elabrous. Scape nearly glabrous, bracts of feltata type. No. 1218. Fig. 16 i. — Leaves not peltate, nearly smooth above, slightly rugose and aculeate below. Scales green, ovate—triangular, acute, entire or lacerate, a little pilose on back. Scape glabrous. No intact bracts left. 82. G. Masafuerae Skottsb. — Stud. 14, Fig. 3; Taf. 3—4. Masafuera: known before only from Q. de las Casas, now found in many places and not at all confined to the canyons, but quite common higher up on the mountain slopes or in the shallow ravines in the alpine region. — Along the road to Las Chozas, open spaces in the forest belt; in the higher treeless tracts between 700 and 1100 m, extensive patches in some places; Q. de las Casas, between stones by the stream and on the rock walls of the canyon (unr.-fr. 8/9 17, no. 400); Q. de las Vacas, reaching farther down owing to the greater supply of water; Q. del Blindado, in the forest, 440 m; Q. Angosta, by the waterfall; Q. de la Loberfa, 280 m, in open forest. Also in the alpine region: THE PHANEROGAMS OF ‘THE JUAN FERNANDEZ ISLANDS 151 on the western wall of the island, not uncommon; near Las Torres, 1350 m; top of Co Correspondencia, c. 1400 m. To my description of 1914 the following notes should be added. Old trunks sometimes as much as */2 m high and 2 dm across; petiole to 11/2 m long, occasionally as much as 12 cm wide at base; largest lamina measured 2,8; m across. Scales densly pilose on back, as in G. feltata. Largest scapes measured 1,67 and 1,75 m, respectively, very heavy, drooping, as much as 26 cm across at base. This species has larger leaves than the other Juan Fernandez species. Besides by other characters, it is distinguished by the branches of the spike becoming conspicuously incrassate in the fruiting state. The fruit is bright orange red turning crimson when ripe. In No. 400 transitions were found between normal leaves and the so-called »ligules», showing that the latter are not ligular structures, but reduced leaves, homologous with the large green leaves as well as with the reduced bracts. I shall return to these organs in another paper. The young petioles of G. Masafuerae have a more agreeable, less acid taste than those of G. felfata; they are quite refreshing, a fact well known to the islanders, who distinguished G. MJasafuerae as a special sort of »nalca» (native name for the petioles) long before it was described by the writer. Area of distribution: Endemic in Masafuera. Halorrhagidaceae. Halorrhagis Forst. A species of Halorrhagis was first collected in Masatierra by Mrs. GRAHAM in 1823 and later by SCOULER, BERTERO, CUMING, GERMAIN, PHILIPPI, REED, MOSELEY and JOHOW, also by me in 1908. I have seen specimens brought by all these collectors. SCHINDLER, Pflanzenreich, also quotes J. D. HOOKER, who never visited Juan Fernandez. All the specimens were referred to one species and identified with Cercodia erecta Murr. Comment. Goetting. II (1780) 3 tape-t = 77. alata Jacq. = 7. erecta (Murr.) ‘Schindler, 1. ¢. IV. 225 (1905) 49, fig. 14. The plant from Masatierra goes under one of these names in all publications and herbaria, also in the monograph of SCHINDLER. In 1908, the same name was used by me, on the authority of SCHINDLER. At that time I had no reason to doubt his statements, with one exception, viz. that he quotes Hf. erecta as a native of Chile (GERMAIN); it does not occur on the continent, and GERMAIN’s specimens came from Juan Fernandez. After I had discovered a number of forms in Masafuera, I began to study H!. erecta from New Zealand and found that the plant from Masatierra was a distinct species. It is evident that SCHINDLER based his description on speci- mens from both New Zealand and Juan Fernandez (Masatierra). Before my visit in 1908 Halorrhagis had been brought from Masafuera only by GERMAIN. No distinction was made between this and the form in Masatierra. Both are quite smooth, but otherwise’ very different from each 152 CARL SKOTTSBERG other. I have seen GERMAIN’s plant and I have collected the same species in several places in 1917. The form I found in 1908 was, however, not this, but a very scabrous one, more like the true /7. erecta; it was mentioned in my account as var. scabva. I have found it in many places in 1917, in flower and fruit. It stands very near erecta, but must be kept separate. Thus, H. erecta does not occur in Juan Fernandez, but is replaced by three different species. On the other hand, it seems that //. evecta of New Fig. 17. a—e Halorrhagis erecta, cult.: a leaves, nat. size, b bud, X 6,6, ¢ sepal and d petal, X 20, e fruits, X 66. f—g H. erecta leg. Cheeseman: f leaf, nat. size; g fruits, X 6,6. h fruits ot H. Colensoi and i of H. cartilaginea, X 6. Zealand comprises two distinct plants. MURRAY described a scabrid plant with alate fruits (see fig. 17 a--g). In Herb. Kew I saw specimens of this kind col- lected by FORSTER, CUNNINGHAM, BUCHANAN, J. D. HOOKER, CHEESEMAN a. 0. IORSTER’S specimens are less scabrous than the rest. But under the same name was a smooth form collected by COLENSO, with small, barely alate fruits, recalling my A. masaterrana in habitus, but differing in the shape and size of the fruit (see fig. 17 h). I have called this H. Colensoz. Hf. cartilaginea Cheesem., later reduced to a variety of H/. erecta (Man. N. Z. Flora 148) was justly reestablished by SCHINDLER. It seems to be THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 153 a very distinct species (fig. 171). To the same subsection Cercodia (Murr.) Schindl. also belong //. /aevis, exalata and fpedicellata, all natives of Eastern Australia. Key to Subsect. Cercodia. Memeescrials, (IGeNtate ted, Pedwclala Schindl. II. Sepals entire. A. Whole plant more or less scabrid. 1. Densely scabro-velutinous. Leaves thick, broadly ovate. FT, cartilaginea Cheesem. 2. Coarsely scabrid. Leaves thin. a. Stem with large, scattered hooks. Teeth of leaves pungent. H, exalata F. Mill. b. Stem with minute prickles, teeth of leaves not pungent. 7. Leaves barely scabrous. Stigma papillae few-celled. Fruit mare than esa fone |* eer ff. erecta (Murr.) Schindl. 6. Leaves very scabrous. Papillae multicellular. Fruit hardly 2 mmileney Meee. 2 Ae. TT agama Skettsb. B. Whole plant perfectly smooth. i eaves larce, natrowauccolate. Se. gees Schindl. 2. Leaves ovate or broad lanceolate. a. Fruit large, generally over 4 mm. x. Narrow ovoid, distincly costate. 7. masatierrana Skottsb. 6. Broad ovoid, not costate. ......... H. masafuerana Skottsb. B-) Proit Smallec.”3muttong: <2)... AZ. \Colensoz, Skottsb. *83. HH. asperrima nov. spec. — Syn. 7. erecta var. scabra, SKOTTS- BERG, Stud. — Fig. 18. Suffrutex semimetralis, glaber sed scaberrimus, fere a basi opposite ramo- sus. Rami tetragoni longitudinaliter ac conspicue 4-lineati, lineis elevatis valde scabris. Folia opposita, superiora saepe alternantia, internodiis sat longis sepa- rata, tenuia, Jaete—obscure viridia subtus paulo pallidiora, ubique asperrima, ovata vel ovato-lanceolata, basi cuneato-rotundata, longe (10—18 mm vel inter- dum ultra) petiolata, lamina argute serrata dentibus distincte apiculatis utrinque I2—14, 30—50 mm longa et 10—22 mm lata, superiora sensim minora, pro magnitudine longe petiolata, anguste lanceolata — ovato-rhomboidea, parcius dentata — subintegra. Inflorescentia spicata foliosa, e dichasiis bracteolatis 3—7- floris composita. Flores brevissime pedicellati (*/2 mm), omnino scabri. Ovarium tetragono-ovoideum, interdum mox alatum, I,2—I,; mm longum et I mm latum. Sepala 4 acute triangulata, minute denticulata, 0a—O 9 mm longa. Petala 4 navicularia, carinata et in carina scabra, breviter unguiculata, 2—2,5 mm longa, viridia—alborosea. Stamina 8 filamentis tenuissimis c. 1,5 mm longis, antheris flavis dorso + rubescentibus, I—1,s5 mm longis. Styli 4, ad 0,8 mm longi, stigmata alba vel rosea papillis longis multicellularibus. Fructus scaber, viridis— obscure violaceus tetragono ovoideus pyriformis exalatus vel 4-alatus, cum sepalis persistentibus 2,7—3 mm longus et 2 mm latus, inter margines alarum autem ad 4 mm metiens. 154 CARL SKOTTSBERG Masafuera: SKOTTSBERG 1908, ster. — Q. de las Casas, quite common (f-fr. ‘4/2 17, no. 457 — f. fructibus alatis); Q. de las Vacas (fl.-fr. 1°/2 17, no. 442 — f. non alata); Q. del Blindado, in the forest c. 440 m; Q. del Vara- dero; on the precipice above Buque Varado, c. 1200 m. Fig. 18. Halorrhagis asperrima; a—f from no. 457: a two lower and b three upper leaves, nat. size; ¢ flower (petals and anthers removed), X 6,6; d sepal and e petal, X 20; f fruits, X 6,6. g—p from no. 442: g two lower, h three upper leaves, nat. size; i flower on male and k on female stage, X 6,6; 1 sepal and m petal, X 20; n typical fruits, o with trace of wings, p one of these seen from the top, all X 6,6. No. 457 and 442 are so unlike each other that I have been inclined to bring them to different species. No. 457 is wholly green, the leaves are broader, especially in the floral region, the branches thinner, the fruit mostly alate with low, broad wings, see fig. 18 a—f. No. 442 is more or less tinged with antho- THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 155 cyanin — though never so bright red as //. erecta —, the leaves are narrower, the fruit only slightly tetragonous and mostly without a trace of wings, see fig. 18 g—p. Fruits from both numbers were sown in the Gothenburg Garden and yielded numerous plants; at present I have 13 specimens of no. 457 and 32 of no. 442 in cultivation. The latter are exactly like the mother plant, but the former are unlike no. 457 and habitually very like no. 442. Still, small wings develop in many fruits as in the mother plant, but not so regularly. 1 suppose the difference in vegetative characters is due to external conditions, but it is less probable that these should influence the development of wings. H. asperrima is very near H. erecta, but this is less scabrous and the leaves are nearly smooth, especially the margin, which, in 4. asferrima, is thickly beset with microscopic aculei. The stigma-papillae are mostly 2-celled and the fruit 3,s—4 mm long in erecéa. Area of distribution: Endemic in Masafuera. 84. H. masatierrana nov. spec. — H. alata (erecta), JOHOW, Estud. 97, SKOTTSBERG, Stud. 10, zoz Murr. nec Jacq.; SCHINDLER in Pflanzenreich p. p. — Fig. 19. Fig. 19. Halorrhagis masatierrana: a two lower, b three upper leaves, nat. size; ¢ flower on male and d on female stage, X 6,6; e sepal and f petal, x 20; g fruits of no. 304, h of JoHow no. 6, all x 6,6. Suffrutex 2—5 dm altus, glaber, laevissimus, fere a basi opposite ramosus. Rami cylindraceo-tetragoni, longitudinaliter 4-lineati, cortice + rufescente. Folia omnia opposita vel superiora florigera alternantia, discoloria, supra laete viridia, subtus glaucescentia, inferiorum lamina ovata, acuta, basi late cuneata, petiolo 3—8 mm longo suffulta, plerumque 15—30 mm longa et g—14 mm lata, 156 CARL SKOTTSBERG regulariter serrata dentibus utrinque 6—9, superiora multo minora, paucidentata- subintegra, basi longe cuneata, suprema bracteiformia. Inflorescentia densa vel laxa, foliosa, e dichasiis bracteolatis plerumque 3-floris composita. Flores brevissime (I mm) pedicellati, omnino laeves. Ovarium subtetragonum et sat distincte 8-lineatum, 0,.—1I mm longum et fere latum, cum pedicello + rufescens, Sepala 4 late ovato-triangularia, suberecta, 0,, mm longa. Petala 4 navicularia alborosea carinata, 3—3,5 mm longa, manifeste unguiculata. Stamina 8 fila- mentis c. 1 mm longis, antheris flavo-roseis 2—2,3 mm. Styli 4, 0,8 mm longi, stigmata rosea papillis pauci-(1—3)cellularibus. Fructus anguste tetragono- ovoideus, in specim. num. 304 cum sepalis 4—4,5 mm longus et I,7—I,9 mm latus, non alatus sed lineis paulo elevatis 8 notatus, fusco-viridis. Masatierra: Not rare on the dry, rocky ridges, also on open, stony ground in the forest belt. Rabanal (JOHow); El Pangal, on the western slope; C. Centinela (JoHoOw); V. Colonial, C. Central (also JOHOW), 570 m (fl.-fr. 78/1 17, no. 304), Q. del Monte Maderugo, road-side in the macal, 240 m, and rocky wall, 390 m; C. Salsipuedes, frequent, 350—650 m (fl. 7°/12 16, no. 172); Porte- zuelo de Villagra, not rare on both sides (fl. °/12 16, no. 34 — also observed by JoHow); ridge between Vaqueria and Q. Juanango, 300 m; Q. Juanango, outer part. South side of the island, Q. Villagra, rare in the forest c. 500 m; east side of B. Chupones, barren slopes. This differs widely from the typical scabrous 7. erecta, but it is more like what I have called 77. Colensoz, especially the more stunted form of A. masa- tierrana, growing on the sunny ridges (no. 34, 172). Of this I got no fruits. JoHow collected fruiting specimens of exactly the same habitus (no. 6). The fruits of these, only 3,s—4 mm long (but hardly quite ripe) come very near H.. Colensot, where, however, they are transversely rugose. Both have traces of wings (see fig. 19 h and 17h). Still, I believe there is only one species in Masatierra and that A. Colensoz is different. Area of distribution: Endemic in Masatierra. *85. H. masafuerana nov. spec. — Syn. 7. alata, JoHOW, Estud. 98 quoad plantam e Masafuera. — Fig. 20. Suffrutex semimetralis patente opposite ramosus glaber laevissimus. Rami cylindracei—obscure tetragoni 4-lineati. Folia subcarnosula laete viridia, inferiora petiolo 8—10 mm longo suffulta, lamina ovato-lanceolata, basi late cuneata, margine incrassata, grosse serrata dentibus utrinque g—1II, c. 45 mm longa et 20 mm lata, superiora sensim minora et brevius petiolata, suprema subrhomboidea subintegra. Inflorescentia laxa foliosa, e dichasiis bracteolatis 1—3 (rarius ad 5)- floris composita. Flores brevissime ('/2—1 mm) pedicellati, omnino laeves. Ovarium ovoideum 1,5 mm longum et I,2 latum. Sepala 4 triangulata margine integro, I—I,2 mm longa. Petala 4 viridia fusconotata, navicularia, carinata, manifeste unguiculata, 2,,—3 mm longa. Stamina 8 filamentis ad 1,5 mm longis, antheris ad 1,6 mm. Styli sepalis aequilongi, stigmata alba, papillis longis sed paucicellularibus. Fructus pyriformis quam in ceteris major, cum sepalis 4,2— 4,6 mm longus, basi 3—3,2 mm latus, fere semper alis omnino destitutus, maturus + violascens, in sicco obscure 4-sulcatus. Adsunt specimina fructibus tubercu- latis nec non alia fructibus irregulariter alulatis. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 157 Masafuera: GERMAIN oct. 1854 (Herb. Santiago!); QO. de las Vacas, on the walls in the outer section (fr. '°/2 17, no. 441 — f. tuberculata); stony beach near Varadero (fr. °”/2 17, no. 1216); Q. de la Loberia, near the shore (fr. **/2 17, no. 485 — f. alulata). Hf. masafuerana differs from all others in the size and shape of the fruit and in the smooth, shiny stems, leaves etc. I have sown seeds of all the different numbers; of no. 1216 I have numerous, of 441 a few plants in cultiva- tion. The difference between them is slight, and the tubercles on the fruit of Fig. 20. Halorrhagis masafuerana; a—f trom no. 1216: a lower, b upper leaves, nat. size; c flower on male and d on female stage, X 6,6; e sepal, X 20; f fruits X 6,6. g—h from no. 441: g petal, X 20; h fruits, X 6,6; i fruits of no. 485, X 6,6. the latter become pronounced only with desiccation, but are quite wanting even in the dry fruits of no. 1216. More material is needed to decide whether no. 441 represents a distinct variety. Of no. 485, the seeds did not germinate. Fruits of all the forms are figured. Area of distribution: Endemic in Masafuera. Umbelliferae. Eryngium L. 86. E. bupleuroides Hook et Arn. — JoHow, Estud. 100; SKOTTSBERG, Stud. ci6, bat. 59 Fiow4:, 2 Fig) .2) ag. 158 CARL SKOTTSBERG Masatierra: Not uncommon near the limit of the forest, along the higher ridges etc., rarely seen below 400 m. — C. Centinela, rather common on the ridge (also observed by JoHow); El Pangal, west branch, on the slopes from 400 m; in the gap between Co Damajuana and Co Yunque, c. 580 m; V. Colonial, C. Central, 400—570 m, scattered; Q. del Monte Maderugo, steep slopes 390— 500 m (fl. *4/1 17, no. 335); Portezuelo de Villagra, on both sides (fl-fr. Dec. 1916, no. 4, also observed by JoHOw); Pto Ingles, central ridge, c. 470 m; the ridge between Vaqueria and Q, Juanango, c. 300 m; forest on a mountain spur west of Co Yunque, 550 m; Q. de la Choza, foot of the steep ridge, 4oo—450 m; south wall of Co Chumacera. Area of distribution: Endemic in Masatierra. #87. E. imaccessum nov. spec. — Plate 12. Text fig. 21 h—p. Arbuscula glaberrima saltem ad '/2 m alta, parce pseudodichotome ramosa. Radix primaria profunde infossa parce ramosa. Truncus primarius ad 25 mm diam., cortice fusco leviter longitudinaliter striato et passim transverse rimoso; rami ultimi cicatricibus foliorum densissime annulati. Folia ad apices ramorum densissime rosulata, semiamplexicaulia, sessilia, valde firma, crassiuscula, obscure viridia et subnitida nervis perplurimis parallelis (circ. 5—7 crassioribus) percursa, late lineari-lanceolata, acutissima, apice pungentia, I0O—15 cm longa et I,s— 2,7 cm lata, basi 1,.—2 cm; maxima visa 17,5—23 cm longa et 2,8—3,8 cm lata, basi I,8—2,6 cm; margine angustissime scariosa, infra apicem dentibus acutissimis pungentibus utroque latere 1—4 (plerumque 2—3) munita, rarissime integerrima. Inflorescentia terminalis monocephala, scapo complanato 6—11 cm longo circ. 5 X 7 mm diam., cavo, viridi-violaceo. Involucrum ad 23-phyllum phyllis crassis durisque, acutis, pungentibus, basi incrassatis, supra obscure violaceis, subtus viridibus, anguste triangulatis, margine paulo incrassatis, 20—25 mm longis et 4—5 mm latis. Capitulum magnum atroviolaceum depresso-hemisphaericum, 21—27 mm altum et 27—37 mm latum; receptaculum vaginis bractearum pro- funde alveolatum, intus cavum, caverna ad 17 mm lata et 13 mm alta, pariete 2 mm solum crasso. Bracteae anguste rectangulato-subspathulatae, firmae, sub- carnosae, basi incrassatae albae, amplexiflorae, apice truncatae et obscure violaceae, margine angustissime albonotatae, ad 7-8 mm longae et basi 2—3 mm latae. Flores numerosissimi, jam deflorati solum visi, staminibus exceptis 7 mm longi. Sepala rectangulata, cochleata, apice truncata et + emarginata, sat crassa, alboviridia apice obscure violacea, 3—3,5 mm longa et I,s;—2 mm lata. Petala ex albo violascentia, duplicato-inflexa, apice subdentata, 2—2,1 mm longa et 0,7 mm lata, mox decidua. Stamina violacea filamentis 5 mm longis apice in- curvis, antheris ovatis 1,; mm longis. Stylopodium crateriforme viride; styli 3—4 mm longi, atroviolacei. Mericarpium fuscum subnitidum 4—5 mm longum et 2,5 mm latum, ambitu truncato-ovale, late et obtuse carinatum, dorso laeve vel inconspicue tuberculatum, sub apice appendicibus calycinis sat magnis or- natum, margine anguste alatum nec non dentatum. Masatierra: Portezuelo de Villagra, on the precipitous walls facing SW. and SE., c. 620 m, about a dozen specimens seen (past fl. *°/12 16, no. 196). A very distinct and conspicuous species of the well known insular type of rosulate dwarf-trees; in spite of growing close to the most frequented spot THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 159 I h t h Fig. 21. a—g Eryngium bupleuroides: a \eaves, $ nat. size; b involucral leaves, outside and inside, nat. size; ¢ bracts, outside and inside; d flower; e id. in fruit; f sepal (inside); g petal; c—gX5. h—p &. énaccessum.: h leaves, } nat. size; i involucral leaf (inside), nat. size; k bracts, inside and outside; 1 fruit, m mericarp from the back, n sepals, o petal, p stamen; k—p X 5. q—u E. fernandezianum: q leaves, } nat. size; r involucral leaves, inside and outside, nat size; s flower, t bract, u sepal; s—u X 5. in the island it seems to have escaped the attention of all earlier visitors. The place where it was found is a perfectly perpendicular and thoroughly inaccessible 160 CARL SKOTTSBERG rock wall, and it cost us considerable trouble to obtain sufficient material. The photograph on Plate 12 is of a complete specimen that had been lowered down from the cliff. Area of distribution: Endemic in Masatierra. 88. E. fernandezianum Skottsb. Stud. 17, Fig. 5. — Fig. 21 q—u. Masatierra: Portezuelo de Villagra, c. 600 m, two specimens at the foot of the crest west of the pass (past fl. 74/12 16, no. 188). Described from a fragment without flowers or fruit, collected by me in 1908, as far as I can understand taken from one of the two bushes seen in 1917. They bore flower heads, but petals and stamens were gone. No ripe fruits were encountered. A complete description is given below. Arbuscula glaberrima 0,7; m saltem alta, erecta, pseudodichotome ramosa, ramis 5—1IO mm crassis, junioribus creberrime annulatis. Folia ad apices ramo- rum confertissime rosulata, firma et crassiuscula, basi tenuiora subamplexicaulia, nervis perplurimis parallelis percursa, nervo centrali magis conspicuo et in pagina inferiore leviter incrassato, linearia, subspathulata, acuta, margine in parte dimidia vel tertia superiore serratodentata dentibus mitibus utroque latere 5—9 (plerumque 6 vel 7), 6—12,3 cm longa et 11—25 mm lata, basi g—12 mm. Inflorescentia terminalis monocephala, scapo 5—6 cm longo et 3 mm crasso superne cavo. Involucrum circ. 18-phyllum phyllis longissime triangulatis acutis basi incrassatis, margine inferne scariosis, supra colore fuscoviridi in violaceum spectante, subtus viridibus, ad 20 mm longis et 5 mm latis. Capitulum circ. 17 mm altum et 20 latum, viridi-violascens. Receptaculum cavum caverna 6—9 mm alta et 8—12 mm diam. Bracteae lineari-subspathulatae, apice trun- cato-rotundatae, basi incrassatae, 5 mm longae et I,5—2 latae, hyalino-albidae, apice violaceae. Flores confertissimi omnes jam deflorati, 6 mm longi et 3 mm diam. Sepala ovata, rotundato-acutata, margine scariosa, viridia, apice viola- scentia, 2,;—3 mm longa et I,5—1I,7 mm lata. Petala... Stamina... Stylo- podium crateriforme obscure viride. Styli 4,5—5 mm longi virides. Mericarpium immaturum 3,5 mm longum, viride, sectione triangulare, apice circumcirca appen- diculatum, dorso et margine appendiculis humilioribus ornatum. I am strongly inclined to regard this as a bastard between £. bupleurotdes and /. zzaccessum. This idea did not occur to me when describing the species, as one of the supposed parents was unknown at that time. In most characters it is intermediate between the two. The cortex is more as in &. zvaccessum, but the internodes sometimes slightly longer. The leaves are exactly inter- mediate in shape and size (see fig. 21), and so are the marginal teeth in shape and number; they are not pungent. The midrib is more conspicuous than in £. inaccessum, where all the principal veins are of the same strength, but less prominent than in £4. dupleuroides. The scape and receptacle are hollow as in the former. The involucral leaves are intermediate in shape, size and number, being green in L. dbupleuroides. The head stands between the two in shape and size and so do the sepals. The fruit is more like that in 4. dupleuroides. No ripe seeds were found. In anatomical structure the fruit takes an inter- mediate position. Both the supposed parents grow close to the place where £. fernandezstanum was found, and its great scarcity speaks rather in favour of THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 161 a hybrid origin. It is certainly one of the very rarest plants in the island. Still, no proofs that it is a bastard can be given at present. The following figures illustrate the intermediate position: = - | Number Size of Number of Number} of = i - , Leneth eerhisn of in- involucral |Size of bracts,| Size of sepals, eae volucral leaves peed De leaves ) mm mm ih leaves mm E.. bupleuroides 8—18(10—11)) 10—12 |10—20X3 3,5—4 X1,s—2) 2—2,5XI—1,2 6 E. fernandezianum)5— 9 (6—7) 16—18 20%4—5 5X 1)s—2125—3.. X1s—1,7| 45—5 E. tnaccessum I— 4 (2—3) | 18—23 |20—25*4—5| 7—8X2—3 3—3,5X1,5—2 ZA Area of distribution: Endemic in Masatierra. 89. E. sarcophyllum Hook et Arn. — JoHOW, Estud. 101; SKOTTSBERG, Stud. 16, Fig. 4, Taf. 5 Fig. 5. Masafuera: coast cliffs, very local: CuMING! JoHOow! — Between Vara- dero and Tierras Blancas (fr. 77/2 17, no. 402); Rodado del Sandalo. Area of distribution: Endemic in Masafuera. Apium L. go. A. fernandezianum Joh.; JoHOw, Estud. 101; SKOTTSBERG, Stud. 17, Ranis VFis. 6. Masatierra: Tres Puntas (JOHOW); barren cliffs in B. del Padre (fr. 7/1 17, no. 291; beg. fl. °/s 17, BACKSTROM no. 291 b; also observed by JOHOW and by the writer, 1908); Tierras Blancas (fl. °/s 17, BACKSTROM no. 1219). Cultivated in a temperate house this species develops exceedingly well and flowers and fruits freely. The branches, which are decumbent and root at the nodes, attain a length of one meter. The fruit is typical of the genus and agrees with that of A. graveolens in structure. A. fernandeztanum is related to A. prostratum Labill., widely dispersed in the Southern Hemisphere; it differs above all in the shape and cutting of the leaves. Area of distribution: Endemic in Masatierra. Ericaceae. Pernettya Gaud. gt. P. rigida (Bert.) DC. — Jonow, Estud. 87. Masatierra: Generally outside the forest but rare below the wooded region; one of the most important shrubs on the rocky ridges, where it forms low thickets. It does not, however, thrive well in the dry western section. We found it on all the ridges to the highest altitude reached (c. 800 m), and lam sure it is a leading species on the top of the Yunque. — Fi. 8/42 16, no. 27; II — 20100. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 162 CARL SKOTTSBERG 8/15 16, no. 89; 4/1 17, no. 259; fr. '"/4 17, no. 623. — A form with larger and thinner leaves was collected at the edge of the Dzcksonia-forest on C. Salsi- puedes (buds */12 16, no. 77), another with exceptionally narrow leaves among rocks near this place (fl. °/12 16, no. 86). Masafuera: Scattered in the higher parts, locally abundant. Ridges and plains above the Chozas village, 600—1350 m (fl. *°/2 17, no. 490); slopes of Los Inocentes, 480-—1400 m, not uncommon. Area of distribution: Endemic. Convolvulaceae. Dichondra Forst. #92. D. repens Forst. Masatierra: Pto Frances, by the stream, c. 40 m. Masafuera: Outer part of Q. de las Casas, among grass (fr. ?"/2 17, no. 551). New for Juan Fernandez. A rather typical D. repens; the young leaves are sericeous, the older sparingly pubescent below and glabrous above, the peduncles 2 cm long or even longer. Compare HALLIER in ENGLER’s Jahrb. XVIII (1892) 83. REICHE, Fl. de Chile V.173 includes all forms under D. re- pens Forst. I suspect this to have been collected by BERTERO, for MONTAGNE, Prodr. Fl. Fern. 356 enumerates »Uredo Hydrocotyles Bert. Hab. ad folia Hydrocotyles?» No Hydrocotyle has ever been reported from the islands; sterile Dzchondra may perhaps be mistaken for a species of that genus. Area of distribution: Subcosmopolitan (Eastern Asia, India, Africa, Mascarene Islands, Australia, New Zealand, North and Central America, South America to Patagonia and Chile, Juan Fernandez). Calystegia R. Br. 93. C. tuguriorum R. Br. — Syn. C. Hantelmanni Phil., JoHow, Estud. 85. Masafuera: Q. de las Casas (no. 412, also observed by JoHOW); Q. de las Vacas, abundant in the outer part; Q. del Varadero. Forming dense carpets on the canyon walls. Both J. D. HOOKER and HEMSLEY united C. Hantelmanni with C. tugurto- rum, but JOHOW kept them separate, declaring that the latter is a small herb, not attaining the size of C. sepium. However, CHEESEMAN, Man. N. Z. Flora 476 states that the stem is slender, prostrate ov climbing, often clothing trees and shrubs to a considerable height. During his monographical studies HAL- LIER came to the conclusion that the two species are identical [ENGLER’s Jahrb. XVI (1893) 548], and I have found it better to follow him. Area of distribution: New Zealand; Chatham Islands; South Chile (Val- divia); Masafuera. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS . 163 Boraginaceae. Selkirkia Hemsl. 94. S. Berterii (Colla) Hemsl. — Jonow, Estud. 85. Masatierra: On the steep rocky slopes above the forests in the centre of the island, rare. Co Damajuana, 500—550 m, very rare; the gap between Damajuana and Co Yunque, 580 m, few specimens (past fl. '*/12 16, no. 132); northwest face of Co Piramide, narrow rock ledge, c. 600 m, few shrubs; Porte- zuelo de Villagra, near the pass (also observed by JOHOW), a couple of speci- mens, 570 m; Q. del Monte Maderugo, rock wall 390 m, rare (past fl. 74/1 17, no:~341). Area of distribution: Endemic in Masatierra; monotypic. Verbenaceae. Rhaphithamnus Miers. 95. Kh. venustus (Phil.) Skottsb. — Syn. Citharexylon venustum Phil. Anal. Univ. 1856; Rt’. longiflorus Miers, Trans. Linn. Soc. XXVII (1870) 98; JoHow, Estud. 78. Masatierra: in the wooded region, extending from Pto Frances to the south slope of Co Chumacera, hardly ever seen below 200 m and much more common higher up; also in the shrubberies on the ridges. Fl. */12 16, no. 11; ae. uo. 40; 79/1216, nor 198; fr. 74/2 17, no. 628, *4/4 17, no. b. — A form with red flowers was found near Plazoleta del Yunque, c. 250 m. Masafuera: GUAJARDO without locality. — In the forests, much more scarce than in Masatierra; on the Sanchez plain, 515 m; Q. del Mono, 475 m; Q. del Blindado, 440 m (fl.-fr. *°/2 17, no. 516). Area of distribution: Endemic. Labiatae. Cuminia Colla. While HEMSLEY, Challenger Report 51 lists three species, erzantha, fer- nandezta and brevidens, JOHOW reduces brevidens to fernandezia. I have seen the material on which C. drevidens was based by BENTHAM, and I think that JoHow is right. The length of the calyx lobes is submitted to some variation; I admit that they are very short in C. drevidens, but otherwise no perceptible difference is found. HEMSLEY ascribed unisexual flowers to Cwumz7nia, JOHOW denied the cor- rectness of this statement and suggested that HEMSLEY was misled by the pronounced dichogamy of the flowers. My examination of numerous living 164 : CARL SKOTTSBERG specimens shows that there are two kinds of flowers, differing in size and in development of the anthers. Female flowers have shorter corolla, exserted style and small staminodes, bisexual flowers have a larger corolla, included style and normal stamens. I believe there are specimens with one kind only, but others exhibit both kinds in the same inflorescence. Flowers with perfect anthers were called male by HEMSLEY. There does not, however, seem to be any difference in the style or stigma between them and the female flowers. Still I have observed cases where the former were shed soon after they had opened. In a few cases female flowers with one or two fertile stamens were met with. JOHOW's statement that Cuminia absolutely lacks »esencias etéreas» is not in accordance with my experience, for C. evzantha at least, has a weak but unmistakable labiatoid smell. 96. C. fernandezia Colla. — JoHow, Estud. 81. — Fig. 22. Masatierra: In the open forests of the ridges and along the rocky crests, never abundant and only once observed below 300 m. — Between Q. de la Fig. 22. Cuminia fernandesia: a male or bisexual, b female flower, with limbs slit open to show stamens and staminodes. 4. Piedra Agujereada and (. Laura (f. magis pilosa); in the higher parts of El Rabanal (also quoted by JoHOw; fl. 7°/3 17, no. 576, f. magis pilosa); slopes of Co Damajuana, 350—530 m; V. Colonial, C. Central (fl. 18/1 17, no. 307); Portezuelo de Villagra (also observed by JOHOW), 540--590 m, scarce (beg. fl. *°/iz 16, no. 192, fr. *"/3 17, no. 192 b); Q. del Monte Maderugo, rocky ridge, 390—500 m; C. Salsipuedes (also observed by JOHOW) 400—625 m, not uncom- mon (beg. fl. *°/12 16, no. 165); Pto Ingles, central ridge, 570 m; between La Vaqueria and Q. Juanango, c. 300 m; ridge west of Co Yunque, c. 550 m; low hill near the camping place in B. Villagra, c. 200 m (fl. ‘/1 17, no. 234); Co Chumacera, south slope (fl. °/1 17, no. 356). Length in mm of calyx corolla anthers style THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 165 Generally perfectly glabrous save for the corolla and some hairs on the calyx lobes. The young leaves are, however, more or less pubescent. No. 576 differs from ordinary /ernandesta in the persistent indumentum on _ petiole, nerves and pedicels, a character otherwise attributed to ervzantha only. This species has larger and broader leaves of a more dull green colour, light violet to nearly white corolla with almost white lobes and filaments, while the flowers of fernandezia are of a bright blue lilac colour. Area of distribution: Endemic in Masatierra; the genus is endemic. 97. C. eriantha Benth. — JoHow, Estud. 82. Masatierra: with the former, but much more scarce. Between Q. de la Piedra Agujereada and Q. Laura, c. 500; El Rabanal (JoHow); V. Colonial, C. Central, c. 500 m; Portezuelo, on the Villagra side in forest (also observed by JoHow), c. 500 m, rare (fl. 7°/1 17, no. 278); C. Salsipuedes (also JoHow), in Dicksonia-grove, 660 m, rare (fl. 1°/1 17, no. 282). Length in mm of calyx corolla anthers style J 7 12—I13 0,75 Vie) - 5 8—9 0,3 Tos Area of distribution: Endemic in Masatierra. Solanaceae. Solanum L. 98. S. fernandezianum Phil. — JoHow, Estud. 82. Masatierra: BERTERO! GERMAIN! — In the wooded region, in moist humus, very local. Pto Frances (JoHOW); C. Chifladores, near the Frances Valley; El Rabanal, in the valley; Northeast slope of Co Damajuana, c. 500 m; V. Colonial, Q. del Monte Maderugo, 235 m (fl.-unr. fr. **/1 17, no. 340); forest on the east slope of C. Salsipuedes, c. 350 m; Pto Ingles (also quoted by JoHOW), central ridge, 470 m; Q. Villagra, near the road just below the pass, c. 550 m (fl.. 2/12, 24/12 16, no. 12, fr. April 1917; beg. fl. °?/7 17, BACKSTROM no. 12 b). The species was described as lacking tubers, but on BERYTERO’s label is written »tubercula gustu amaro». This is very puzzling, for JOHOW could not find the tubers, nor were we able to discover any. In order to find out if they develop during the winter, I told Mr. BACKSTROM to dig for them, but he could not find them. I have raised several specimens from seeds; they have flowered and fruited, but behave like annuals and do not form any tubers. I have propagated them from cuttings, using the basal axillary shoots. Ina wild state the plant branches profusely and seems to be perennial. Area of distribution: Endemic in Masatierra. 99. S. Robinsonianum Bitter in FEDDE’s Repert. XI (1912), 7.— Syn. S. furcatum, JOHOW, Estud. 83. 166 CARL SKOTTSBERG Masatierra: PHILIPPI 1864! — El Rabanal (JoHow); Pto Ingles, west branch, 190 m (fl.-fr. ?°/1 17, no. 311); B. Villagra, ravine near the camping- place, c. 180 am) (fliefr.’ "/1- 177;%10)9260). Santa Clara: JoHow. — Morro de los Alelies, rare (fl.-fr. 7°/1 17, no. 343). *Masafuera: near the Chozas village (fl. 17/2 17, no. 506); Q. del Vara- dero (fl.-fr. 12/3 17, no. 508). — New for this island. Professor BITTER kindly examined my dried material as well as living plants raised from my seeds and communicates the following new description: »Addenda ad descriptionem: Herbaceum, rectum, ca. 40—60 cm altum, divaricatim ramosum; rami Ca. 4 mm diam., obtuse angulati, lineis decurrentibus parum prominentibus instruct, pilis simplicibus pluricellularibus acutis curvatim) accumbentibus primo crebris serius sparsioribus obsiti; internodia in ramis florentibus 5—8 cm longa; folia superiora false geminata, inaequalia; petioli 1,;—2 cm longi, jam a basi vel fere a medio sensim latius alati; laminae ovatae vel late ovati-lanceolatae, majores 835 -5,9—0,5:5,5-0 usque ad 14: 8)5 cm, minores 5:3, 6% 357, 7: 457,534 oem basi late cuneatim vel subrotundatim in petiolum alatum abeuntes, infra medium latissimae, ad apicem versus magis sensim angustatae, subacutae vel obtusius- culae, dentibus grossis late triangularibus in utroque-latere ca. 5-7 obtusis vel obtusissimis instructae, membranaceae, utrinque virides, subtus vix pallidiores nitidioresque, utrinque praecipue in venis, sparsius in mesophyllo pilis simpli- cibus acutis pluricellularibus curvatim accumbentibus crebris obsitae; vena media et venae later. prim. in utroque latere 7—8 (raro 9—11) curvatim ascendentes subtus prominentes; inflorescentiae laterales, a foliis remotae, 7—11-florae; pedun- culus 12—25 mm (tandem — 27 mm), longus, semel furcatus; rhachides cr. 6—(tandem)g mm longae, flores plerumque densiuscule in rhachidum apicibus secuti; pedicelli 5—11 mm longi, in statu florifero erecti; calyx campanulatus, cr. 4:3 mm, profunde in lobos lineari-lanceolatos acutos inaequilongos (exteriore ceteris longiore). 2—-3:0,8 mm_ partitus, extus sicut pedunculus, rhachides et pedicelli pilis brevibus simplicibus pluricellularibus acutis accumbentibus prae- ditus, intus praecipue in parte connata glandulis minutis breviter stipitatis crebris obsitus; corolla stellata, diam. 15—17 mm, extus violacei-suffusa, ceterum albida, in parte interiore stella flavescente striis tenuibus obscuris (fusci-violaceis) radianti- bus insignata, profunde in lobos lanceolatos acutos 8: 2,5 mm extus pilis simpli- cibus acutis brevibus crebris obsitos intus pilis brevibus similibus in vena media tantum sparsis, in parte apicali crebris praeditos partita; filamenta paulum inaequi- longa, I,s—2 mm longa, gracilia, intus pilis pluricellularibus simplicibus vel rarius semel ramosis acutis densis instructa; antherae anguste ellipsoideae, 3:0,8 mm, utrinque emarginatae, poris introrsis subapicalibus obliquis; ovarium subglobosum, I mm diam., glabrum; stylus stamina manifeste superans, 6 mm longus, fere a basi cr. ?/s longitudinis pilis densis simplicibus pluricellularibus acutis sensim minoribus patentibus obsitus; stigma styli apice paulum crassius, obtusum, subbilobum; calyx in statu fructifero auctus, diam. 7—8 mm, lobis latius lanceolatis cr. 3—4:2 mm subacutis vel obtusiusculis baccae accumbenti- bus; baccae maturae globosae, non satis magnae, ca. 5—8 mm diam., parum nitidae, tandem nigerrimae, succo intense purpurei-violaceo impletae; granula THE PHANEROGAMS OF ‘THE JUAN FERNANDEZ ISLANDS 167 sclerotica duo subapicalia subglobosa parva, ca. 0,6 mm diam.; semina ca. 15 — 16, oblique reniformia, valde applanata, I,6— 1,8: 1—1,3:0,5 mm, manifeste minute reticulata, in baccis maturis extus succo violaceo intense tincta.» Area of distribution: Endemic. #100. S. (Jorel/a) masafueranum Bitter et Skottsb. nov. spec. — Plate 14. The following description was written by Professor BITTER. Herbaceum, ad 1,75 m altum; rami robusti, + ve flexuosi, 4—6 mm diam., subteretes vel paulum obtuse angulati, lineis decurrentibus parum prominentibus instructi, pilis simplicibus pluricellularibus acutis parvis diaphanis curvatim accum- bentibus plerumque sparsis obsiti, tandem subglabrescentes; internodia in ramis majoribus elongatioribus 9—14 cm longa; folia solitaria vel superiora saepe false geminata, tunc inaequalia; petioli 1,;—4 cm longi, ad laminam versus sen- sim magis alati, pilis simplicibus pluricellularibus acutis parvis crebris obsiti; laminae late oblique lanceolatae vel oblongi-lanceolatae, infra medium Jatissimae, basi oblique cuneatim vel rotundatim in petiolum alatum abeuntes, ad apicem versus magis sensim angustatae, acutae vel fere sensim acuminatae, integrae, majores cr. 12:5, 15:6 usque ad 17,5:7 cm (an etiam majores‘), minores 9: 3,3 usque ad 12: 4,5 cm, laminae omnes herbaceae, utrinque sordide virides, subtus paulum pallidiores nitidioresque, utrinque in venis et in mesophyllo pilis simpli- cibus pluricellularibus acutis sparsis, in margine paulo densioribus obsitae; vena media, venae later. prim. 6 curvatim ascendentes et venae later. secund. illas reticulatim conjungentes albide flavescentes subtus prominentes; venae majores subtus sicut petioli et rami in statu sicco striolis brevissimis parum prominenti- bus densis praeditae (cellulae arena crystallina impletae!); inflorescentiae laterales, extraaxillares, fere semper paulum infra nodos ortae, simplices vel semel bre- viter furcatae, 5—8-florae; pedunculus 1,5;—(tandem in statu fructifero!)3 cm longus, rhachides 4—6 mm tantum longae, floribus ergo densiuscule secutis; pedicelli in statu florifero erecti, 6 mm longi, in statu fructifero basi deflexi, IO—II mm longi; calyx campanulatus, 3 mm longus, apice fere 3 mm diam., in dentes 5 ovati-lanceolatos obtusiusculos 1 mm longos 0,5; mm latos basi membranis diaphanis conjunctos partitus, extus sicut pedunculus, rhachis et pedicelli pilis paucicellularibus acutis curvatim accumbentibus densiusculis ob- tectus, intus glandulis minutis breviter stipitatis crebris obsitus; corolla alba, stellata, diam. 10-14 mm, in lobos 5 lanceolatos acutiusculos cr. 5:3 mm extus pilis simplicibus paucicellularibus acutis accumbentibus crebris in parte apicali densis et partim subramosis obtectos, intus in venae mediae parte superiore et in mesophylli parte apicali pilis simplicibus paucicellularibus obsitos partita; corollae tubus infra staminum insertiones cr. I,5 mm longus, glaber; filamenta I,5—(tandem)2 mm longa, intus (praecipue prope basim) pilis simplicibus pluri- cellularibus acutis densis ad apicem versus sensim sparsioribus praedita; antherae flavescentes, ellipsoideae, utrinque emarginatae, cr. 2,2 :0,8—I mm, poris introrsis apicalibus obliquis tandem paulum longitudinaliter dehiscentibus; ovarium sub- globosum, I mm diam., glabrum; stylus 5,;—6 mm longus, in parte apicali subgeniculatim incurvatus, paulo supra basim glabram cr. ”/g longitudinis pilis tenuibus patentibus fere omnibus simplicibus (perpaucis semel subramosis) pluri- cellularibus acutis densis sensim minoribus obtectus, in parte apicali glaber; 168 3 CARL SKOTTSBERG stigma subglobosum, styli apice manifeste crassius; calyces fructiferi parum aucti, diam. cr. 5 mm, lobis magis triangularibus baccae submaturae accumbentibus; baccae submaturae globosae, 6—8 mm diam.; semina oblique reniformia, valde applanata, 2:1,5:0,5 mm, minute reticulata; granula sclerotica in bacca non reperi. Masafuera: In forests. On the Sanchez plain, 515 m (fl. ?°/2 17, no. 526); near the Chozas village (fl.-unr. fr. 3/3 17, no. 363). Area of distribution: Endemic in Masafuera. Nicotiana L. 101. N. cordifolia Phil. — JoHow, Est. 83; SKOTTSBERG, Stud. 7, Wat. 16 .Picc.4. Masafuera: Coast cliffs close to the landing place near Q. de las Casas (fl. *°/2 17, no. 1220); Q. de las Vacas, on the canyon wall, not far from the entrance, scattered specimens (also observed by JOHOW and by the writer in 1908, only locality known hitherto; fl.-fr. 7°/2 17, no. 392); Q. Angosta, in the gorge; ©. del Varadero, numerous fine shrubs near the entrance; along the shore of Tierras Blancas, not uncommon (fl.-fr. '*/2 17, no. 399). Area of distribution: Endemic in Masafuera. Scrophulariaceae. Mimulus L. 102. M. parviflorus Lindl. — Jouow, Estud. 82. var. externa nov. var. — A plantis in Chile lectis differt foliis omnibus petiolatis (superioribus brevius sed semper distincte), margine sat grosse et irregulariter serratis, nec non pedunculis petiolum aequantibus sed foliis multum brevioribus. Planta perennis pilosa. Masafuera: GERMAIN! — OQ. de las Casas (also observed by JOHOW), wet places under overhanging rocks etc.; Q. de las Vacas (fr. *%/2 17, no. 497); Q. Angosta, at the waterfall; Q. de la Loberfa (fl.-fr. ’"/2 17, no. 486). It is hardly possible to ascertain the right position of the insular form unless all the material from the continent is revised. GERMAIN’s specimens were labelled MW. parviflorus var. by PHILIPPI, and JOHOW remarks that the Masafueran plant differs from the continental one in the short pedicels and in the pubescence. As all specimens collected in Masafuera are of the same kind, while there is no similar form in the collections from Chile, I have described the former as a variety. It is not impossible that J7. pzloszusculus Kunth from Peri is the same, but as I have not seen authentic material and the descrip- tion is very brief, I must leave this question open. Area of distribution: J7. parviflorus is common in central and south Chile. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 169 Euphrasia L. “103. E. formosissima nov. spec. — Plate 15, fig. 1, Plate 20, fig. 6—7; text:/fie..:23) Eueuphrasia pluriennis lignosa, plerumque 15— 20, interdum usque ad 70 cm alta, ad radices Pernettyae rigidae parasitica. Truncus primarius cylindrico- tetragonus, ad 6 mm crassus, cortice obscure fusco inferne glabrato, fere a basi Opposite ramosus, ramis ramulosis ramulis iterum ramosis, junioribus distinctius tetragonis bifariam hirsutis, pilis crassis plerumque bicellularibus. Folia carno- sula late — anguste ovata, 8—18 (rarius ad 20 et ultra) mm longa et 5—8 mm lata, basi in petiolum 2—3 mm longum subiter vel sensim angustata, margine leviter revoluta dentibus humilibus latissime truncatis utrinque 3—5 (plerumque 4) instructa, supra subglabra, parcissime glandulosa sed secus nervos pilosiuscula, subtus in parte inferiore ad nervos parce pilosa vel subglabra, infra marginem et inter nervos incrassatos reticulo glandularum humilium capite ferrugineo pulchre notata. Bracteae quam folia caulina minores ac latiores. Flores ad apicem caulis ramorumque breviter sed densiuscule spicati, sessiles. Calyx pro genere sat typicus, lobis lineari-subspathulatis apice incrassatis, extus nervis exceptis subglaber, intus secus nervos pilosus pilis unicellularibus et glandulis ferrugineis commixtis, 5—6 mm longus, profunde partitus. Corolla alba 10—13 mm longa, tubo 4—5 mm longo ad finem anthesis haud elongato, subglabro, intus inferne obscure luteo; labium superius extus pilosum, 6—7 mm longum et 5-——6 latum, bifidum lobulis reflexis; labium inferius glabrum 7—8 mm longum et ad 10—II mm latum, in centro macula magna flavo-aurantiaca distincte cir- cumscripta notatum. Filamenta staminum posticorum sub 3 mm, anticorum circ. 3.5 mm longa, inferne lutea, dein viridescentia, antherae ad 1,4 mm longae, ferrugineae, subglabrae-glabrae, duorum staminum posticorum loculus unus paulo longius calcaratus. Stylus ad 9 mm longus, dilute viridis, parce pilosus, exsertus; stigma alboviride capitatum. Capsula matura valvis obovatis minute emarginatis et abrupte apiculatis, superne setulosa, 2,s—3 mm longa et 2—2,: mm lata. Semina oblique fusiformia I—1I,: mm longa straminea, testa longitudinaliter sulcata et inter juga transverse sculpta. Masafuera: in the higher parts, not uncommon in the alpine region. Ridges above Q. Sanchez (fl. 7°/2 17, no. 386b); north of the Casas gorge from 800 to 1400 m (fl.-fr. '*/2—‘/s 17, no. 386); C. Atravesado, common; C. del Barril, from 750 m (fl. */3 17, no. 547). A rather interesting discovery. Two types of Euphrasza are known from the southern hemisphere, sect. 77zfidae from Chile and sect. Lweuphrasia sub- sect. Australes from Australia and New Zealand. The new species has nothing to do with the South American species, and as far as I can see, it is not nearly related to the Australes, but belongs to the Semzcalcaratae, known before only from the northern hemisphere. Within this section /. formosissima is the only perennial species. Among boreal ones it might be compared with Z£. grandiflora Hochst. (Azores), or with species from Eastern Asia. It lies near at hand to regard it as one of the older endemics in the islands, as it differs 170 CARL SKOTTSBERG from all other species in its mode of growth (see below) and as it stands so far apart from all the congeners on the south hemisphere. The arrangement of the glands on the lower leaf surface is illustrated in fig. 23 b. These glands are of two kinds, smaller with a short, narrow, uni- cellular stalk and a globose two-celled head with dark brown contents and larger (much less numerous), sessile, like a semiglobose papilla, divided by a radial wall and with yellowish contents. Both kinds were described by WETT- STEIN in his monograph of the genus, p. 18—-19. The long pedicellate glands so common in many other species seem to be wanting. The smaller kind also occurs on the upper side of the leaf and on the calyx. Fig. 23. Euphrasia formosissima: a leaves, showing general outline, b lower side of leaf, Cc ts, d lower side of bract, a—d X 2; e two calyx lobes, inside, X 4; f stamens, X 124; g capsule, X 6,6; h seed, X 20. The growth of other perennial Luphrascae was explained by WETTSTEIN l. c. 14. The shoots die off after hawing produced flowers, while lateral branches, which are sterile the first season, become floriferous the next year, and so forth. £. formosissima represents another type. To some extent it is like many others, for the primary stem and a number of side branches may produce flowers simultaneously, sometimes even a few shoots of a third genera- tion flower at the same time. It occurs that a branch dies after having flowered, but very often this is not the case: the branch survives, the top continues to grow above the floral region and produces a new set of flowers, and this may be repeated again; thus, the new species differs from all others. If, as WETTSTEIN thinks, the perennial species are genetically older than the annual, /. formosissima seems to exhibit the most primitive type of Ewphrasia known. Area of distribution: Endemic in Masafuera. THE PHANEROGAMS OF ‘THE JUAN FERNANDEZ ISLANDS 171 Plantaginaceae. Plantago L. 104. P. fernandezia Bert. — JoHow, Estud. 77. Masatierra: In the shrubberies along the high ridges in the centre of the island, rare. North face of Co Damajuana, 530 m, few specimens; the depres- sion between this peak and Co Yunque, 580 m, rare; V. Colonial, C. Central, 570 m, very few plants; Portezuelo de Villagra, c. 550 m, near the path in two places, perhaps twenty specimens altogether (fl. *°°/12 16, no. 7; only locality known before, JOHOW etc.) rock ledges on the perpendicular wall of Co Piramide, ©, 575)m: Professor PILGER sent me the following notes on this species: »P. fernandezia ist mit P. princeps Cham. et Schlecht. von den Hawaii- Inseln verwandt: sie ist dieser Art nicht nur habituell ahnlich, sondern zeigt auch grosse Ubereinstimmung im Bliitenbau. Fiir eine entferntere Verwandt- schaft konnten auf dem amerikanischen Festland hochstens einige Arten der Cleistantha-Gruppe (P. Candollei, oreades, Sodiroana) in Betracht kommen. Zwei Bliitenstande der Exemplare von P. fernandezia zeigen eine mehr oder weniger weitgehende Verkiimmerung der Staubblatter. Die normale Form hat Bliiten mit lang herausragenden Antheren. Zu dieser Form sind Uber- gange vorhanden, indem die Antheren etwas zwischen den dauernd aufrechten Corollenzipfeln sich hervordrangen und + reichlich Pollen entwickeln. In anderen Bliiten bleiben die Antheren von den aufrechten Zipfeln umschlossen, sie sind ziemlich gross, entwickeln aber keinen Pollen. Der Griffel ragt aus der Bliite hervor. Man k6énnte an einen Vergleich mit dem Dimorphismus bei der Cleistantha-Gruppe denken, doch ist der Fruchtknoten in den Bliiten mit redu- zierten Staubblattern nicht stark entwickelt wie dort bei den geschlossenen Bliiten, sondern bleibt klein und von der Ro6hre frei.» Area of distribution: Endemic in Masatierra. 105. P. truncata Cham. subsp. Skottsbergii Pilger. — P. Skottsbergiz Pilger in SKOTTSBERG, Stud. 6. Masatierra: B. del Padre (SKOTTSBERG 1908). *Santa Clara: scattered on the table-land (fr. 78/1 17, no. 350). Professor PILGER remarks to this species: »P. Skottsbergii ist eine charak- teristische Form, die allerdings der P. truncata subsp. firma sehr nahe steht. Am besten wird man sie neben P. firma auch als subspecies zu P. truncata stellen». Area of distribution: P. truncata is known from Central and South Chile, the subspecies is endemic in Masatierra and Santa Clara. Rubiaceae. Oldenlandia L. *106. O. thesiifolia K. Schum. — Syn. O. uniflora Ruiz. et Pav. on L. Masatierra: V. Colonial, in the small stream at the foot of the chapel hill. 172 CARL SKOTTSBERG Rather like SCHUMANN’s illustration in Flora Brasil. VI: 6, tab. 127. Agrees very well with specimens from Chile. Glabrous with few hairs on the ovary. Undoubtedly of recent introduction, as it has not been recorded by previous visitors in spite of growing in Cumberland Bay. There is perhaps no reason to regard it as introduced with the human traffic; the valley is sometimes visited by stray birds from the continent, and they very likely brought seeds of this and some other species in mud adhering to their feet. Area of distribution: Brazil, Argentina, Paraguay, Uruguay, Chile from Valparaiso to Chiloé, Masatierra. Nertera Banks et Sol. 107. N. depressa Banks et Sol. — SKOTTSBERG, Stud. 6. Masafuera: almost confined to the subalpine and alpine districts, where it occurs in moist places, round large stones, in small depressions in the ground etc., generally associated with mosses. — Q. de las Casas, in wet moss at the waterfall, 215 m; high land north of Casas, not uncommon from 750 to about 1100 m (fr. 17/2 17, no. 413, 7*/2 17, no. 413 c); among rocks at Las Torres, 1370 m (fl.-fr. no. 413 b); C. del Barril, 650—750 m. Area of distribution: Mexico to Fuegia; subantarctic—circumpolar; Hawaii; Masafuera. Coprosma Forst. 108. C. triflorum (Hook. et Arn.) Benth. et Hook. f. — JonHow, Est. 73. — Vig. 24 a—c. Masatierra: in the open forest and in the shrubberies along the ridges, not uncommon from 3—400 m, ascending to the highest peaks and ranging from Pto Frances to Co Chumacera: Fl. ¢),*4/sa) 16, no. .263,44/19 16, sn0tSo. fl. 2 1/12 16, no. 151 (a few fruits from 1916 still on the trees); past fl. ¥ Gliel MOM AG tk. “osttT iNOrs Oheow la M7. Olle Mistaken for /7/cppotis triflora Ruiz. et Pav. by BERTERO, Ann. sc. nat. XXI. 347; called Psychotria? triflora Hook. et Arn., Bot. Misc. III. 359, P. Hookeri G. Don Gen. Syst. Gard. HI. 585. The flowers have not been described, but BENTHAM and HOOKER FIL. recognized it as a member of the genus Cofrosma, Gen. plant. II. 139. Generally strictly dioecious. Corolla of ¢ 6—7 mm long, fleshy, greenish brown, as a rule 4-cleft; 4 stamens with thin filaments, 6—7 mm long, anthers 4,5—5 mm with thick, brown connective. Female flowers with 2,; mm long ovary; corolla dark lilac brown with narrow tube measuring 2—3 mm and four strongly revolute narrow linear segments, 2—2,; mm long. Styles white, 9 mm long. In one case a number of bisexual flowers were found on a branch of a female tree. They have a large, normal ovary with the usual long styles. The corolla is. more as in the male flower, but only 4—-5 mm long; there are 4 stamens with 2,5 mm long anthers but very short flaments. The pollen contained numerous sterile grains. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 173 Very like C. foliosum A. Gray from Oahu (Hawaii), which differs in the narrower leaves and in the shape of corolla and stamens. Area of distribution: Endemic in Masatierra. 109. C,. pyrifolium (Hook. et Arn.) Skottsb. — Syn. Psychotria pyrt- folta Hook. et Arn., JoHOw, Estud. 74. — Fig. 24d. Masatierra: An important forest tree, extending over the entire wooded region from Pto Frances to Co Chumacera and ranging from 200 m, or perhaps less, to 650 m at least. We did not find flowering specimens (sterile, '°/12 16, mo, 161, *°/12 16, no. 1222, 18/1 17; no\'308); but ripe fruits were gathered, °/, 17, no. 600. Masafuera: Scattered through the forested region; lowest altitude observed 280 m (fr. 17/2 17, no. 435); not uncommon up to 950 m. One small tree was Fig. 24. a—c Coprosma triflorum, a male, b apparently bisexual, c female flower; d C. py77- folium, male flower. All X 3. found in the alpine fern-beds, c. 1200 m, not far from the Correspondencia Camp (fl. ¢ °/3 17, no. 501); also in some of the deep gorges, by the stream, not much above sea level, f. i. in Q. de las Casas (past fl. 2 14/2 17, no. 499). — Specimens from open and sunny stations (nos, 499 and 501) differ from others by the firmer, bright green leaves. The removal of Psychotria pyrifolia to Coprosma is quite welcome from a geographical point of view. BERTERO |. c. suggested that it was a mere form of his Azppotts; HOOKER and ARNOTT, |. c. 360, called it Psychotrza?, while HEMSLEY and JOHOW dropped the question mark. It is dioecious and the structure of the flower is the same as in C. “7florum; accordingly, it differs widely from Psychotria. The male flower has not been described before; it is brownish green with yellow anthers. The corolla measures 8—g mm and is deeply cleft with linear, acute, reflexed lobes; the stamens are inserted at the base of the corolla tube; the filaments measure 8—g, the anthers 5—6 mm. C. pyrtfolium, sometimes rather like the preceding, is a much larger tree with thinner and broader, more pyriform leaves, indistinctly serrulate in the upper half and with narrower meshes in the net of venules. As was pointed 174 CARL SKOTTSBERG out by JOHOW, it is also distinguished by the presence of domatia on the leaves. It also differs in the somewhat larger and more obovoid fruit as well as in the shape of the male flower, as seen from my figures. C. pyrifolium is related to C. laevigatum Cheesem. from Rarotonga and to other species from the same region, than to its congener in Masatierra. and probably more nearly allied to these Area of distribution: Endemic. Galium L. *r10. G. masafueranum nov. spec. — Fig. 25. Perenne, sat pusillum, + decumbens. Radix primaria tenuis persistens, multiceps. Caules numerosi, tenuissimi, fragiles, inferne ramosi, repentes, ad \ va ) . x ; i v / Fig. 25. Galium masafueranum: a branch, nat. size; b lower side of leaf, X 24; ¢ in- florescence and d flowers, * 5; e normal, unripe fruit; f ripe fruit with appendages; g mericarp from ventral side with do.; h fruit with appendage on pedicel. e—h x 10. nodos radicantes, saltem ad 2 dm longi, manifeste 4-lineati et secus lineas dense patente-retrorse setoso-hispidi. Folia qua- ternaria, subsessilia vel brevissime petiolata, elliptica, plerumque manifeste mucronata sed non pungentia, secus marginem revo- lutum nec non in nervo mediano subtus longe setosa, ceterum pilis nonnullis in- spersa vel subglabra, 4—g mm longa et 2—5 mm lata (plerumque 6 X 4 mm), in- feriora jam emarcida conspicue minora. Inflorescentiae laterales breves foliis sub- aequilongae ~ duplo longiores; pedunculus setosus, + curvatus, denique apice recur- vus, ad 13 mm longus, apice 2—4-foliatus, plerumque 1—3-florus, floribus brevissime pedicellatis. Calyx nullus. Corolla luteo- viridis, 2,;—3 mm diam., lobis mucronu- latis 1,2—I,3 mm longis et 0,8—0,9 latis quattuor rarissime quinque, passim setosis. Stamina 0,5 mm longa. Stylus fere ad basin bifidus, ad 0,6 mm longus. Discus viridis incrassatus. Ovarium longe hispi- dum. Fructus 2,5; mm latus et 2 mm altus, cum setis ad 3,5 X 3 mm metiens. Meri- carpia_ reniformia, dense _ setoso-hispida, setis rectis vel paulo curvatis nunquam uncinatis; superne interdum appendicibus longiusculis (ad 2 mm) apice nudis ceterum setosis instructa. Masafuera: In the alpine region, trailing in the moss mats, probably not uncommon. Western precipice, a short THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 175 distance north of Buque Varado, c. 1230 m (fl.-fr. ‘/3 17, no. 367 b); fell-fields near the Correspondencia Camp, c. 1200 m; in moss among the rocks of Las Torres, 1370 m (fl-fr. **/2 17, mo. 367); C. del Barril, 1110 m (fr. 1/3 17, no. 530); Los Inocentes, 1375 m. The curious appendages do not appear to have been found in any other species. Their nature is problematic. In the cases where each carpel ends in such a tail-like process it would lie near at hand to regard them as enlarged, persistent styles, but this explanation is impossible, as they are situated outside the disc; inside this the traces of the styles will be found. In several instances as many as three were found on one carpel (fig. 25 g): in two cases such an appendage was found attached to the pedicel (fg. 25h). Finally, many plants did not have any at all. G. masafueranum is related to a number of Andine species, such as G. Juegianum Hook fil., ucinulatum DC., canescens Kunth and andicola Krause. The latter three are larger plants with many-flowered inflorescences and with much longer hairs on the fruits and leaves; G. fwegzanum has larger leaves, long pedicellate flowers and a glabrous stem, and all of them have barbed setae om the fruit. Area of distribution: Endemic in Masafuera. Campanulaceae. Wahlenbergia Schrad. In Monogr. Camp. (1830) 160 ALPHONSE DECANDOLLE described IV. fernan- desiana A. DC. Four years later CoLLA (Mem. Accad. Torino XXXVIII. 118, tab. 35) described Campanula Larraini Bert. ined. COLLA stated that, to judge from DECANDOLLE’s diagnose, C. Larrazni was different. The material quoted by DECANDOLLE and collected by Mrs. GRAHAM, DOUGLAS and SCOULER shows that his species comprised several forms, for Mrs. GRAHAM collected W. Grahamae Hemsl. and what I call W. fernandeziana s. str. and DOUGLAS and SCOULER collected W. Larraznz. This explains why A. P. DECANDOLLE, Prodr. VIII. 438 referred W. Larraini to fernandeziana as a synonym. The question of nomenclature becomes complicated by the fact that there are two types among BERTERO’s plants, both under no. 1443. I do not know if COLLA did see both, anyhow, his description and plate answer very well to one of them, and this I have retained here as IW. Larrainz. DC. Prodr., HEMSLEY and JOHOW write Wahlenbergia Larraini Bert., Colla, but COLLA did not admit the genus Wahlenbergia, but brought his species to Campanula. HEMSLEY, Challenger Report 45 rejects Larraznz, but describes the new species Grahamae, based upon a part of Mrs. GRAHAM's material. JOHOW admits one species only, W. fernandeziana A. DC. For some reason he compares the narrow-leaved species (my Larraznz) with IV. Berterot, which is a quite different thing; he declares that he found all sorts of transitions between Lar- raint and Grahamae. It is true that there are forms apparently intermediate between Grahamae and fernandeziana s. str., but nobody who has seen these 176 CARL SKOTTSBERG species growing wild should propose to unite Larrazn? with Grahamae. JOUOW believes that the size and shape of leaves and flowers undergo so great changes with the external conditions that all differences will be explained in this way. He writes p. 75: »En los terrenos himedos i sombrios se encuentran ejemplares robustos con hojas muy anchas que convienen con la descripcion de la W. Grahamae Hemsl.» — but this species is at home on the steep rocky ridges and grows exposed to full sunlight; nor is it true that the typical /ermandeszana srows »en los lugares ménos fértiles». All the species are small shrubs with strong, persistent root. The in- florescence is terminal, and the flowering branches die in the autumn, at least the upper, floral part, while innovations are developed from their lower leafy part or from older branches. Key to the species formerly united under IV. fernandeztana. I. Corolla campanulate, white, with narrow contracted lobes; calyx segments narrow, entire; leaves lanceolate or linear-lanceolate, + acuminate. W. Larraini. Il. Corolla + broad campanulate to funnel-shaped with dark purplish veins; calyx segments dentate. Leaves lanceolate-ovate. A. Leaves glabrous or nearly so, narrower toward the base: corolla campa- mulate with w= serectnlODes <2. ee aa seco 8 eee ee eee WV. fernandesiana. B. Leaves pubescent, with broad base, semiamplexicaul; corolla very broad=with: broad: revolute lobes. =... we W. Grahamae. 111. W. Larraini (Bert., Colla) Skottsb. — Syn. Campanula Larraini Bert. ined., COLLA |. c.; JoHow, Estud. 75, sub IV. fernandeztana. — Fig. 26 a-—e. Masatierra: BERTERO! no. 1443 ex p.! SCOULER! DouGLas! — On the low, dry ridges between the valleys; Pto Frances, Loma Incienso, in shrubberies, 360 m; Q. de la Pesca de los Viejos, c. 225 m, in a small ravine (fl. 3/19 16, no. 127); C. Salsipuedes, one plant in the height of c. 75 m, several higher up, c. 465 m (beg. fl. 8/12 16, no. 51 — f. parce pilosa); between Q. Juanango and La Vaqueria, c. 300 m. — A form with spread hairs on the leaves was also collected by SCOULER and DouGLas! Suffrutex erectus plerumque glaberrimus, dense ramosus, ramis suberectis sat dense foliosis. Caulis glaber. Folia coriacea, lanceolata vel lineari-lanceo- lata, 20—30 mm longa, 4—8 mm lata, basi angustata, apice acuminatim pro- ducta, argute serrata, supra glabra vel secus medianum pilosa, subtus glabra vel rarius pilosiuscula. Flores contracte paniculati. Calyx glaber lobis angustis- simis acutis, integris vel rarissime denticulo glandulaeformi uno vel altero munitis. Corolla + anguste campanulata, glabra vel secus marginem loborum pilosiuscula, alba, 14—15 mm longa, fauce 6 mm diam., expansa ad 15 mm lata, lobis angustis erecto-patentibus apice convolutis, 6—7 mm longis et basi 4 mm 1 BERTERO named the plant in honour of a certain Mr. LaRRaIN, who seems to have communicated some species from Juan Fernandez. ‘This gentleman is not included in the list of collectors compiled by Jonow, and I have not seen herbarium labels with his name. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 177 latis. Stamina 3,2—3,5 mm longa, filamentis triangulato-linearibus ad 1,; mm longis, basi setis nonnullis instructis, antheris ad 1,8 mm. Stylus 4 mm longus, sub apice incrassatus et parce pilosus. Area of distribution: Endemic in Masatierra. 112. W. fernandeziana (A. DC. ex p.) Skottsb. -- Jouow, Estud. 75. — Plate 20, fig. 9; text fig. 26 i—l. Masatierra: Mrs. GRAHAM! BERTERO no. 1443 ex p.! Gay! SkKorTs- BERG 1908 no. 417! — On the higher ridges in the centre of the island, or in open stations below the forest. North face of Co Damajuana, 530 m; Co Pira- mide, near Portezuelo, abundant in one place; V. Colonial, C. Central, 570 m, Fig. 26. a—e Wahlenbergia Larraini: a leaves (no. 51), lower and upper surface; b same of no. 127, upper and lower surface; ¢ flower, d stamen, e pistil; a—c nat. size, d—e xX 5. f—1 W. fernandeziana. f leaf of no. 267 (upper surface), g no. 325 (lower surface), h 305 (do.), i flower, k stamen, 1 pistil; f—i nat. size, k—l1 X 5. m—q W. Grahamae: m—n leaves, upper and lower surface, o flower, p stamen, q pistil; m—o nat. size, p—q X 5. with the following (beg. fl. *8/1 17, no. 305); Pta San Carlos, near the sea, rare (fl. °/1 17, no. 267); Pto Ingles, central ridge (fl. '9/1 17, no. 325); ridge between Co Yunque and Villagra, c. 200 m (fl. '/1 17, no. 264). A f. e/ata with long, thin, less densely foliate branches and very lax inflorescence was collected near Portezuelo (fl. *°/3 17, no. 1223). Suffrutex erectus subglaber, ramis curvato-erectis dense foliosis, caule ple- rumque glabro. Folia subcoriacea, lanceolata vel lanceolato-ovata, sessilia, basi et apice sensim angustata, dense et argute serrata, reticulatim nervosa, glabra vel supra secus medianum puberula, 15—30 mm longa et 5—g mm lata. In- florescentia + laxe paniculata, floribus longe pedicellatis, pedicellis glaberrimis, bracteis foliosis ovato-triangularibus. Calyx glaber apice violaceus lobis triangu- latis 5—6 mm longis basi 2—2,5 mm latis, margine parce glanduloso-denticu- latis. Corolla ampla campanulata, extus puberula, 15—16 mm longa, aperta I12— 20100. The Nat. Hist, of Juan Fernandez and Easter Isl. Vol. II. 178 CARL SKOTTSBERG usque ad 20 mm diam., fauce 9 mm diam., alba, inferne pulchre purpureo- venosa, lobis 6—7 mm longis basi 4—4,5 mm latis, paulo reflexis, planis, mar- gine puberulis. Stamina filamentis 3 mm longis a basi lata subiter angustatis, valde setosis; antherae 2 mm longae. Stylus ad 5,5 mm longus, versus apicem incrassatus nec non dense setosus, lobis stigmatiferis margine et subtus pilosis. Area of distribution: Endemic in Masatierra. 113. W. Grahamae Hemsl. — JoHow, Estud. 75 sub fernandeziana, — Plate 15, fig. 1; Plate 20, fig. 10; text fig. 26 m—q,. Masatierra: Mrs. GRAHAM! MOSELEY! — Mountain ridges in the centre of the island. V. Colonial, C. Central, 570 m (fl. 18/1 17, no. 306); Portezuelo de Villagra, near the pass and abundant in places, 550—600 m (fl. 17/12 16, no. 1226; fl. 24/12 16, no. 1224, f. minus scabra ad praecedentem vergens; fl. 10/, 17, no. 1225, f. laxior); B. Villagra, rock wall at the back of Q. de la Choza, c. 400 m. No. 1226 is the most extreme Grahamae that I have seen. Of the others nos. 1224 and 1225 have smaller flowers, measuring only 20 mm across. The possibility of bastards occurring between this and the former should be considered. Suffrutex effusus, vage ramosus ramis prostrato-adscendentibus longis sub- flexuosis, pubescenti-setosis foliosissimis. Folia chartacea, sessilia, ovato-lanceo- lata vel ovata, acutiuscula, basi lata fere semiamplexicaulia, dense sed interdum indistincte denticulata, utrinque pubescenti-scabra, subtus in nervis dense pilosa, 20—35 mm longa et 10—16 mm lata. Inflorescentia pseudo-umbellata, foliis triangulato-ovatis instructa, haud multiflora sed ob magnitudine florum valde conspicua. Calyx glaber lobis triangulatis 5—7 x 2—3 mm, inferne dentatis. Corolla late campanulata vel sub fauce leviter constricta fere globoso-campa- nulata, extus puberula, alba, eximie purpureo-venosa, cum lobis 13—15 mm longa, expansa ad 25 mm lata, fauce 16—17 mm diam., lobis revolutis late triangularibus 7—8 mm longis et 8—10 mm latis, margine puberulis. Stamina filamentis 3—4 mm longis a basi lata valde angustatis; antherae I1,5—2 mm longae. Stylus 6,;—8 mm, versus apicem incrassatus et cum pagina inferiore stigmatorum setosus. Area of distribution: Endemic in Masatierra. 114. W. Masafuerae (Phil.) Skottsb. Stud. 6. — Syn. W. tuberosa Hook. fil., JoHOW, Estud. 78. Masafuera: GERMAIN! GUAJARDO! — coast cliffs from B. Toltén to Vara- dero and Loberia, rather common; also in the outer part of the canyons, on rock ledges and in crevices. Only once observed in the highlands, north of Casas, c. 1200 m. Fl.-past fl) Feb. 1917, ‘no. 428. f. rosea: A typo differt corolla praecipue basin versus pulchre rosea. — Masafuera: Q. de la Loberfa, one large specimen found among the typical plants (fl. *7/2 17, no. 480). When describing Luphorbia Masafuerae (Bot. Zeit. XI1V.647) PHILIPPI felt sure of the generic position, but later he began to doubt the correctness, as shown by GUAJARDO’s specimen, labelled in PHILIPPI’s hand-writing: »Eu- phorbia? Masafuerae Ph. (nonne potius Wahlenbergia?) Masafuera 1869 Leybold>. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 179 In Stud. 6 I quoted this species from Masatierra also. There were two sheets of a Wahlenbergia in Herb, Lund; they had no labels, but on the sheets was written in J. G. AGARDH’s hand-writing: »Juan Fernandez, D. DOUGLAS no. 48». As D. never visited Masafuera, 1 concluded that they came from Masatierra. They did not belong to the species known before from Masatierra; thus, only IV. Masafuerae remained, or a new species. Habitually they looked like the former, but differed in the closer serrature of the leaves; that were glabrous and had a somewhat different venation. The basal parts of the speci- mens were missing. Comparing these specimens a second time with the rich material of W. Masafuerae now at hand, I found that the differences quite forbid an identifica- tion. I further found that, apart from these specimens, nothing like them had ever been collected in Juan Fernandez, nor did such a species exist among DOoUGLAS’s plants in Herb. Kew. During my visit to that herbarium last year I had a look at the Wahlenbergiae from Saint Helena, and was at once struck by the great likeness between W. “éntfolia A. DC. and the plants in Herb. Lund. Some time ago I sent for both and was able to establish their identity. The examination of a capsule of the alleged DOUGLAS plant revealed that it is 2-valvous as in all the St. Helena species, while all Juan Fernandez species have 3-valvous capsule, Certainly we can draw no other conclusion than that the statement on the Lund specimens is quite false. And if we go back to DOUGLAS’s journal, only lately published, we read under no. 48: »Campanula(?) leaves alternate, sessile, lanceolate, serrate, smooth above, slightly pubescent underneath» etc. DOUGLAS no. 48 is W. Larraini. The person who distributed the plants in question to Lund has made a rather dangerous mistake. Area of distribution: Endemic in Masafuera. 115. W. Berteroi Hook. et Arn. — JOoHOw, Estud. 75. — Syn. Campa- mula gracilis var. revoluta Colla |. c. 119. — Plate 20, fig. 8. Masatierra: BERTERO no. 1440! »ad rupium fissuris in montibus editiori- bus et in Goats island». Rediscovered in Masatierra by us, but not in the »higher mountains»: B. Cumberland, Pta San Carlos, two small specimens (fl. 9/1 17, no. 265); in the sand outside the caves, one large and beautiful plant (A. 7/12 16, no. 43). Both localities near the sea. Santa Clara: BERTERO (»Goats island»); JoHow! — Morro de los Alelies (also observed by JOHOW), fr., some few fl. left, *°/1 17, no. 345. W. Berterot has (just as W. Masafuerae) a large, lactiferous subterranean bulb, formed by the base of the primary stem together with the primary root. From the surface of this bulb spring the vegetative-floral shoots, each appar- ently of one year’s duration. The shape of the corolla is unlike that in other species. Its colour is a deep rosa, the base crimson. Style and stamens are rose-coloured, anthers golden yellow. The style is much longer than in other species, the stigma visible in the mouth of the corolla. As the figure of HooKER and ARNOTT in HOOKER’s Journ. I (1834) plate 137 is incorrect as to the flower, another was prepared. The flower figured here was trimerous throughout with 6 sepals etc., an anomality sometimes observed. Area of distribution: Endemic in Masatierra and Santa Clara. 180 CARL SKOTTSBERG Lobelia L. 116, L. anceps Thunb. — JoHOow, Estud. 76. — Syn. L. rupicola Colla, Mem. Accad. Torino XXXIX (1834) 120. Masatierra: El Pangal, wet rocks near the waterfall, 215 m (fl. 7/1 17, no. 221); Pto Ingles, moist rocks near the sea (fl.-unr. fr. ?°/1 17, no. 314; also found by BERTERO and by JOHOW). Masafuera: coast rocks near Q. de las Casas, where water trickles down (also observed by JOHOW); in the Casas canyon, humid rock wall; in the interior of Q. de las Vacas (fl.-past fl. *°/2 17, no. 496); Q.del Varadero, at the waterfall in the small branch of the canyon. A variable species that ought to be studied on material from different countries. In Juan Fernandez, the colour of the flower ranges from a deep lilac blue to almost white. REICHE, FI. Chil. V.87 writes »planta anual?», but the Juan Fernandez plant is perennial, as usually stated. Area of distribution: South Africa, Australia, New Zealand, Chile from prov. Maule to Valdivia; Juan Fernandez. Compositae. Lagenophora Forst. 117. L. Harioti Franch. — Syn. L. hirsuta, SKOTTSBERG, Stud. 5. — Fig. 27 f—i, 28 f—g. Fig. 27. a—b Lagenophora hirsuta, orig. PoEPPiG: a involucral scales, b ray floret. c—e L. nudicaulis from Falkland Isl., leg. SKOTTSBERG: c, d involucral scales, e ray floret. f—i L. Harzoti from Masafuera: f involucral scales, g ray floret, h expanded ligule, i disc floret. — All X 10. Masafuera: Alpine fell-fields, scarce. Highland north of Casas, from 1130 m to the top of Las Torres, 1370 m (fl.-fr. 14/2 17, no. 484); C. del Barril, c. 830m, rare, 9725 m more common (fl.-fr. 1/3 17, no. 536); Los Inocentes, ¢. 1375 m. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 181 The insular plant is certainly identical with Z. HYariof# Franch. Miss. sci. Cap Horn V. 344, of which I have seen type material from Herb. Mus. d’hist. nat. in Paris. Another question is if this is different from ZL. xudicaulis (Comm.) P. Dus. (fig. 27 c—e, 28 a—b). ZL. Harioti (fig. 28 c—e) is larger and more slender, the leaves less crowded, more distinctly dentate and always pubescent. The involucral scales are larger, c. 4 mm, rarely exceeding 3 mm in xmdzcaulis. The flowers are slightly larger, but the achenes look very much the same in both. There may be as much reason to regard LZ. Harioti as a form of the other, in which case the name var. gract/zs (Alboff) should be used. ALBOFF, Contr. Fl. Terre de Feu II. 27 placed his variety under L. hirsuta Less. It has Fig. 28.. Achenes of Lagenophora. a,b L. nudicaulis, a orig. COMMERSON, b ex herb. Link (Berlin). c—g ZL. Harioti: ¢ Cord. Rio Manso leg. Retcue, d Cord. Linares leg. REICHE, e Fuegia leg. SKoTYSBERG no. 151, f Masafuera, g Masafuera, dorsal view. h—m J. hirsuta. h orig. Porpric, i Cord. Rio Manso leg. REICHE, k—1 Cord. Valdivia leg. NEGER, from both sides and from the back, m Cord. Nahuelbuta leg. ReicHe. n—o L. pumila: n leg. BasTIAN, o leg. Cockayne no. 79. p—q L. fetiolata, p leg. COLENSO, q leg. Haasr. — All X 16. nothing to do with this. ZL. Aérsuta (fig. 27 a—b, 28 h—m) is known by the shape of the leaves, that are larger and more deeply crenate, and especially by the long beak of the achene, which also has strong lateral nerves absent in the others, as seen from fig. 28 h—m; these nerves are also visible in the figure of PoEPPIG and ENDLICHER, Nov. gen. et spec. plant. I tab. 26, though not mentioned in the descriptions. I have figured one achene of the original material (fig. 28 h). L. Harioti has been mistaken for Azrsuta on account of the hairy leaves, also by the writer (Zur Flora des Feuerlandes, Wiss. Erg. Schwed. Siidpolar-Exp. IV: 1: probably all the specimens quoted belong to ZL. Hariofz). FRANCHET regarded his species as »bien différente de ZL. Commersonii (= nudicaulis) et voisine de LZ. Forster’ (= pumila)»: comp. fig. 28 n—o. Still, it is much nearer to the former. ZL. pumila differs in the shape of the leaves and bracts and also 182 CARL SKOTTSBERG in the smaller achenes with their very short beak. The achene of L. Hariot is more like that in L. fetcolata Hook f. (fig. 28 p—q). Both pumila and petio- lata come from New Zealand. All the species in question form one small natural group. The type material of Z. Harioté was out of flower: ALBOFF describes these as light rose-coloured. In the plant from Masafuera the colour varies from white to rose, turning + crimson with age. I have seen L. Harioté (determined as hirsuta or sometimes nudicaulis) from the following stations: Fuegia, Ile Clarence, Sholl Bay, HArioT 1°/5 1883 (type); Ushuaia, in pratis alpinis '’/s 1902, SKOTTSBERG no. 151; Chile, Cord. Linares, 2200 m, Jan. 1897, REICHE; Cord. of Rio Manso, goo—1400 m, Feb. 1896, REICHE; Cord. of Chillan, 2000 m, Jan. 1895, REICHE. Area of distribution: Andes of South Chile to Fuegia; Masafuera. Erigeron L. In the large collections of Santiago, Kew etc. Ihave not met with any near relatives of the island species. It is noteworthy that not one of the numerous alpine species, described from Chile by PHILIPPI, seems to be related to the new forms from the mountains of Masafuera. In order to get the opinion of a specialist, I asked Professor VIERHAPPER to examine my material and to compare my new species with the collections in the Vienna museums, what he most willingly did. He thinks, and here I quite agree with him, that all the island species, in spite of their different habit, are related to each other; 4. rapecola is very unlike the rest, but the principles of its morphology are the same as in the others. Professor VIERHAPPER mentions the following species as more or less related to the insular forms: £. lepidotus Less. (Hawaii), spzculosus Hook. et Arn. (Chile), w/¢givosus Benth. (Columbia) and heteromorphus Rob. (Mexico). HEMSLEY compared £&. fruticosus (Juan Fernandez) with £. Darrellianus Hemsl. (Bermuda). Prof. VIERHAPPER adds: »Als besonders wichtiges gemeinsames Merkmal erscheint mir die + tiefe Serratur der Blatter, die z. B. bei spzculosus genau so ist, bei /epzdotus allerdings nicht, der aber dafiir sonst dem /ruticosus sehr nahe zu stehen scheint». 118. E. fruticosus DC. — JoHow, Estud. 51; SKOTTSBERG, Stud. 4. — Fig. 29 a—c. Masatierra: Rocky and stony ground on the high ridges or near the sea, also in the dry western parts of the island, but not reported from the eastern section. North-east corner of El] Yunque, c. 580 m; V. Colonial, C. Central on the crest, 570 m (also JoHow); Cave no. 6, one specimen; Portezuelo de Vil- lagra (also observed by JoHOW), 500—600 m, not uncommon (fl. 8/12 16, no. 9); Q. del Monte Maderugo, road-side in the macal, 240 m, two specimens; C. Salsipuedes, c. 600 m (fl. 8/12 16, no. 73); between La Vaqueria and Q. Juanango, rather abundant; south slope of Co Tres Puntas; C. de las Cabras, 380 m; B. Chupones, among rocks in the grass (fl. °/1 17, no. 255). THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 183 Masafuera: Scattered on the ridges and in the canyons; Q. de las Chozas, Dicksonia-grove; Q. de las Casas (also observed by Jonow), rare; Q. Inocentes, c. 500 m, abundant; ©. Angosta; Q. del Varadero, rare; slopes above Las Chozas, 700—750 m, in the grass (fl. **/2, 7°/s, no. 524); C. del Barril, c. 750 m. f. majuscula n. f. — Fig, 29d—e. — A typo differt statura majore et magnitudine omnium partium. Rami ad 70 cm longi et 6,5 mm crassi, dense foliosi. Folia ad 18 cm longa et 2 cm lata, sat profunde lacerata. Panicula amplissima. Flores 2 ad 7 mm longi, tubo 3, limbo 4 mm. Achaenia ut in typo. Masafuera: Q. de las Casas, one specimen in the bottom of the gorge in a shady corner (fl. °/3 17, no. 375). — Looks like a »gigas-form». Aster erigeroides Hook. et Arn, (=E. fruticosus) 8 proliferus Comp. Bot. Mag. II. 48 is a large-leaved form, but nothing is said about the size of the heads etc.; it is described »corymbis quasi Jateralibus», an expression difficult to understand. Area of distribution: Endemic. Fig. 29. a—e Evigeron fruticosus: a outer and inner scales, b ray and ¢ disc floret; d—e f. ma- juscula, ray floret and achene. f—i &. /uteoviridis, f outer and inner scales, g—h ray and disc floret, i achene. — All X 6. *119. E. luteoviridis nov. spec. — Fig. 29 f—i. Euerigeron suffruticosus polycephalus subglaber luteoviridis. Radix pri- maria persistens lignosa longissima (ad 58 cm longa visa), valde ramosa, multi- ceps. Caulis lignosus plerumque a basi ramosus; innovationes basales internodiis inferioribus elongatis, ceteris brevissimis folia rosulatim conferta gerentibus. Rami florigeri 20—30 cm longi, rigidi, glabri, striati. Folia inferiora dense rosu- lata, anguste spathulata vel lineari-lanceolata, acuta, in petiolum semiamplexi- caulem longissime attenuata, plerumque 5— 8 cm longa et 4—6 mm lata, versus apicem dentibus acutis parvis incrassatis utrinque circ. 5 munita, reticulatim nervosa nervo mediano permanifesto, pallida, lutescenti-viridia, subglabra (pilis nonnullis marginalibus) vel glaberrima, valde firma quasi coriacea; folia caulina inferiora similia, superiora linearia, integra, 2—4 cm longa et 2—3 mm lata. Capitula ex axillis supremis orta, subumbellatim disposita, pauca — sat nume- rosa, circ. 8 mm diam., inferiora pedunculis elongatis 3—5 cm longis, bracteo- latis, simplicibus vel infimis parce ramosis, glabris vel parce pilosis suffulta. Involucrum 5—5,5 mm altum, 4,;—5 mm diam., subglabrum. Squamae lanceo- 184 CARL SKOTTSBERG latae, acutae, hyalino-marginatae, minute laceratae nec non parce piligerae, paucae exteriores breviores 2,;—3 X I mm, intermediae numerosae 4,5—5 X I,2 mm, intimae angustiores glaberrimae. Flores radii omnes & ligulati albi glaber- rimi, tubo circ. 2 mm longo, ligula angusta suberecta apice indistincte lobulata circ. 3 mm longa; flores disci § corolla 3—3,5 mm longa lutea. Achaenia ovata compressa subglabra, 1,5 mm longa et 0,e—0o,7 mm lata. Pappus circ. 2,5 mm longus. Masafuera: In the alpine region; edge of western precipice, above Buque Varado, c. 1200 m (fl.-fr. ‘/3 17, no. 558); C. Atravesado, c. 1350 m (fl 1°/. 17, no. 500). No. 558 differs from the type (no. 500) in the broader, more coarsely serrate leaves. Near &. fruticosus, but undoubtedly quite distinct. The light colour and the thick leaves characterize the new species; /. fruticosus is pure green, much more villous, and has much thinner leaves, generally also a distinct trunk, which attains the height of a few feet. Area of distribution: Endemic in Masafuera. *120. E. Ingae nov. spec. — Plate 16, fig. 1; text fig. 30 a—d. Euerigeron perennis basi suffruticosus hirsutus oligocephalus. Caudex sat profunde infossus lignosus di- polychotome ramosus, radicibus numerosis chordi- formibus; innovationes basales longi subrosulatim foliosi. Caules florigeri ad 30 cm vel ultra alti, sat robusti, adscendentes—erecti, sulcato-striati, pilis longis albis pluricellularibus + dense villosi. Folia basalia rosulata, infima tempore floris jam emortua, spathulata, in petiolum basi dilatatum longe angustata, acuta vel obtusata et mucronata, superne grosse serrato-dentata dentibus late triangu- latis mucronulatis utrinque I—3(—4), sat obscure viridia, undique sparse et longe villosa, 4—7 cm longa, 8—12 im lata; caulina sensim minora, spathu- lato-linearia, breviter petiolata, margine dente uno vel altero instructa, suprema sublinearia, 2—4 cm longa et 3—5 mm lata, integra. Caulis sub capitulo ter- minali leviter incrassatus. Capitula terminalia et ex axillis supremis I—3 (rarius ad 5) lateralia, pedunculis tenuibus villosis ad g—12 cm longis caulem superanti- bus suffulta, 16—18 mm diam. Involucrum 7—8 mm altum et 10 mm crassum. Squamae, exterioribus paucis brevibus exceptis, 6,;—7,5 mm longae et I—I,2 latae, lanceolatae, acutatae, margine et dorso villosae, intimae margine irregula- riter hyalino-laceratae, parce pilosae—glabriusculae. Flores 2 ad g mm longae, albae, corolla 7—8 mm longa, tubo perparce villoso. Flores ¥ 5,5 mm longi lutei. Achaenia valde compressa, margine incrassato setoso, in latere basi et apice pilis nonnullis inspersa, 2—2,; mm longa et I mm lata, pappo circ. 2 mm longo. Masafuera: Las Torres, on rock ledges, 1370 m (fl. '°/2 17, no. 391); Los Inocentes, c. 1350 m (fl.-fr. °/3 17, no. 556). — Named in honour of Mrs. INGA SKOTTSBERG. var. Innocentium nov. var. — Fig. 30 e—h. — A typo differt foliis con- spicue latioribus, basalibus spathulatis 4—7 cm longis et 1—2 cm latis, late rotundato-obtusis, mucronulatis, margine grosse sinuato-dentatis, caulinis ovato- lanceolatis subintegris vel integris, 2—2,5 cm longis et 7—12 mm latis. Inflore- THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 185 scentia magis contracta. Involucri squamae 5—6 mm longae; flores 2 7 mm, % 4 mm longi. Achaenia 1,5 mm solum longa, pappo aequilongo. — Masa- fuera: Los Inocentes, c. 1400 m (fl.-fr. °.3 17, no. 557). — Only very few specimens observed; habitually somewhat different on account of the broader leaves and also notable for the smaller flowers, but more material is required in order to decide whether it deserves the rank of a species. Area of distribution: Endemic in Masafuera. Fig. 30. a—d Evigeron Ingae: a largest outer, and inner scales, b ray and c¢ disc floret, d achenes; e—h var. /nmocentium, e outer and inner scales, f ray and g disc floret, h achene. Lee6s #101. E. turricola nov. spec. — Plate 16, fig. 2; text fig. 31 a—g. Euerigeron perennis caespitosus parvulus monocephalus. Rhizoma adscen- dens, radicibus numerosis fibrosis, multiceps. Folia carnosula firma, dense rosu- lata, spathulata, petiolo laminam aequante usque ad duplo superante, obtusato- rotundata et apiculata, dentibus utrinque 1—2 sat manifeste mucronatis ornata, 3—4 cm longa et 3—7 mm lata, saltem secus marginem pilosa vel undique hirsuta, pilis longis pluricellularibus. Scapus 5—12 cm altus, sulcato-angulatus, strictus vel leviter curvatus, subglaber—hirsutus, monocephalus vel rarissime ex axillo supremo capitulum alterum producens, perparce foliosus superne nudus, folia caulina 3—7 linearia acutiuscula 1—2,5 cm longa, summa vel interdum omnia valde reducta bracteiformia. Capitulum circ. 15 mm diam. Involucrum 5X6 mm. Squamae una vel altera brevis, triangulata 2,5 x I mm solum 186 CARL SKOTTSBERG metiens, ceterae 4—5 X I—I,2 mm, lanceolatae, acutae, denticulatae, dorso dense hirsutae; intimae subglabrae, margine solum dense ciliatae nec non scariosae et denticulatae. Flores 2 albi, circ. 9 mm longi, limbo angusto + distincte tridentato patente, 4,5—5 mm longo, tubo extus pilis nonnullis insperso. Flores 8 flavi, 4,.—5 mm longi. Achaenia + oblique ovata, compressa, margine in- crassato hirsuto ceterum subglabra vel sub apice pilis paucis bifurcatis inspersa, alpine» habit. The leaves are variable as to shape and degree of pubescence, but I think all the specimens collected belong to the same species. Area of distribution: Endemic in Masafuera. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 187 122. E. rupicola Phil. — JoHow, Estud. 51. — Fig. 31 h—k. Masafuera: GERMAIN! without locality. — Common on the coast rocks: between Q. del Ovalo and Q. del Pasto; entrance to Q. del Mono (beg. fl. !%/s 17, no. 417 b); between Q. de las Casas and Q. de las Vacas (beg. fl. 1° 1°/2 17, no. 417); Q. Angosta, in the canyon; Q. del Varadero (fl. 1*/3 17, no. 514b); Rodado del Sandalo, abundant in places (fl. '"/2 17, no. 514); Q. de la Loberfa, not uncommon. Seen once only in the alpine region, west of ©. Sanchez, 1100 m, one specimen (fl. ?°/2 17, no. 541). Of this peculiar plant only one small and incomplete specimen existed before, viz. the type in Herb. Santiago, so JOHOW believed that it was a very rare species, while in reality it is quite common in many places on the coast cliffs. As the description of PHILIPPI is incomplete, I shall add the following notes. The morphology is the same as in the other species without a trunk; there is a stout, woody primary root of the multiceps type; the stems are woody, profusely branched, more or less prostrate, forming dense mats. The upper part of the flowering shoots dies in the autumn, while te base persists and carries innovations. The leaves are mainly basal but seldom so densely rosulate as in the others; they are long spathulate, entire, very obtuse, fleshy and perfectly glabrous, but otherwise rather variable in outline, 3— 7 mm broad and about 2—4 cm long, of which about one half belongs to the petiole. The upper leaves are reduced to bracts. The branches end in a terminal head; numerous lateral ones are developed in the axils of the upper leaves, borne on very long, filiform pedicels with some leaves below and bracts above; sometimes these leaves support a third generation of heads. As a rule, the pedicels are glabrous; the upper part, however, slightly pilose in no. 541. Often enough all the branches are abbreviated; if elongated, they are prostrate — ascending and rise a few cm only above the ground. The head measures 7 mm across, the involucre is 4,5 mm high and 5 mm wide, the scales 4—4,5 mm long, glabrous, the outer setulose near the tip, the others quite glabrous with scariose denticu- late margin. The 2 flower measures 7, the § 5 mm. The achenes are typical, 1,5 mm long, hispid along the thickened margin and on the flat surface. Pappus 2—2,5 mm. Area of distribution: Endemic in Masafuera. Gnaphalium L. 123. G. spiciforme Sch. Bip. — Syn. G. spicatum Lam. var., SKOTTS- BERG, Stud. 5. — I*ig. 31 I—m. Masafuera: In the alpine region, apparently rare; first found by the writer in 1908. Co Correspondencia, c. 1350 m (past fl.-fr. 1°/2—*/s 17, no. 414); C. del Barril, 1290 m. The determination of this plant remains uncertain. Anybody familiar with the present state of confusion in the genus will understand the difficulty of classifying a species belonging to the purpureum assemblage. It is not G. spicatum; this wide-spread species has been introduced to Juan Fernandez and 188 CARL SKOTTSBERG is easily known from the alpine Graphalium of Masafuera. This is strictly perennial with numerous rosulate innovations; the scape is simple or very little branched and has few leaves. The leaves are greenish grey above, snowy white with long, dense wool below, and the same kind of long, crisp wool thickly envelops the scape and heads. The latter form a dense, short almost globular spike, surrounded by leaves; sometimes there are some lateral glomerules, sessile. or pedunculate. The compact globose spikes and the long crisp wool give to this form an appearance different from that of all forms of purpureum or spicatum that I have seen. The involucral scales number 15 to 20, all the outer are short and broad, 3 X I—I,2 mm, acute, greenish hyaline with dark brown apex; the inner are larger, 4X I—I,5 mm, obtusely apiculate to square, denticulate, green in the centre with broad hyaline edges and a dark brown or reddish brown tip. The head measures 4 X 2,5 mm. In G. sficatum the head is smaller and narrower (3 x I,s—2 mm), and not so dark, with all the scales acuminate or acute (fig. 31 n). The species was brought to G. spzciforme after a detailed comparison with the type, LECHLER no. 1250, of which I have seen material in Upsala, Kew and Marburg. It includes two forms, one very like my plant, another with a leafy stem, grey wool and sessile clusters of heads in all the upper axils. The latter is = G. mucronatum Phil. in Herb. Santiago; REICHE brought this to spzciforme, but I think it is better to regard mucronatum as distinct and to bring a part of LECHLER no. 1250 to this. I have seen the following specimens of G. sfzczforme, as limited above: ae Point, LECHLER no. 1250 ex p. (type), fig. 311; Port Famine, KING (s. n. G. falcatum, very like the plant from Masafuera); Orange Harbour, WILKES (s. n. G. consanguineum); Bay San Nicolas (CUNNINGHAM). DUSEN’s G. spicatum from Lake Argentino in Patagonia (no. 5742) is similar in most respects. To judge from our present knowledge of G. spzczforme, this is a magellanic species with an outpost on Masafuera. The sterile specimens of 1908 were compared with G. consanguineum Gaud. The fresh material clearly shows the difference between them, and consanguineum is brought to sfzcatum by most authors. Area of distribution: South Patagonia; Fuegia; Masafuera. Abrotanella Cass. *124. A. crassipes nov. spec. — Fig. 32 a—h. Pulvinatim caespitosa, caespite rotundato—subplano, denso nec non sat magno. Folia creberrime rosulata, linearia, haud marginata, curvato-patula, glabra, plerumque 10—15 mm longa, infra apicem 1I,5—2 mm lata; lamina car- nosa, supra punctulis depressis inspersa, apice rotundato-subconica, minutissime apiculata, versus basin leviter angustata deinde in vaginam subhyalinam ner- vosam intus glandulosam dilatata. Glandulae ferrugineae stipitatae clavatae multicellulares. Scapus in ramis terminalis, 1—2-foliatus, primo inter folia + absconditus, statu fructifero elongatus nec non incrassatus et carnosus, ad THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 189 IO—1I5 mm longus et 1!,;—2,; mm crassus, basi semper tenuis, leviter angulatus, monocephalus. Capitulum circ. 3 mm diam. Involucrum 6—8-phyllum squamis ovatis—latissime obovatis, obtusis, + truncatis, viridibus apice + roseo, margine hyalinis, uninerviis, basi praesertim glandulis (supra depictis) munitis, 2—2,2 mm longis. Flores obscure rubri, pauci—sat numerosi (ad 16 visi), circ. 3 mm longi: Q perpauci, ¥ magis numerosi, omnes fertiles, stylus floris ¥ distincte bifido fere ut in 2. Calyx nullus. Achaenia ovoideo-fusiformia, obscure tetragona, Fig. 32. Abrotanella; a—h A. crassipes: a branch, nat. size, b leaves, X 5, ¢ scales, X 17, d scape with head, X 2,5, e Q and f % flower, X 10, g achenes, X 17, h gland, X 150. i—k A. linearifolia, i scale and k achene, X 17. 1—m A. Mose/eyi, 1 scales, m achenes, X 17. apice leviter concava et margine humili circumvallata, glabra, 2—2,5 mm longa, pappo plane destituta. Masafuera: Only found on the top of Los Inocentes, 1325—1400 m, abundant in places (past fi.-fr. °/s 17, no. 502). A. crassipes is nearly related to A. /inearis Berggr. (New Zealand) and to A. linearifolia A. Gray (West Patagonia, Fuegia). It differs from the former in the shape of the flowers and achenes and especially in the leaves, which have a long pilose sheath in A. /mearis. I have not seen type material af A. /zneart- folia, but specimens from. Pto Churucca in Herb. Kew are rightly named and Igo CARL SKOTTSBERG agree very well with the type; only, the involucral scales are said to be »sub 2—3-nervatae» (Proc. Amer. Acad. V.137) and I find them 1-nerved (fig. 32 i). The achenes have a corona of 4—5 long bristles (fig. 32 k). These bristles are present in all the achenes, while there is no trace of them in A. crassipes. On the other hand, the glands described above seem to be absent. The plant found by the writer in Tierra del Fuego and named linearifolia (K. Sv. Vet.- Akad. Handl. LVI (1916), 318, Taf. 23, Fig. 27) differs from the ordinary one in its much more compact growth and in the much shorter, more canaliculate leaves. Perhaps it ought to be separated from “nearzfolia, but the coroniform pappus is the same. It is, however, possible that A. crassipes is not confined to Masafuera. In Herb. Kew I found an unnamed specimen, labelled »Otway Harbour, Gulf of Penas, Patagonia. Coll. Mr. MOSELEY Jan. 1876. REED June 1876». It is very like my plant; the leaves are a trifle narrower, the scape not at all in- crassate, in spite of the achenes being ripe, but some scapes in crasszpes are less elongated and incrassate than the rest. The scales (fig. 321) have the same general outline and there are a few glands of the same kind. The achenes are more distinctly clavate, without any beak at all (fig. 32 m). At present, I distinguish this as A. Moseleyi nov. spec. Further material is required to decide whether it is better to bring it to A. crassipes. Area of distribution: Endemic in Masafuera. Robinsonia DC. DECANDOLLE, Prodr. VI, established three sections: Symphyochaeta (RA. macrocephala), Eleutherochaeta (RX. Gayana and thurifera) and Eleuthero- lepis (XR. gracilis). In several respects R. macrocephala differs so much from the other species that the most modern taxonomists would reserve the name Robinsonia for this species alone. According to my opinion, we loose more than we gain by splitting up an isolated genus like this, the species of which are, without the slightest doubt, genetically related to each other but lacking closer connection with all other genera. That, in many orders, apparently small characters must be, and have always been, used to distinguish the genera, is one thing; but to make this a rule is wrong: this mechanical and spiritless splitting-up process may, perhaps, satisfy the ambition of the authors of all these »new genera» but it has become a burden to all other botanists. Thus, I retain Rodznzsonza in its old sense, distinguishing two subgenera, Symphyo- chaeta, comprising A. macrocephala, and Eleutherochaeta to receive all the other species, which have free pappus setae and deciduous corolla. I propose to subdivide the latter subgenus into two sections, Symphyolepis with soldered (R. Gayana and ¢thurifera) and Eleutherolepis with free involucral scales (R. eventa, Masafuerae, gracilis). There are some other differences between these sections. In Symphyolepis the base of the pappus forms a conspicuous coronula and the style of the ¢ flowers is shortly bifid with acute penicillate branches; in Eleutherolepis the pappus setae are free almost to their very base, and the style is deeply bifid with linear-truncate branches. Of the three species THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS IQ! of this section &. graci/is differs from the others by its small, narrow leaves and less constricted involucre with obsolete calyculus. Habitually, it is equally distant from all the other species. Key to the genus Robinsonia. I. Pappus of 2 flower forming a tube, including the persistent corolla. Leaves with a very dense net of small pellucid venules between the nerves. Subg. I. Symphyochaeta (DC.) Skottsb. R. macrocephala Dene. II. Pappus bristles free, caducous as also the corolla. Leaves with a much Boarser: net, of Veins, <.:.-.---..- Subg. II. Eleutherochaeta (DC.) Skottsb. A. Involucral scales soldered. Leaves distinctly acuminate. Sect. 1. Symphyolepis Skottsb. 1. Scales cohaerent almost to their tips. Inflorescence rounded — flat. Leaves c. 12 cm long, with numerous lateral veins almost parallel LE] OTC DVSSRRTS 616 oe MP ee PRO En Coe R. Gayana Dene. 2. Scales cohaerent to 7/3. Inflorescence pyramidal. Leaves c. 20cm long with lateral veins forming a conspicuous net with large meshes. R. thurifera Dene. B. Scales free. Leaves acute, but not distinctly acuminate. Sect. 2. Eleutherolepis DC. 1. Calyculus conspicuous. Leaves 8—15 cm long. a. Leaves fleshy with thick base, leaving semilunar, flat scars. R. evenia Phil. b. Leaves coriaceous with less thickened base, leaving linear, PROMMIMEM ty SCA tS ee soso on anne R. Masafuerae Skottsb. 2. Calyculus indistinct. Leaves about 3 cm long. A. gracilis Dene. 128. R. macrocephala Decne. — JoHow, Estud. 57. — Fig. 33 a—c. Masatierra: Pto Frances, Loma Incienso (called thus after the Robin- sonia), c. 360 m, scattered specimens (past fl.-fr. 1/12 16, no. 131, also observed by JoHow); hills between Pto Frances and Rabanal (JoHow); Co Centinela, steep slope near the entrance to El Pangal, some few shrubs c. 300 m (fr. A e195 no. 573). The pappus of this species is quite unique, forming a closed, tough and firm tube with a wrinkled surface. The achenes, glabrous in other species, are setose in this. The calyculus consists of few, minute bracts. The involucre in the @ measures 10—12 mm in length by 8—g across; the head is about 15 mm long and 10 mm broad. The scales number 17 or 18 and are firmly soldered. The ray florets are 5 to 8, but sometimes only 2 or 3, sometimes none at all, according to Jonow. They are about 10 mm long; the ligule is atropurpureous, boat-shaped, 2 I mm, not smaller than in other species in contrary to older statements. The disc florets are yellow with reddish lobes, 8—8,; mm long including the ovary. When the fruit is ripe, the involucre bursts open in a quite irregular fashion. The achenes are much larger than in other species, 7—8 mm. 192 CARL SKOTTSBERG Unfortunately, the flowers of the few male plants seen were all gone already on the 13th of, December, so they remain unknown. There were only few female heads still in flower and even these fairly advanced. DECAISNE, Ann. sc. nat. I. 28 writes: »Vidi specim. unicum in herb. Mus. Paris, a Claudio Gay relatum. Floret Januario. Fructum perfecit Februario». Gay visited Masatierra in February (JOHOW |. c. 38). His specimens are in fruit, and I do not think they flowered in January, as every specimen seen by us must have flowered already in November. According to JOHOw, only Gay and himself ever col- lected this species, but in DC. Prodr. VI. 448 we find the following statement: »leg, cl. LARRAIN et CLAUD. GAY»; comp. above, foot-note on p. 176. Area of distribution: Endemic in Masatierra. 126. R. Gayana Decne. — JOHOw, Estud. 58; SKOTTSBERG, Stud. Taf. 5, Fig. 1. — Fig. 33 h—k, 35 d—e. Masatierra: Common on all the ridges from Pto Frances to Co Chuma- cera; abundant in the higher parts where it forms a conspicuous feature in the vegetation of the rocky crests; occasionally found at lower altitudes (280 m on the north, 190 on the south side of the island). — Fl. Dec. 1916, nos. 22 and i232 ou 23.9e4 and 137.2. fr, Aiprilarol7,.no,-23)b. Male plant: Heads less numerous than in 2, 8—9 mm high, with fully expanded rays up to 12 mm across (larger than the female, contrary to the statement of JOHOW); involucre 5—6 x 4,5—5 mm; flowers yellow, ray florets generally 8, 6—7 mm long, ligules almost plane, 3,5—4 * 2—2,5 mm, style slightly bifid with short triangulate, acute lobes; disc florets 6—7 mm long, lobes of style incrassate, conical, abruptly narrowed with apiculate tips. Female plant: Heads as many as 100 in the corymb, 5—6 mm high by 4,5—5 across; involucre 4—4,5 mm in length and width; flowers greenish yellow, 7—8 ray florets 5 mm long, ligule + erect, boat-shaped, + bifid; disc florets 4 mm long not counting the exserted style, which is thick with small lobes. Achenes 2—2,5 mm long, those of the ray florets shorter and more curved. Both sexes with very few and small calycular bracts. Area of distribution: Endemic in Masatierra. 127. R. thurifera Dcne. — JoHow, Estud. 59. — Fig. 33 d—g, 35 a, f, g. Masatierra: BERTERO! SOHRENS! — On the ridges, among shrubs and small trees, rare. C. Chifladores, one specimen 350 m, numerous about 450 m (past fl. 17/2 17, no. 619 &, 620 g). Portezuelo de Villagra (also observed by SOHRENS), south-west slope of Co Piramide, 575—600 m, about 20 trees seen (fl. 24/1 17, no. 358 2, 367 J): one plant epiphytic on Blechnum cycadifolium. DECAISNE’s description refers to the female plant only, and the specimen brought by SOHRENS is also female. Apparently the male plant is described below for the first time. Heads less numerous than in 9, where there are several hundred in one panicle, 7 mm high, 10 mm across; involucre 4,5—5 by 3,5-4 mm, cylindric- campanulate, of about 13 scales, united to above the middle, glabrous with rather abrupt, penicillate apex. Ray florets generally 5 (sometimes 6), 7 mm long, ligule dirty yellow, 4 x 2 mm, tridentate, a little fleshy; style long exserted, THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 193 lobes ovate with penicillate apex. Disc florets 5 mm long, not counting the exserted stamens, yellowish green; stamens distinctly thickened below the anther, style a little exserted with orbicular-triangulate, slightly divergent stigmas, their tips penicillate. Pappus composed of very few bristles. The following notes complete the earlier short descriptions of the female plant. Head 4 mm high and 4,5 across, involucre only 3 < 3 mm, smaller than in all the other species, with few inconspicuous calycular bracts. Ray florets generally 8, sometimes as few as 5, yellowish green, only 3—3,2 mm long with a small (1,5 X 0,7 mm), boat-shaped ligule; style thick with short lobes, yellowish green; disc florets of the same colour, 3 mm long, not counting the exserted style; stigmas short, divergent. Achenes 10-costate acc. to DECAISNE, 5—10- costate acc. to Jouow; I found them 5—8-costate, ribs little prominent and not equidistant. The achene is crowned by a narrow ring much less conspicuous than in &. Gayana. On most of the achenes examined there are one or more thick, white stripes, which proved to be composed of the resin exuded on the stem of this and other Robinsonias, but not met with on the achene in any other species. These clumps of resin were observed by DECAISNE, l. c. 24: »le R. thurifera a ses fruits munis de cOotes trés saillantes et formées par une substance médullaire blanchatre», The marginal achenes are shorter (1,5 mm long) and more curved than those of the centre, where they are 2 mm long, a difference also noted in &. Gayana. There are only 5 to 10 bristles, while, in all the congeners, they are much more numerous. Area of distribution: Endemic in Masatierra. 128. R. evenia Phil. — JoHow, Estud. 60. — Fig. 33 l—o, 34 b, 35 b. Masatierra: GERMAIN! REED! DOWNTON! MOSELEY], all without locality. — Not uncommon in the eastern and central parts, along the higher ridges, at the limit of the forest among brushwood, or in the humid, D7cksonza-rich woods. — C. Chifladores, one large, apparently not epiphytic specimen at 350 m, many in the forest 500 m, all epiphytic; between Q. de la Piedra Agujereada and Q. Laura, c. 500 m, on Lecksonia, abundant; C. Centinela, 560 m, on Dzck- sonia, abundant; the depression between Co Damajuana and Co Yunque, 580 m, on Thyrsopteris; V. Colonial, C. Central, 570 m, on Dzcksonia; Portezuelo de Villagra, among shrubs c. 600 m, several trees, all growing on tree-ferns (fl. J 3/2 16, no. 25); C. Salsipuedes, in Dzcksonza-grove, 670 m, seedlings (no. 74) and adult plants, all epiphytic (fl. 17/1 17, no. 82 9, 83 4); dense forest on a mountan spur west of Co Yunque, c. 500 m, epiphytic; Q. Villagra, c. 500 m, on Dicksonia. Curiously enough it is nowhere stated that 7. evenza is an epiphyte. It grows on the trunks of tree-ferns; only in one single case a large and old speci- men was found growing on the ground, but I am inclined to believe that also this had germinated on the base of a fern tree, sending, as it often does, roots into’ the soil, and that the original substratum had disappeared. The seedling has fleshy, densly pilose, coarsely serrate leaves. Even the old, perfectly glabrous leaves are much more fleshy than in other species, a circumstance perhaps connected with the epiphytic life; they are as much as 1 mm thick (not counting the midrib), and the veins are hardly visible. 13 —20100. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 194 CARL SKOTTSBERG Fig. 33. Flowers of Rodinsonia. a—c KR. macrocephala Q: a ray, b disc floret, ¢ achene. d—g RX. thurifera: d ray, € disc floret of o, f ray, g disc floret of 9. h—k R. Gayana: h ray and i disc floret of o, j ray, k disc floret of 9. 1—o &#. eventa: 1 ray, m disc floret of of, n ray, o disc floret of 2. p—s A. Masafuerae. p ray, 4 disc floret of o’, r ray, s disc floret of 9. t—x R. gracilis: t ray, u disc floret of o, v ray, x disc Horet of 9. — All x 8. THE PHANEROGAMS OF THE JUAN FERNARDEZ ISLANDS 195 Male plant: Head 8 mm high, 10—12, rarely as much as 15 mm across. Involucre of about 13 scales, 6—7 KX 4—4,5 mm, campanulate, with 5 or 6 small calycular bracts. Ray florets 8, pure yellow, 8—1o mm long; ligule tridentate, 3,2—4 X 2,5 mm; style hardly exserted, bifid; pappus rays numerous (25—30) in all kinds of flowers. Disc florets greenish yellow, 6—7 mm long; style deeply bifid with truncate branches. Female plant: Head about 10 mm high, 7—8 across; involucre cylindrical, 7—8 X 5 mm; ray florets 8, yellow, 7—8 mm long; ligule 2,5 x 2 mm; style greenish, deeply bifid; disc florets 6 mm long, yellow with greenish, barely exserted style. Young achenes slightly costate with pilose costae. Area of distribution: Endemic in Masatierra. *129.. R. Masafuerae_ nov. spec. — Plate 17, fig 1; text fig. 33 p—s, Bava oe Cc, ok, 1. Arbuscula dioeca ad 2 m alta, trunco primario certe ad 7 cm crasso, pseudo- dichotome-subumbellatim ramosa, ramis junioribus viridibus, cicatricibus lineari- bus foliorum manifeste notatis. Folia ad apicem ramorum more generis conferta, cras- siuscula, subcoriacea, dilute viridia, glaberrima, lanceolata—anguste obovato-lanceolata, in apicem acutum sensim angustata, versus basin valde contracta, margine in parte dimidia superiore minute denticulata, nervo mediano utrinque conspicuo, basi latissimo, secun- dariis indistinctis, angulo acuto exeuntibus, plerumque 7—12 cm longa et I,;—2 cm lata, sed magnitudo summopere variabilis (folia ad 18,2 X 4,7 cm visa, in specimine sylvicola unico invento ad 22 « 3,5 cm). Inflorescentia terminalis valde ramosa corymbosa, bracteis sat parvis instructa; pedicelli capitulorum I—2 cm longi, superne pilosi. Capitula ¢ 7 mm alta et I0 mm crassa; involucrum campanulatum, 5 mm altum et 4 mm latum, addito calyculo a bracteis perpaucis angustis ad 2 mm longis formato. Squamae circ. 12, 5 mm longae et 1,5 mm latae, glabrae vel dorso pilis nonnullis inspersae, apice penicil- latae, crassae, margine scariosae, liberae sed manifeste imbricatae. Flores ligulati 8, flavi, 8 mm longi, ligula 3—3,5 * 2 mm, irregula- riter tridentata, in tubum abrupte angustata, stylo breviter bifido paulo exserto. Flores disci 6 mm longi, viridi-lutei, antheris flavis; stylus inclusus apice bifidus ramis truncatis penicillatis. Pappus multiradiatus. Capitula 2 magis numerosa certe ad 75, 8 mm alta et 7 mm lata; involucrum cylindrico-campanulatum, sub apice constrictum, basi truncatum et pilosiusculum, 6—7 mm altum basi 4—5 mm a Fig. 34. Leaves of a Robinsonta Masa- Juerae, b R. evenia. Nat. size. 196 CARL SKOTTSBERG latum, calyculo conspicuo e bracteis 5—6 (vel ultra) ad 4 mm longis formato. Squamae ad 14, ut in gd, sed magis adpressae et conspicue longiores. Flores ligulati 8, flavi, 8 mm longi, ligula 2,2—2,5 mm longa et 1,;—2 mm lata, sub- tridentata; stylus exsertus ramis patentibus. Flores disci viridescentes, 6 mm longi, staminodiis minutis, stylo exserto ut in praecedente. Achaenia fusiformia, fusca, 9—10-costata, c. 2,2 mm longa; pappus setis circ. 25 basi annulum minutum formantibus. Masafuera: On the high ridges, generally growing in the fernbeds, often much damaged by the goats, which had killed a great many specimens. — Edge of western precipice, above Buque Varado, c. 1250 m; near the Corre- spondencia camp, 1130 m (fl.-past fl. '®/2 17, no. 434 &, no. 1233 0); Las Torres, 1370 m, rare; C. Atravesado, 1350 m, one small tree; C: del Barril) © 7o@mms; Los Inocentes, 740—950 m, among Dicksonza, also seen growing on the trunk of this; slopes of O. Varadero, c.650 m. One aberrant sterile plant was observed in ©. de las Chozas in dense forest, with long, thin leaves (no. 374); one seedling was found in the Casas canyon. Of special interest as the first Rodzmsonza reported from Masafuera. It is very near 7. evenza, but as a rule not epiphytic and differing in many respects. The leaves (fig. 34) have a broader midrib and are more narrowed towards the base, the base is less incrassate and the scars left linear, not triangular or semilunar as in evenza. The female head is a trifle smaller and has a more conspicuous calyculus. The details of the flowers show some differences, as seen from my illustrations. Area of distribution: Endemic in Masafuera. 130. R. gracilis Decne. — JOHOw, Estud. 61. — Fig. 33 t—x, 35 h. Masatierra: In the shrubberies of the ridges, not seen at a lower altitude than 530 m, scarce. Between Q. de la Piedra Agujereada and Q. Laura, robust specimens; C. Centinela (also observed by JOHOW), 530 m, forming small thickets and exceptionally well developed (fr. 78/3 17, no. 577); crest of the island above Pangal, c. 800 m, very scarce; V. Colonial, C. Central, c. 570 m, solitary; Portezuelo de Villagra, one ¢ plant near the SELKIRK memorial (buds 3hie, beg. fl. '/12 16, no. 24), and a few @ plants, one of which grew on Blech- num cycadifolium (fl. **/12 16, no. 195), C. Salsipuedes, 730—750 m, solitary (bes, f1es/1e/16,9n0. 78 Gasi/ 2). JOoHOW writes |. c. »ligulas numerosas i mucho mas largas que en las demas especies». HEMSLEY’s figure of a male plant shows 8—1Io ligules (pl. 55). My observations, made on abundant living material, are not in accordance with JOHOW’s statements. Inflorescence comparatively small, only containing 20—25 heads. Caly- culus of few, minute bracts. Involucre 4 mm high and 4 mm across, little exceeded by the florets, of about 12 scales, firmly pressed together in 2, much less so in g. The male head is 7—8 mm across, the ray florets generally 8, bright yellow, 5 mm long; ligules 2,5 x 2 mm, 2—3-dentate. Style slightly bifid with erect branches. Disc florets yellow, 4 mm long, style deeply bifid with truncate branches, hardly reaching the full length of the stamens. THE PHANEROGAMS OF ‘THE JUAN FERNANDEZ ISLANDS 1Q7 The female head is 4 mm across; ray florets 8, yellow, 4 mm long with small ligule (c. 1 X 0,9 mm), 2-—3-dentate. Disc florets cylindric, yellowish green, 3,2 mm long, with exserted bifid-truncate style and minute staminodes. The achenes are 1,2 mm long, costate, glabrous according to JoHow, but minutely scabrid. Area of distribution: Endemic in Masatierra. Fig 35. a—b tp of style in male disc florets, a of R. thurifera, b of R. eventa, X 50. © Q head of &. Masafuerae, X 5. d—i achenes: d—e FR. Gayana, ray (d) and disc (e); f—g PR. thisz- Jera, ray (f) and disc (g); h R. gracilis (disc); k—i R. Masafuerae, disc (k) and ray (i). d—i X 17. Rhetinodendron Meisn. 131. R. Berterii (Dcne) Hemsl. — JOHow, Estud. 55. — Fig. 36. Masatierra: in the high parts of the central section, generally in open, low forest, rarely observed below 500 m, scarce. — FE] Rabanal (JOHOW); the depression between Co Damajuana and Co Yunque, 540 m; V. Colonial, C. Central (also observed by JOHOW), on both sides of the crest, 535—570 m, many specimens (past fl. 2 15/; 17, no. 301); Portezuelo de Villagra, a few trees on a rock ledge, c. 600 m (beg. fl. 14/12 16, no. 134 d, 135 2); Q. Seca, in the forest, one seedling; C. Salsipuedes (also observed by JOHOw); highest part of Q. Salsipuedes, rare (fl. J 7/12 16, no. 173) ridge west of Co Yunque, c. 530 m; Q. Villagra, in the forest c. 500 m, solitary (also observed by JOHOW); south slope of Co Chumacera (fl. 2 °/1 17, no. 253). According to JoHow, Rhetinodendron generally is an epiphyte, and later 198 CARL SKOTTSBERG authors have repeated this statement, adding their own reflections. I ‘shall return to this question when dealing with my ecological observations: at present it suffices to tell that, in spite of the very special attention paid to this plant, of which I have seen much more than earlier visitors, I have been unable to discover a single case where an adult specimen was epiphytic. Once, two seedlings were observed on 7hyrsopterts, but seedlings of Robinsonias as well as of Dendroseris micrantha are often found growing on the trunks of ar- boreous ferns. The structure of the head in LRhetinodendron is interesting and will be described with full details later; this time I only refer to my figures. The pap- b Fig. 36. MRhetinodendron Berterit; a 9 head, b & head, ¢ o@ flower, d achene — All X 8. pus is peculiar, for it forms four different portions, two larger lateral and two smaller, dorsal and ventral; the achene is flattened in tangential direction. Area of distribution: Endemic in Masatierra; monotypic. Centaurodendron Johow. 132. C. dracaenoides Joh., Estud. 63; SKOTTSBERG, Stud. 5, Taf. 5, Fig. 2—3. Masatierra: only on the high ridges encircling Cumberland Bay, very rare. Forest on the very steep slope of Co Damajuana, one specimen seen; Portezuelo de Villagra, three specimens near the path just below the pass, c. 550 m, three more seen on the slope of Co Piramide close to the pass, 580 m (no. 1234); northwest face of the same mountain, on rock ledges, several trees. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 199 The first who collected Cextaurodendron was not JOHOW, but BERTERO; still, B. did not discover it. Under the name of Dendroserits micrantha, BER- TERO no. 1602! (Herb. Kew) there is one sheet with pieces of an old inflore- scence of the latter together with a leaf of Centaurodendron. JOHOW discovered this in July, 1892, and found it again in June, 1895; to judge from a specimen in his herbarium it was also gathered by SOHRENS, probably in April, 1895, on the same occasion when S. found Rodznsonia thurt- fera, for it lies together with this under the same name. I rediscovered it in August, 1908, and found the old inflorescence, just as JoHow did, but there was no trace of new ones. From this I concluded that it flowers late in the summer or even in the autumn, and I expected to get flowers this time, but was cruelly deceived. For all I could do was to state that in certain years not a single specimen produces flowers, and that the year 1916—17 was one of them. Area of distribution: Endemic in Masatierra; monotypic. Dendroseris D. Don. The type of the genus is D. macrophylla D. Don, Phil. Mag. XI (1832) 388, collected in Masafuera by CUMING (Herb. Kew!). One year later DECAISNE, in Arch. de Bot. I (1833) 513, described the genus Rea Bert. ms., with the following species: R. macrantha Bert. et Dene (Masatierra), erterzaza Decne (Masatierra), fzznata Bert. et Decne (Masatierra), 2evzzfolzca Dene (Masatierra), micrantha Bert. et Decne (Masatierra), margimata Bert. et Dene (Masatierra: Portezuelo), mo//zs Bert. et Dcne (Masatierra: Portezuelo). In Comp. Bot. Mag. I.32 HOOKER and ARNOTT reduced Rea to Dendroserts, listing the following species: D. macrophylla (with R. macrantha as syn.), Bertertana, pinnata, nerit- folia, micrantha, marginata and mollis. JoHow, Estud., reduced these seven species to four, declaring D. ber- tertanma identical with pzzuata and marginata with macrophylla and rejecting D. mollis altogether. Of this, only the leaves, described as »subincano-velu- tina» — glabrous in all the others — are known. There is no specimen in Kew, nor in Paris, and I have not been able to discover the type. JOHOW guessed that a stout specimen of Guaphalium citrinum (= cheiranthifolium) or some other similar plant was mistaken for Dendroseris. This is hardly probable. In the locality given, Portezuelo, nothing like it is found, unless a young Aodzn- sonta or Rhetinodendron was mistaken for a Dendroseris. With regard to BER- TERO's ability as an observer, this is difficult to believe. The treatment of Dendroseris in JOHOW’s flora is not quite satisfactory; still, he is not to blame. He had no opportunity to see BERTERO’s types, his own material was scarce; further, when HOOKER and ARNOTY reduced Kea macrantha to D. macrophylla they indicated the wrong course later followed by Gay and Jouow. HEMSLEY pointed out the validity of D. margznata, but JoHow brought this to macrophylla. He was equally critical against his own discoveries, for the peculiar mzcrantha from the coast rocks was only with hesita- tion classified even as a variety, v. pruznata. In the case of D. Berteriana, 200 CARL SKOTTSBERG JoHow based his opinion on BERTERO himself, who was inclined to regard it as a noticeable form of fzuvata. While I regard fruznata as a species, I follow JoHow with respect to Sertertana. Finally, JoHow added one new species, D. gigantea, which has remained very little known until now. Returning to D. macrophylla, it is evident that Rea macrantha is different from this. Both have incised leaves, but their shape is not the same and there are many differences in the flower and fruit. 2. macrophylla is endemic in Masafuera, macrantha in Masatierra. Rea marginata is certainly distinct from both, not only in the leaves, which are entire, but also in flower and fruit, described here for the first time. Finally, var. marginata of JoHOW, at home on the coast rocks, is distinct from the true margznata and will receive the new name JD. UMitoralis. Thus, D. macrophylla in JOHOW’s sense includes four species, one from Masafuera, two confined to Masatierra and one to a rock off Masatierra and to Santa Clara. These four species differ from the rest in many respects. The stem is branched, hollow with numerous diaphragms, the receptacle is also hollow, its surface plane, with numerous fibrils round the points of insertion of the achenes. The head is very large, the flowers bright orange, the achenes distinctly winged, the pappus formed by short, irregular, + rufous bristles. I am pretty sure that the »splitting-up school» would reserve the name Dendroseris for these four species. My reasons for not doing so were explained under Xodznsonza. The remaining species have creamy white flowers and unvinged, obovoid, slightly compressed achenes; here the accordance between them ends, and they belong to two rather different types. In D. pemnata and D. regia (nov. spec.) the stem is simple, but otherwise as in D. macrophylla; the receptacle (unknown in D. regia) is hollow, but concave and only slightly fibrillose. The remaining species (D. micrantha, pruinata, neritfolia and gigantea) have a branched and solid stem, and a solid, concave and naked receptacle. DD. pimnata and regia are intermediate between the macrophylla-type and the mzcrantha-type. Either, we should retain Dendroser?s in the old sense, or three genera must be recognized. Further, D. ezgantea differs from all its congeners in the shape of the corolla. In view of these circumstances I propose to subdivide the genus in the following manner. I. Stem branched, hollow. Head 3—5 cm wide; receptacle fibrillose; flowers orange yellow; achenes winged. -__.-.... Subgen. I. Eudendroseris Skottsb. A. Leaves chartaceous, grossly dentate-lobate. 1. Leaves oblong-ovate, + regularly lobate. Anthers dark. Achenes mostly broader than long. Cotyledons entire. D. macrantha (Bert. et Dene) Skottsb. 2. Leaves often as broad as long, irregularly lobate-serrate. Anthers yellow. Achenes twice as long as broad. Cotyledons trilobate. D. macrophylla D. Don. B. Leaves coriaceous, entire. 1. Anthers dark. Achenes thick. D. marginata (Bert. et Dene) Hook. et Arn. oAwAmthers;: yellow, Atchenes, very flates. «22.4 D. litoralis Skottsb. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 201 II. Head smaller. Flowers whitish. Achenes not winged. A. Stem simple, hollow. Receptacle fibrillose. Leaves pinnate. Subgen. I]. Phoenicoseris Skottsb. 1. Pinnae deeply bifurcate with linear segments. D. pinnata (Bert. et Dene) Hook. et Arn. 2. Pinnae + ovate, dentate. a. Petiole winged, but not auriculate. D. pinnata var. insignis (Bert.). bil Petiolesdurieuldte:su: ci tees! tk. oteinimmen Deringer Skottsh B. Stem branched, solid. Receptacle naked. Leaves entire. Subgen. III. Rea (Bert. ex p.) Skottsb. 1. Leaf-margin minutely denticulate or entire. Ligule 5-dentate. Sect. 1. Aurea Skottsb. a. Leaves petiolate. Involucre funnel-shaped, c. 5 mm high. +. Leaves coriaceous, narrow lanceolate. D. neritfolia (Dene) Hook. et Arn. + +. Leaves chartaceous, ovate-lanceolate. D. micrantha (Bert. et Dene) Hook. et Arn. b. Leaves sessile. Involucre campanulate, c. 10 mm high. D. pruinata (Joh.) Skottsb. 2. Leaves densely dentate-serrate, very thin. Ligules deeply 5-cleft. Sect. 2. Schzzoglossum Skottsb. D. gigantea Joh. BENTHAM and Hooker FIL., Gen. plant. I. 219, describe the achene of the Dendroseriidae in the following terms: »Achaenia a dorso compressa sub- trigona v. 2—3-alata». This description fits the genus /’zchza, a genus not very near Dendroseris. The achene of Zhamuoseris is unknown. The marginal achenes of Dendroseris are more or less compressed from the back, triangular in section, with two lateral and sometimes also with a ventral wing in £u- dendroseris. But all the other achaenia in the head are compressed from the sides, and their wings in Ludendroseris are dorsal and ventral only. This dimorphism clearly stands in relation with the different position in the head. It is illustrated below, fig. 37. 133. D. macrophylla D. Don. — JoHow, Estud. 70 p. p. min. — Fig. siaiasoim: Masafuera: CuMING! DowNTON! — On the walls of the canyons, also higher up on the ridges and on the top of the lofty coast cliffs in some places, scattered. Ravines above Q. Sanchez, solitary specimens; Q. de las Casas, rock ledges (A. ?%/2 17, no. 362; also observed by JOHOW); Q. de las Vacas (also JoHow); Q. Inocentes, some large trees on the cliffs c. 500 m; Q. Angosta, in the narrow gorge; Rodado del Sdndalo, on the top of the cliff; Q. de la Loberia, some fine trees, c. 250 m. There is a good illustration of this species in CURTIS, Bot. Mag. t. 6353. The specimens growing in the fissures on the canyon walls are small and seldom more than 2 or 3 m high; in more suitable places there are much larger ones, at least 5 or 6 m high and with a trunk 15 or perhaps 20 cm thick, so 202 CARL SKOTTSBERG that D. macrophylla surpasses all the other species of Ludendroserzs in size and beauty. It differs from these in several respects. The leaves are thinner, papyraceous when dry, pure dark green. Their shape is variable. The large rosulate leaves measure about 30—35 x 26—28 cm; they are broad ovate with cordate base, + acuminate, grossly and irregularly dentate with triangular, cal- lose teeth. The petiole attains a length of 20 cm; it is canaliculate, winged, but comparatively slender. Other rosette leaves are orbicular in outline or even broader than long (18 cm long and 23 cm broad in one of my specimens), cordate, deeply inciso-lobate, at least in their lower half, with linear-triangular segments. The inflorescence is composed of few but very large heads on long pedicels; the lower supporting leaves are petiolate, subpinnate or sinuate, the uppermost sessile, entire, auriculate. The corolla is c. 45 mm long, the ligule 3,5 mm across. The anthers are yellow. The style is yellow, with 3 mm long branches. The achenes (only few mature ones seen) are 5—6 mm long, brown, striate, the base narrow. In the marginal ones only the ventral wing is distinct, the lateral wings being poorly developed. Embryo with very short radicle and trilobate cotyledons. Area of distribution: Endemic in Masafuera. 134. D. macrantha (Bert. et Decne) Skottsb. — JoHOow, Estud. 70 sub D. macrophylla. — Fig. 37 b—d, n. Masatierra: BERTERO no. 1600! — Portezuelo de Villagra, on the south- west side of the pass (also observed by JOHOW), c. 580 m, two small specimens seen (beg. fl. */1 17, no. 330). BERTERO describes this as a tree 8 to 15 feet high and found in the shady woods of the high mountains. His specimens probably came from Porte- zuelo, as they bear no. 1600, for no. 1601, D. marginata, is collected in this locality. To judge from BERTERO’s words, it seems he saw DP. macrantha in more than one locality. In spite of a diligent search, we only found the two dwarf trees mentioned above, growing almost concealed among dense shrubs of Pernettya, etc. D. macrantha differs from macrophylla in many respects. The leaves are oblong, about twice as long as broad (large blades 25— 30 cm long by 13—14 wide), firmer than in the latter, shorter petiolate (6—7 cm), narrow cordate at base but hardly decurrent, coarsely and regularly dentate-lobate; upper leaves decurrent on the petiole, subentire or entire, uppermost sessile, auriculate. Heads more numerous than in macrophylla. Flowers numerous, corollas as in D. macrophylla, 35—40 mm long; anthers lilac brown; connective tip yellow with numerous low, dark papillae, not found in other species. Style darker than in macrophylla. Achenes quite different, broader than long (3,5—4 by 4,5—6 mm), obscurely wrinkled, with narrow but distinct wings. Embryo 3—3,5 mm long; cotyledons entire, broader than long. Area of distribution: Endemic in Masatierra. 135. D. marginata (Bert. et Dcne) Hook. et Arn. — JoHOw, Estud. 70 sub D. macrophylla; var. marginata Joh. ex REICHE, Flora de Chile V.5. — Plate 18; Plate 20, fig. 11; text fig. 37 e—i, o. THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 203 Masatierra: Portezuelo de Villagra (BERTERO no. 1601!), on the north- east precipice of Co Pirdmide, rock ledge, c. 625 m (fl.-fr. 3/12 16, no. 3). Most likely the plants observed on the face of Co Damajuana, c. 530 m, and at the back of Q. de la Choza, c. 450 m, in both places quite out of reach, belonged to this species. JOHOW wanted to explain DY. marginata as a local form of macrophylla, due to external conditions. Both, however, grow in the same place and do not Fig. 37. Achenes (a—l) and embryos (m—p) of Dendroseris subgen. Eudendroseris. a D. ma- crophylla, two achenes from the side; b—d PD. macrantha: b marginal achene, ventral view, e—d central achenes, side view; e—i D. marginata: e—f marginal achenes, e ventral and f lateral view; g—i central achenes, g—h lateral, i ventral view; k—1 D. “toralis: k central achene, lateral view and 1 marginal achenes, dorsal view. m embryos of D. macrophylla, n of D. macrantha, 0 of D. marginata and p of D. litoralis. — All X 5. only differ in the leaves. BERTERO’s specimens were sterile, but another collector, not quoted until now, found the inflorescence (in fruit). A specimen in Herb. Kew labelled »Rea longifolia Phil., Reed oct. 1872», consists of a leaf of D. micrantha and a piece of an old inflorescence of D. marginata. The leaves are coriaceous, broad ovate with rounded or acuminate tip and cordate base. They are quite entire, but the edge shows scattered warts, visible on the leaf near the rock in plate 18. The petiole is broadly alate. The upper leaves are sessile as in the former. The (hitherto undescribed) 204 CARL SKOTTSBERG inflorescence is rather large, more condensed, with numerous heads; these are smaller than in the former and the involucre much more urceolate. The open head in full bloom measures 3,5; cm across; it only contains about 50 florets. The corolla attains a length of 40—42 mm; the whitish tube is soft papillose in the middle part, the deep orange coloured ligule is 3 —3,5 mm across. Anthers dark lilac brown with tails considerably longer than the gland of the filament. Style yellowish white below, upwards covered with brown papillae, also on the back and sides of the branches, while their upper surface is clothed with low, light yellow stigma papillae. Branches of the style 4,5 mm long. Achenes larger than in any other species, 7 or even 8 mm, brown, thick, with longitudinal furrows; marginal ones curved with distinct lateral wings and a ventral keel, the others with thick dorsal and ventral wings. Embryo very large, 5—6 mm long, with entire, orbicular cotyledons. Pappus turning dark rufous with age. Area of distribution: Endemic in Masatierra. *136. D. litoralis nov. spec. — Syn. D. macrophylla Joh. ex p. — Plate 19; plate 20, fig. 12; text fig. 37 k—l, p. Arbuscula I,5;—2 m alta habitu praecedentis sed truncus primarius crassior. Folia ad apicem ramorum pauca dense rosulata; lamina obscure viridis sub- elaucescens subtus pallidior, coriacea-carnosula, valde nervosa nervis crassis albidis, late ovata, obtusa, cordata, margine laevi + undulato, ad 25 cm longa et 18 cm lata; petiolus sectione triangularis crassus inferne saltem canaliculatus et marginatus, ad 12 cm longus. Inflorescentia terminalis late paniculata mul- ticeps, foliosa foliis sessilibus amplexicaulibus. Pedunculi 2—4 cm longi. Capi- tula magna, florida ad 5 cm diam. Involucrum campanato-urceolatum, squamais externis viridi-glaucis ovato-triangulatis acutis, interioribus linearibus subhyalinis penicillatis. Receptaculum fibrillis laceratis usque ad 1 cm longis munitum. Flores numerosi certe ad 100; corolla ad 35nim longa, tubo angusto albido basi excepta molliter papilloso, ligula crocea 3—3,; mm lata stylo aequilonga vel brevior. Autherae dilute flavae, caudae glandula paulo longiores. Stylus flave- scens viridi-papillosus, ramis ad 4 mm longis. Achaenia sordide straminea, illa florum marginalium lateribus alatis ventroque carinato, cetera tenuia lateraliter valde compressa, 5—6 mm longa, basi valde angusta, sat late alata, alis dense ac longitudinaliter striatis. Embryo 3,;—4 mm longus, cotyledonibus integris suborbiculatis radicula paulo longioribus. Pappus sordide stramineus. Morro Vinillo, a small rock off the south coast of Masatierra (P. ARRE- DONDO a. o.); cultivated specimens, raised from seeds gathered in Vinillo seen in some of the gardens in Masatierra (plate 19). Santa Clara: Crevices in the coast cliffs of the main island, and on Morro de los Alelies (fl-fr. ?°/: 17, no. 342). Also observed by JOHOW. Very near D. marginata, but distinct. The leaves are thicker, with smooth edge, the heads larger with twice as many flowers, the anthers and style yellow, the tails of the anthers shorter. The achenes are very different and the embryo considerably smaller. These achenes are mature, for I have raised numerous specimens from the portion collected by me in Santa Clara as well as from others of absolutely identical appearance gathered in the garden of ARREDONDO, who introduced the species from Morro Vinillo. My plants are all of the same THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 205 type and agree perfectly well with the herbarium specimens as to the leaves. They are 3—4 years old but have not yet flowered. Area of distribution: Endemic in Santa Clara and Morro Vinillo. 137. D. pinnata (Bert. et Dene) Hook. et Arn. — JoHoOw, Estud. 71. Masatierra: On the higher ridges, at the limit of the forest, among shrubs and small trees, not uncommon. — Between Q. Laura and Q. de la Piedra Agujereada, c. 600 m, large trees; El] Pangal, west branch, c. 400 m, and on the narrow crest above Pangal, c. 800 m; north face of Co Damajuana, ;00— 530 m; V. Colonial, C. Central (JoHOW), solitary in the forest down to 350 m; Portezuelo de Villagra (JOHOW), scattered on both sides, 5—600 m, one specimen (no. 1235) growing on Blechnum cycadifolmm:; C. Salsipuedes (JOHOW), 600 —750 m, scattered in the shrubberies (beg. fl. °/12 16, no. 79 — one single tree, all others sterile); QO. de los Helechos, on Dicksonia; between Q. Salsipuedes and Pto Ingles; Pto Ingles, central ridge c. 470 m. var. insignis Bert. ms., Johow. — Syn. Rea Bertertana Decne. — Masa- tierra: ©. de la Piedra Agujereada, in wet forest c. 600 m; in the gap between Damajuana and Yunque, c. 580 m, in the humid forest, no. 164 (BERTERO no. 1505! [type] labelled »El Yunque Majo 18302; the Kew sheet consists of two leaves and one [complete??| inflorescence, the latter very small). Regarding D. Bertertana (insignis), the material collected hitherto is very scarce. BERTERO alone has found the flowers, and very late in the season (May). The very few plants seen by us suggested a pzunata-form of extreme shade and moisture. A seedling, quite like that of typical pzmnata, was found with them. The primary leaves are simple but denticulate and reniform. Area of distribution: Endemic in Masatierra. #138. D. regia nov. spec. — Plate 17, fig. 2; text fig. 38. Phoenicoseris D. finnatae affinis. Truncus simplex rec- tus ad 180 cm altus et 13 cm diam. visus, cicatricibus foliorum manifeste notatus, apice rosulam giganteam foliorum circ. 30 cm altam et 1 m latam gerens. Folia plantae juvenilis maxima, ad 150 cm longa et 57 cm lata a me visa, plantae adultae minora, 30—65 cm longa. Lamina crassiuscula, 3—6-pinnata, ambitu late vel latissime ovata, discolor subtus glaucescens, 24— 36 cm longa et 16—31 cm lata, praecipue subtus secus nervos = ir- regulariter pulverulento-pubescens mox glabra; pinnae laterales sinuato-dentatae, versus basin irregulariter et profunde dentato- laciniatae, flabellatim nervosae; terminalis triangulato-semiorbi- cularis basi totam latitudinem folii fere occupans, plicata, + pro- funde divisa et margine irregulariter dentata, pedato-nervosa; pinnae haud raro valde approximatae sese tegentes; petiolus ad 20 cm longus, crasse carnosus, inferne auriculis membra- naceis pulchre reticulato venosis, versus basin angustatis, apice rotundatis et approximatis, 6—11 cm longis et 0,s—2,5 cm latis ornatus. Cetera ignota. LA SM ‘ Awan’, $izia\e | ae ave Spb s a) Yo 2 6, %4o a se SS ue) big pgm wt go) Ceonwinpiet a cee Soe SEES H CE ROENE Ma SS) Soe. Ghee 5) cece. etn). co Ue segs, REE sine oie alee gcc Beer eruitre me NSIS VENER a) ee ae ee eT CO eee Aa ole 718, SEC RGOHERMANUIE Le tian un fey Ss des ie Gs Soe nt et NR ee EERE 20) SrhoavaldaviensismGnsebs settee oS Sass we eg) es ee Se pS a Ed ad oe Guaphalimealdunateoides Remy \ 2. 0. + + 2 2 + + <-eyHA lesen 2G0sor ree ee BEFLEZICNUNE. OG ee che) Se sees oS) ge asst Si oS tls ese) Se ee ee ete es ae 234 Gnaphalium cheiranthifolium Lam. spicatum Lam. : spiciforme Sch. Bip. . . stachydifolium Lam. Greigia Berteroi Skottsb. Gunnera bracteata Steud. insularis Phil. znsilaris, SCHINDLER . Masafuerae Skottsb., peltata Phil. . . : pyramidalis Schindl. . . Halorrhagis asperrima Skottsb. Heleocharis maculosa (Vahl) R. Br. Heterocarpus fernandezianus Phil. CHECLAAUGICS wal ieee masafuerana Skottsb. masatierrana Skottsb. melanocephala, JOHOwW . vincentina Clarke . Hordeum murinum L, .... nodosum L, . Hypochoeris glabra L, Tsolepis fuscopurpurea Steud. . Juania australis (Mart.) Dr. Juncus acutus L. . . Koeleria micrathera (Desv.) Gris. capillaceus Lam. Chamiissonis Kth. . Dombeyanus Gay Jernandezianus Steud. imbricatus Lah, planifolius R. Br. . Lactoris fernandeziana Phil. . Lagenophora Harioti Franch. hirsuta, auctt. . . Lapsana communis L, ... Lardizabala biternata R. & P. Libertia formosa Grah. Ligusticum fernandezianum Phil. . grandiflora Phil... .. Lobelia anceps Thunb, rupincola Colla Supa Alun. 6-62 28 Lolium multiflorum L. .... Loranthus Berterot YH. & A. Luzula cfr. alopecurus ... — *insularis Skottsb. . .. Lythrum hyssopifolia L. .. Malva nicaeensis All... .. Margyracaena Skottsbergii Bitter panmilotdels 2 s-% Margyricarpus *digynus Bitter . CARL SKOTTSBERG THE PHANEROGAMS OF THE JUAN Marrubium vulgare L. : Matthiola incana (L.) R. Br. . Medicago denticulata Willd. . —. lupulina Lo. ss ee — maculata Willd. SAME GAC Atte it isle wy m8 Megalachne Berteroniana Steud. Melilotus indicus (L.) All... . — parviflorus Dest. . Melissa officinalis L. .. . Mentha aquatica L.. . . Se pilenivmivls Migs. . 6. + 5 ee Mreropsis mand DC. . 2... +. Mimulus parviflorus Lindl. var. : Modiola caroliniana (L.) Moench. .. . Monocosmia monandra (R. & P.) Pax Myrceugenia fernandeziana (H. & A.) Berg — Schulzei Johow . are ; Myrteola nummularia (Poir.) Berg. . Nasturtium aquaticum (L.) Wg. ..... Merenadepressa B.S. 5 6 ee es Pcewans cordifolia Phil. ...... Nowodworskya imberbis Phil... 2... @chaeaviaelegans Phil) ....-... @enothera mollissima L. . .... — propingua Spach. .... Oldenlandia thesiifolia K. Schum. . . Oreobolus obtusangulus Gaud. . Oryzopsts bicolor (Vahl) Speg. . Beaisicouaiculatar lee. te ees ep laxac Fie 8) Al. — micrantha Bert. . Pantathera fernandeziana Phil. Parietariadebilis Forst. . ... . PUSH RAGING ee os Paronychia chilensis DC. ‘ Paspalum distichum. Lb... 2 2. — fernandezianum Colla ......-. Peperomia.Berteroana-Miqu... . . «+ + « = —etietaandezsaiaMigqe ties. x SR MAAOAITTEUARDET Eo 1am a BO ( — nummulariaefolia Griseb....... — Skottsbergi:-C. DC... Bemethya rigida. WC. Bor. we) Petroselinum sativum (L.) Hoffm. ..... plidlats intermediavBoses:: ..G5.> 2: a Phrygilanthus Berteroi (H. & A.) Reiche . . — tetrandrus (R. & P.) Eichl. .. Miivealiscperuvianal Lic. A mak. es Se B. PRbESTOMS, RiSea Ps ve. ai mi sh an ee Piptochaetium bicolor (Vahl) Presl. . .. . -—® laewissinaunal. Blilet i sr sane yet eserves Rea annual ibste i 6 @ < ais FERNANDEZ —“— es WN = & ££ he hb Dh QQ NNN 236 Bieamipcatemsis sie pesmi: )o-ptcitie teeta Plantago fernandezia Bert. ... . — lanceolata L.. ==) MMAlOLe es. poets eee nee =~ )*Skottsberei Pilger - bo Podophorus bromoides Phil. Polygonum aviculare L. ...... — hydropiperoides Michx .... Polypogon chilensis (Kth) Pilger . ee soins Widener oe ao oo — imberbis (Phil.) Johow ... . Psychotria pyrvifoia H.& A. .... Ranunculus caprarum Skottsb. . == TENATCa HITS eles, acre prea valcke es Raphanusesativuish leer ert a: cs Je myelts fea Berrerana Wone™ so = see Rhaphithamnus /ongzflorus Miers — venustus (Phil.) Skottsb. .. Rhetinodendron Berterii (Decne) Hemsl. Rhodostachys elegans (Phil.) Mez . Robinsonia evenia Phil. .. . = Gayana Dene 2 o. = ABOLACIIS OCHS J8 ca arene 31 a — macrocephala Dene. . . — Masafuerae Skottsb. ..... = sthuGiteras Che =. (caus. se Roubieva multifida (L.) Mog. . Rubs oeoIdes Smt) aay) a eee Inumnmexqalc etose lagen aeay rst eee — conglomeratus Murr. — crispus L. Sewoy Seen Be See iui cine vas. ane eee ee PRU AN OKACTEOSA DG. Ne tip tas tee E= tchalepemsiswlan tl ssa eo. ce Sagina \chilensissNaud........ . - Salicornia peruviana Kth .... . Sanicula slibertaiGi. é& Schl. >. = Santalum fernandezianum F. Phil. Scirpus cernuus Vahl. —— a OdOSUSBROLtOn ss oc ei Selkirkia Berterii (Colla) Hemsl.. . . BeneciOpvilloanis: Wes coo a uss Setaria geniculata R.& S.. — imberbis R.&S...... Silencvaallicall 53 ais @ be Silybum Marianum Gaertn. ... . Sisymbrium officinale (L.) Scop... . Solanum argenteum Dun... ..... = demmandezianom Phil... . = > TE LUCALUI, JOHOW, 2 «ss <>» — miasatfueranum Bitter & Skottsb. — Robinsonianum Bitter . Sonchus fallax, JoHow =—— Oleraceus du, . . . 2 + CARL SKOTTSBERG 215 171 223 223 171 99 216 216 100 214 100 73 123 218 218 205 163 163 197 110 193. 192 196 191 195 192 217 130 216 216 216 216 220 220 217 121 oe) 117 105 105 163 22 214 214 217 225 218 223 165 165 167 165 227, 22 THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 2 mupiord temandertana (Phil) Skottsb. . . «'s Foss 2% ole hs oo RO MRL eyo ete tae aesataIta-(ET).)! SKORSDE x ie wea, Ge OA Ve, Be RN tr ee RES Rie mae AD eared ara .COMertMord Studs Wan inde eee a oe We ae eco eR BRORERE, . ~ 7201 Pte ncrumaeer (Eni) ROWEBs . xia s+. 2 4s woe ee bo we nD fips ose MESES ODOW nn) a bee de, ee ke oy ow SORIA, Sees. Barta CUSDIIStAG NVI, (6 ws oe ee Ble 84 we dae, 6 a ose RR PORNO 277 RemIOMa CUTS fe 4 eos daw aed 8 SE dw Wo en « ORE. BET | ~217 MEECMHOR, JOUGW 6 nies a we we, Oe Hose 2 es SOLU oun, ang RACRORTE QUE ihe (ah eel et wk Ww WA) ate ees nk, a ee SSE. MN ROC IR GE Pemeeeteanicematia «EL. a) 3 re 6 eae ck mee AN ae yen od OM Teed) BP RApE . 96 ICRMAPAPATANG: StOUG. Sir ta" sy, eke est pas Pj EM bare saan BOD Linuiey , =—98 RE ALESOER OT DUSCE, es ces. SKS le ea ye AOR? LOE) Bog... erg Pema dernandezianyni Dahist.” . 2 3s.s sso 3 aos 2p © a0 a 6 a eees or, eee RMSE SDANS CON sick oa oy. vol ce ob Sine bv SN Po eS be ay et RNOLD. TDI BRASS 6 Ola fe So ae ek sk Oe kB ee oe 2 OR. PIS 2 Tee Bemeesestadosa Gaetth. 6002 BROTHERUS, Cum 3 tabulis (21—23). Ex Insula Paschalis (Isla de Pascua, Easter Island) solum ut videtur duae species muscorum adhuc cognitae sunt, Campylopus sp. e cratere Rano Aroi et Papillaria pascuana Thér. n. sp. in sched., quarum cl. F. FUENTES in opusculo »Resena Botanica sobre la Isla de Pascua» (Instituto Central meteorologico y geofisico de Chile. Publicaciones. N° 4, Santiago de Chile 1913), pag. 144 mentionem fecit. In itinere 1917 a cl. C. SKOTTSBERG una cum domina INGA SKOTTSBERG 14 species lectae mihi communicatae fuerunt, quarum 7 novas (Papillaria illa addita) infra descripsi. Faas Dicranaceae. Campylopus Brid. 1. C. introflexus (Hedw.) Mitt. Hanga Ho Orno\|, in cryptis rupium planitiei (542). Area: In zona temperata late distributa. 2. C. (Trichophylli) turficola Broth. n. sp. — Tab. 21, Fig. 1 a—i. Dioicus; gracillimus, caespitosus, caespitibus densis, late extensis, dila- bentibus, fuscescenti-viridibus, opacis. Caulis erectus, 4—8 cm altus, haud tomen- tosus, dense foliosus, apice curvatulus et breviter attenuatus, simplex vel parce ramosus. Folia sicca arcte imbricata, humida erecto-patentia, canaliculato con- cava, e basi breviter oblonga lanceolata, mutica, singula hyalino-apiculata, c. 3 mm longa et c. 0,6 mm lata, marginibus superne subconniventibus, integris vel subintegris, comalia breviora et latiora; nervo basi tertiam partem follii 1 = La Pérouse Bay. 16— 2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 242 Vv. F. BROTHERUS occupante, dorso laevi, in sectione transversa cellulis ventralibus laxis, inanibus; cellulis laminalibus rhomboideis, basilaribus internis laxe rectangularibus, hya- linis, externis multo angustioribus, alaribus numerosis, subrotundato-hexagonis, teneris, fuscis, in ventrem dispositis, fuscis. Caetera ignota. Rano Kao, in turfosis lacus crateris copiosissime, 110 m. s. m. (520, 521). Species statura gracillima follisque muticis dignoscenda. 3. CC. (Trichophylli) hygrophilus Broth. n. sp. — Tab. 21, Fig. 2a—f. Dioicus; gracillimus, caespitosus, caespitibus densiusculis, dilabentibus, late extensis, laete viridibus, intus fuscescentibus, nitidiusculis. Caulis erectus, 1O—13 cm altus, parce fusco-tomentosus, dense foliosus, innovationibus proliferis, similibus, supremis I—4, caudatis, usque ad 2 cm longis. Folia sicca imbricata, humida erecto-patentia vel patentia, rarius falcatula, canaliculato-concava, e basi oblonga sensim lanceolato-subulata, mutica vel brevissime pilifera, ad 5 mm longa et ad 0,7 mm lata, marginibus superne subconniventibus, integris; nervo basi tertiam partem folii occupante, dorso laevi, in sectione transversa cellulis ventralibus ducibus magnitudine aequalibus; cellulis laminalibus anguste rhomboideis, basi- laribus internis laxe rectangularibus, hyalinis, externis multo angustioribus, alari- bus numerosis, subrotundato-hexagonis, teneris, in ventrem dispositis, fuscis. Caetera ignota. Rano Kao, in turfosis lacus crateris, in aqua copiose, 110 m. Ss m, (516—5 10). Species praecedenti affinis, sed foliorum forma et nervi structura digno- scenda. 4. C.(Trichophylli) dicranodontioides Broth. n. sp. — Tab. 22, Fig. 1a—h. Dioicus; gracilis, caespitosus, caespitibus densis, late extensis, lutescenti- viridibus, intus fuscescentibus, nitidis. Caulis erectus, 3-—5 cm altus, haud tomen- tosus, dense foliosus, simplex vel innovando ramosus. Folia plerumque + distincte falcata, canaliculato-concava, lanceolato-subulata, singula + longe pilifera, pilo erecto, subintegro vel serrulato, 4—5 mm longa, marginibus superne subconni- ventibus, integris vel superne minutissime serrulatis; nervo basi tertiam partem folii occupante, dorso laevi, in sectione transversa cellulis ventralibus ducibus magnitudine aequalibus; cellulis laminalibus rhomboideis, incrassatis, lumine angusto, basilaribus internis laxe rectangularibus, hyalinis, externis multo angusti- oribus, alaribus numerosis, subrotundato-hexagonis, teneris, fuscis, in ventrem dispositis. Setae solitariae vel binae ex eodem perichaetio, cygneae, c. 7 mm altae, stramineae. Theca regularis, breviter oblonga, sicca plicata, pallida, ore rubro. Operculum oblique rostratum. Calyptra fimbriata. Rano Kao, locis lapidosis in cratere, circ. 125 m. s. m. (545, 548—550), cum f:a perfalcata (544). — Rano Aroi, ad moles, circ. 400 m. s. m. (547). — Etiam ab Exp. »Albatros» 21 Dec. 1904 a cl. A. AGassiz lectus (66, fia per- falcata). Species a praecedentibus foliis + distincte falcatis oculo nudo jam digno- scenda, habitu Dzcranodontiz. MUSCI INSULAE-PASCHALIS 243 5. CC. (Trichophylli) saxicola Broth. n. sp. — Tab. 22, Fig. 2 a—h. Dioicus; gracilis, caespitosus, caespitibus densis, viridibus, intus fuscescenti- bus, opacis. Caulis erectus, vix ultra 3 cm altus, haud tomentosus, dense foliosus, simplex vel parce ramosus. Folia + distincte falcata, canaliculato-con- cava, e basi lanceolata sensim longe subulata, breviter vel longius stricte pilifera, 3—4 mm longa, marginibus superne subconniventibus, apice minutissime serru- latis; nervo basi tertiam partem folii occupante, dorso laevi, in sectione trans- versa cellulis ventralibus laxis, inanibus, dorsalibus substereideis; cellulis lami- nalibus rhomboideis, basilaribus internis laxe rectangularibus, hyalinis externis multo angustioribus, alaribus numerosis, .ovali-hexagonis, teneris, in ventrem dispositis, fuscis. Seta solitaria, c. 7 mm alta, sicca flexuoso-erecta, humida cygnea, lutea, laevis. Theca irregularis, oblonga, sicca plicatula. Caetera ignota. Rano Kao, locis lapidosis, circ. 125 m. s. m. (522, 526—529). — Rano Aroi, ad moles, circ. 400 m. s. m. (525). — Hanga Ho Orno, in fissuris rupium planitiei (523, 524). — Vaintu Rova, in fissuris humidiusculis, circ. 300 m. s. m. (530). Species a praecedente statura minore, nervi structura nec non theca irre- gulari dignoscenda. Habitu Dicrano longifolio Ehrh. similis. Fissidentaceae. Fissidens Hedw. 6. F. (Amblyothallia) pascuanus Broth. n. sp. — Tab. 22, Fig. 3 a—e. Dioicus; tenellus, gregarie crescens, viridis, aetate fuscoruber opacus. Caulis 2—5 mm altus, infima basi fusco-radiculosus, dense foliosus, simplex. Folia pauci-multijuga, erecto-patentia, stricta, infilma minuta, superiora raptim multo majora, lanceolato-linearia, breviter acuminata, acuta, ad I mm vel paulum ultra longa et ad 0,15 mm lata, elimbata, integerrima, lamina vera paulum ultra medium folii producta, lamina dorsali ad basin nervi enata ibidemque angustata; nervo lutescente, aetate rufescente, subcontinuo; cellulis subrotundis, minutissimis, minutissime papillosis. Caetera ignota. Hanga Ho Orno, ad moles (532). — Vaintu Rova, in caverna humida, elite. 300 m. S..m. (531). Pottiaceae. Weisia Hedw. 7. W. flavipes Hook. fil. et Wils. Rano Kao, ad rupes circ. 300 m. s. m. (534). Area: Australia orientalis, Tasmania, Nova Zelandia. 8. , .W../sp. Hanga Ho Orno, in pascuis (534 b). 244 V. F. BROTHERUS Orthotrichaceae. Macromitrium Brid. 9. M. sp. Rano Kao, in lapidosis in cratere, circ. 125 m. s. m. (536). Bryaceae. Bryum Dill. 10. B. argenteum L. var. lanatum (Palis.) Bryol. eur. Hanga Ho Orno, in pascuis frequens (507, f:a gracilis). Area: In terris calidioribus late distributa. Bartramiaceae. Philonotis Brid. 11. Ph. laxissima (C. Miill.) Bryol. jav. Rano Kao, circ. 300 m. s. m. (552). Area: India orientalis, Madagascar. Meteoriaceae. Papillaria C. Mill. 12. P. pascuana Thér. ap. FUENTES lI. c. sine descriptione. — Tab. 23, Fig. 1 a—f. Dioica; gracilescens, caespitosa, caespitibus plerumque densis, rigidis, su- perne laete viridibus, intus lutescenti-fuscescentibus. Caulis primarius elongatus, repens, dense ramosus. Caules secundarii procumbentes vel penduli, flexuosi, ad 16 cm longi, dense foliosi, laxiuscule pinnatim ramosi, ramis patulis, vix ultra 2 cm longis, simplicibus, attenuatis vel obtusis. Folia decurrentia, sicca arcte imbricata, humida erecto-patentia, e basi cordata subsensim in laminam elongatam, lanceolato-ligulatam, acumine longiusculo, subulato instructam atte- nuata, c. I,g mm longa et c. 0,75 mm lata, marginibus integerrimis, basi + undu- latis et minutissime denticulatis; nervo lutescente, infra apicem folii evanido; cellulis rhomboideis, incrassatis, lumine angusto, papillosis, subobscuris, mar- ginalibus laevibus, pellucidis, basilaribus internis elongatis, angustis, laevibus, pellucidis, infimis aureis. Caetera ignota. Rano Kao, locis lapidosis in cratere, circ. 125 m. s. m. {554—557). — Rano Aroi, in cratere, c. 400 m. s. m. (558). — Vaintu Rova, in crypta humidiuscula, circ. 300 m. s. m. (553). Species P. cuspidiferae (Tayl.) Jaeg. valde affinis, sed foliorum forma dignoscenda. MUSCI INSULAE-PASCHALIS i) th On Fabroniaceae. Fabronia Raddi. 13. KF. macroblepharioides Broth. n. sp. — Tab. 23, Fig. 2a—g. Autoica; tenella, caespitosa, caespitibus plerumque densiusculis, late exten- sis, laete viridibus, sericeo-nitidis. Caulis elongatus, repens, + radiculosus, dense ramosus, ramis brevibus, densiuscule foliosis, simplicibus, obtusis. Folia sicca plerumque homomalla, humida erecto-patentia, ovato-lanceolata, pilo + elongato, integro terminata, marginibus erectis, fere e basi jam dense serratis, dentibus + distincte ciliiformibus, patulis; nervo tenuissimo, ad medium folii evanido; cellulis elongate rhomboideis vel elongate hexagonis, alaribus subquadratis. Seta 5 mm vel paulum ultra alta, tenuissima, straminea. Theca e collo distincto cylindrica, fusca. Peristomium normale. Operculum oblique rostratum. In cratere Rano Kao, locis lapidosis; 120 m. s. m. (559/563). Species /. macroblepharidi Schwaegr. affinis, sed foliis brevius ciliatis, laxius areolatis nec non theca majore dignoscenda. Rhacopilaceae. Rhacopilum Palis. 14. Rh. cuspidigerum Schwaegr. Rano Kao, locis lapidosis, c. 125 m. s. m. (566, 567). Area: Ins. Sandwicenses, Samoa, Norfolk, Nova Caledonia. 246 Vv. F. BROTHERUS Explicatio tabularum. Tab. 21. Fig. 1a—i Campylopus turficola n. sp.: a planta sterilis ('/1); 6--d folia (!7/1); e—g apices ('59/1); A pars basilaris (17/1); Z sectio transversa partis inf. (1/1). Fig. 2a—f Campylopus hygro- philus n. sp.: a planta sterilis (4/1); 4, ¢ folia (17/1); @ apex (18/1); e pars basilaris ('7/1); f sectio transversa partis inf. (17/1). Tabs 22: Fig. ra—h Campylopus dicranodontioides n. sp.: a planta fertilis (1/1); —d folia (1/1); e apex (7!%1); 7 pars basilaris ('7/1); g sectio transversa partis inf. (1/1); # theca ('?/:). — Fig. 2a—h Campylopus saxicola n. sp.: a planta fertilis (‘/1); b—d folia (12/1); e apex (4/1); f pars basilaris (17/1); g sectio transversa partis inf. (17/1); & theca (!7/1). — Fig. 3 a—e Fissidens pascuanus n. sp.: @ planta sterilis ("/:); 6 eadem (!7/1); c, d folia (99/1); e pars apicalis (47/1). Tab. 23. Fig. 1a—f Pafillaria pascuana Thér. nomen: a planta sterilis (4/1); 6—d folia (7/1); é apex (17/1); f pars basilaris (17/1). — Fig. 2a—g Fabronia macroblepharoides n. sp.: a planta fertilis (4/1); 6 eadem (27/1); ¢, @ folia (°/1); e pars apicalis (47/1); f, g thecae (29/1). = S _ A Z a! == OTe ap ° a GN I~ IO “) : + ~ —— == = > : = = Le ish Kor. Jf. || qj HAZE 6o. 2 VU y 2 z: J 5 ce CGA LOS V4 ; Z OBS Z a —Z) 2n OMS Og 2 oC, ZZ a 2 and Easter R de ce ae a ; “WOLD: = — Nat. Hist. Fuan Fernan > SED | SS 5 | oes ae ieee ee SS en, —aane LODDeGapare SS y zdez and Easter Isl. Vol. Il. Nat. Hist. Fuan Fernar > ule mit) A vy: Mi ny vA ery PLATE 23. Nat. Hist. fuan Fernandez and Easter Isl. Vol. II. \ \} | i! \) \J| \\ /d\ ‘ ANN \ ASRS | WSSSSSSK \ Cane ER Wi SAR AASSNS val Re SS Mit WNW His ( J Zi, ZA CFA <2 = SS SS — —— he a Ned SN No J0( i iN \ \ = i i 2d)\ | i it | , Jj! Dulin lic iit | \ i yt i ¥ auNige Sie MN Thain y Mi : Hut flac Mf ie Se tT Hid -\* g 5 | | \ atl Pe a ee es SS rags a Thérese Ekblom del. ; F Sige aS me +!) yl © 9. Marine Algae from Easter Island. By F. BORGESEN. With 50 text figures, When Professor SKOTTSBERG asked me to work out his collection of alge from Easter Island I undertook with pleasure this task, as the isolated situa- tion of this island made its algal flora of a special interest. The collection consisted of about 20 glasses, a few stones with crustaceous algz and some dried specimens, mostly the same as contained in the glasses. All the material brought home was collected at low tide between high and low water mark and in one locality, Hanga Piko, on the west coast of the island; sublitoral alge were only collected in tide-pools. As is understood from this, it is a small collection only, and as several of the glasses contain the same large species, it was to be expected that the number of species could not be great. At the examination of the glasses, however, fragments of other larger species were found upon and among the larger algz, and also some smaller epiphytic and endophytic species. Alltogether 68 species have been listed. Regarding several of these species the material is so scarce that there was hardly enough for microscopic preparation, while it was quite insufficient for a more detailed examination, a circumstance likely to have an influence upon the exactness of the determination. At all events, the present incomplete list of species shows that this very isolated island possesses a both rich and varied marine flora of which a more thorough investigation, including the sublitoral vegetation, is highly desirable. If I would try to give my opinion on the more characteristic features of the algal flora, judging from the present material, I might say that many of the species are small creeping ones which by means of strong hapteres or rhizoids are able to fasten themselves to the substratum; such species are f. i. Gelidium pusillum, Ptilothamnion Pluma, Laurencta claviformtis, Chondria repens, Gymnogongrus aequicrassus, Hypnea Esperi, Herposiphonia tenella, Dictyopteris repens etc. I have also come across proportionally many species of crustaceous alge. On the whole most of the species are small, the largest one found is Sargassum spec. Many species form low tufts and have a cartilaginous 248 F. BORGESEN thallus. These different types of the thallus are most certainly adjustments to the very strong surf that continually beats the shores of this extremely exposed island. As could be expected from the situation of the island, 27° s. Lat., the algal flora has a subtropical character with a few pronounced tropical forms f. i. Halimeda Opuntia, Valonia ventricosa and Chnoospora fastigiata. No Cau- lerpa was collected. It is of a special interest to note that some species hitherto known only from the West Indies or adjacent parts of the Atlantic are found at the shores of Easter Island, thus contributing to the remarkable correspondance between the floras of the Pacific and of the West Indies. Regarding this question compare my remarks in »The marine Algz of the Danish West Indies», vol. II, p. 491 etc. Finally I want to express my best thanks to the specialists who kindly assisted me with the determination of certain groups. I am much indebted to Mme. Dr. A. WEBER VAN BOSSE and Mme. PAUL LEMOINE for having determined respectively the Squamariacee and the Lithothamniacez found in the collec- tion, and to Dr. H. E. PETERSEN who took the trouble to name the small fragments found of Ceramzum. The drawings are for the most part made by mag. scient. O. ROSTRUP under my inspection. Chlorophyceae. Fam. Ulvaceae. Ulva L. U. Lactuca L. LINNE, Spec. Plant. II, 1753, p. 1163. The specimens are small, some few cm high only, forming low tufts, the thallus being folded, crisped and entangled as is often the case with this species on exposed localities. Area of distribution. Seems to occur in all seas. Fam. Chaetophoraceae. Endoderma Lagerh. E. viride (Reinke) Lagerh. — Fig. 1. LAGERHEIM, G., Bidrag till Sveriges algflora (Ofvers. af K. Vetensk. Akad. Férh., 1883, p. 75). I. Huper, Chaetophorées epiphyt. et endophyt. (Ann. Sc. Nat., 7. sér., Bot., tom. 16, 1892, p. 326). BOrcGesen, F., The Mar. Alg. of the D. W. L., vol. I, p. 10, Il, p. 416. — Entocladia viridis Reinke, in Bot. Zeit., 1879, p. 476, tab. 6, figs. 6—9. In the thick walls of Laurencia claviformis an Endoderma was present, most likely referable to &. virzde, which according to my concepcion of species MARINE ALGAE FROM EASTER ISLAND 249 (I. c.) is a rather variable plant. Quite as in the West Indian plants the present one formed a reticular tissue, the endophyte utilising the larger place left above the walls of the peripheral cells in the host plant. The cells are 8—15y. long and 7—8y, broad, of a rather irregular shape, often about oval, mostly as long as broad, but shorter ones are common too. aa Fig. 1. LEudoderma viride (Rke) Lagerh. from Laurencia claviformis. In one of the cells zoo- spores: areyseen, C2, 400) The cells contain much starch. The shape of the chromatophore could not be determined, probably because the material had been kept too long in formaline. Cells with zoospores were found now and then. Area of distribution: Atlantic coasts of Europe and N. America, West Indies, Easter Island. Fam. Cladophoraceae. Cladophora Kiitz. C. spec. Together with Gelidium pusillum a small Cladophora was found upon a shell. It had vigorous rhizoids deeply immersed in the substratum. The basal cells were 2—3 mm long and 150—175 y, thick, the upper cells shorter, but a few cells only were developed. The cell walls were thick and stratified. The largest plant found was 5 mm long. C. (Aegagropila) socialis Kitz. — Fig. 2. Kurzinc, Spec. Alg., p. 416. Tabulze Phycologice, vol. IV, tab. 71. ReEin- BOLD in A. WEBER VAN Bosse, Liste des Algues du Siboga, I Myxophyceae, Chloro- phyceae, Phaeophyceae, p. 82. To this species, originally described by KUTzinG from Tahiti, I refer a plant forming extensive, soft, Vaucheria-like tufts. On the upper side of these tufts the filaments are not much branched, often being quite simple for a considerable distance. The ramification is subdichotomous, the side branches 250 F. BORGESEN growing out at acute angles from the mother filaments and soon attaining the same strength as these. In the upper part of the thallus the cells are of much variable length, 5—10—-20 times as long as broad or even more. The diameter of the cells varies from 50—I100. being mostly 70—8o u.. The basal parts of the thallus have a different appearance. Here the cells are much shorter, the filaments (Fig. 2 a) are irregularly bent and curved, and much less regularly ramified; and furthermore we find here a great number of rhizoids. These rhizoids serve for the most part to attach the plant to the substratum but also to keep the whole thallus together as many of them fasten Fig. 2. Cladophora socialis Kitz. Parts of plants showing ramification and formation of riizoides: sa i1G2'2 "1, b,c tc. 1. themselves to other filaments. The rhizoids are sometimes long, sometimes short; in the long one cross walls are often present while the short ones have no cross walls. They generally end in a small coralliform disc. There is no cross wall separating the rhizoid from its mother cell. The wall of the filaments is thin in the young parts of the thallus, thick and stratified in the older parts. By its ramification and mode of growth as well as by the fact that the rhizoids are not cut off from their mother cells by cross walls this plant at the first glance reminds rather much of C/ladophoropsis, but it is of course easily distinguished by the presence of cross walls at the base of the branches. Furthermore it must be remembered that the formation of cross walls in the filaments of Aegagropila takes place in the same way as f. i. in Cladophora, MARINE ALGAE FROM EASTER ISLAND 251 while cell division in Cladophoropsis (as far I have been able to observe on preserved material of CZ. memébranacea) is realised in the peculiar way I have called segregative cell division. The species, originally found at Tahiti, has been collected at the Lucipara Islands by Mme. WEBER. Area of distribution: Pacific Ocean, Malay Archipelago. Fam. Valoniaceae. Valonia Ginn. V. ventricosa J. Ag. Specimens as large as dove’s eggs are found. They were fastened to other alge and were also themselves very much utilized by epiphytes. The specimens seem to agree very well with others collected by me in the West Indies (I. c., vol. I, p. 27). In both cases they were fastened to the substratum by means of rhizoids growing out from the small lentiform cells in the basal end of the large cell. In some of the specimens a good number of small roundish cells were found enclosed in the large mother cell, as described by MurRAY.' On the base of my West Indian material I arrived at the conclusion that these bodies owed their origin to abnormal conditions, but after I have seen the Pacific specimens I agree with MURRAY in considering them as quite normal organs of vegetative reproduction formed by a kind of free cell division — segregative as I have called this peculiar division modus found in so many related tropical forms. Another sample of algz contained some few specimens of more elongated shape about 2 cm long and 3/4,—1 cm broad. I dare not say, if these specimens more rightly ought to be referred to V. Forbes Harv. but had they come from the West Indies I would without doubt have considered them as forms of V. ventricosa. As is the case in this species they were covered by epiphytes, especially MWelobesiacee, while V. Forbesti, according to Mme. WEBER’, is less used as substratum. Area of distribution: West Indies, Malayan Archipelago, Tahiti, Easter Island. Rhipidiphylion Heydr. Rh. reticulatum (Asken.) Heydr. — Fig. 3, 4. Heypricu, F., Beitrige zur Kenntnis der Algenflora von Ost-Asien besonders der Insel Formosa, Molukken- und Liu-kiu-Inseln. Hedwigia, 33. Bd., 1894, p. 281. — Anadyomene reticulata Askenasy, Forschungsreise S. M. S. »Gazelle» IV. Theil, Botanik (Algen), p. 5. The thallus (Fig. 3) forms small, 2—4 mm broad, cristate and excentric bodies fixed by means of vigorous rhizoids to the host plant (Ga/axaura). ‘ Murray, G., On Halicystis and Valonia, in Murray, Phycological Memoirs, Part II, 1893, p. 5O—1. 2 WeseEr-VAN Bosse, A., Liste des algues du Siboga, I, Myxophyceae, Chlorophyceae, Phaeophyceae, p. 59. 252 Fig. 3. Young plant of RAcpidiphyllon reticulatum (Asken.) Heydr., c. *%/1. F. BORGESEN The rhizoids grow out from the lower- most ends of the cells in the basal part of the plant (Fig. 4b). They are irregularly bent and have thick walls. They grow down- wards along the wall of the main cells and contribute highly to the strengthening of these. Reaching the host plant they become divided in several thinner filaments which are termin- ated by irregular coralliform lobes. I have not seen any cross walls in these rhizoids. From the base of the main cells even high up in the thalius such rhizoids are formed. The growth of the plant is acropetal. The young cells are at first seen as small outgrowths upon the end of the mother-cell. These outgrowths are often rather large before they become separated by a wall (comp, Fig. 3). The ramification always takes place in the same plane. When the apex of a cell touches the neighbour branch it becomes fastened to it by a thick cellulose ring, formed in the same way as in Wzecrodictyon umbilicatum (Fig. 44). The large cells are about 200 », wide and as much as 600u. long. The cells contain a wall plasma with numerous chromatophores forming a dense network. Numerous pyrenoids are regularly distributed in the chromatophores. Fig. 4. Rhipidiphylion reticulatum (Asken.) Heydr. a summit of plant, c. ®/1; b base of small Specimen, (c,. oY: MARINE ALGAE FROM EASTER ISLAND 253 The flabelliform mode of ramification reminds very much of Anxadyomene, but in Rkzpzdiphylion the space between the large cells remains open no small cells being formed here. ASKENASY described the plant as an Anadyomene, but I think HEYDRICH was right in making it the type of a new genus. Rhipidiphyllon differs from A/rcrodictyon mainly by its acropetal, fanshaped ramification resulting from the fact that the new cells are produced from the upper end of the mothercell, all being directed upwards under acute angles and all lying in the same plane. It also differs in the less regular and less frequent junction of the cells. Aszpzdzphyl/on is also a much smaller plant; the largest specimens I have seen are about 5—6 mm in diameter, as stated by ASKENASY. It seems to me rather doubtful if the plant so beautifully figured by OKAMURA in his »lllustrations of the Marine Algz of Japan», Vol. 1, no. 6, pl. 30 really belongs to this genus. I am much more inclined to consider this a WMicrodictyon, The shape of the thallus, the mode of ramification, the shorter and less marked main cells, the more spreading branches, not distinctly gathered at the end of the main cells with much larger angles between them, and the lowermost being directed downwards, all suggests M/ccrodictyon. Area of distribution: West Australia, Formosa, Easter Island. Microdictyon Dene. M. umbilicatum (Vell.) Zanard. — Fig. 5. ZANARDINI, Iconographia, Phycologica Adriatica, vol. I, p. 79, tab. XIX, 1860. Some few small frag- ments were found growing on HHypnea. Regarding the _ limita- tion of species in A/zcrodictyon there still is some uncertainty, as pointed out by REINBOLD.' In the treatise quoted below he points out that the net- work arises in three different ways, namely by means of a simple anastomosis (» Ver- dickungsring», BITTER) as is the case in MW. umbilicatum, by pseudohapters, the attach- 1 REINBOLD, Tu.,in A. WEBER vaN Bosse, Liste Algues Siboga, Fig. 5. Microdictyon umbilicatum (Vell.) Zanard., part of i. ps 66: specimen, c. 7/1. 254 F, BORGESEN ment points being crenulated cell-ends as is the case in W/. pseudohapteron Gepp and finally by means of true tenacula as found in JZ. Montagnez Dene. Regarding the mode of net formation the present plant quite agreed with M. umbilicatum and therefore I do not hesitate to refer it to this species. It agrees very well with my West Indian specimens and specimens from Cadiz. Area of distribution: Mediterranean Sea, West Indies, Red Sea, Australia, Sandwich Islands ete. Fam. Codiaceae. Halimeda Lamx. H. Opuntia Lamx. forma Reinschii (Hauck) Barton. Barton, E. S., The genus Halimeda, p. 21. Okamura, K., Icones, vol. III, pl. 148. — Halimeda Reinschii Hauck, Ueber einige von I. M. Hildebrandt im Rothen Meere und Indischen Ocean gesammelte Algen. (Hedwigia, 1886, p. 167.) The specimens gathered seem to agree very well with the description of HAUCK and the figures of Mrs. GEPp (Miss BARTON). It is a rather small plant forming compact tufts ca. 5—6 cm high. The base consists of a short stem from which branches are issued in all directions. The lowermost joints are often broader than long and very thick, followed by a few ones nearly cylindrical and longer than broad, but soon the joints again become shorter, broader than long and trilobed. Higher up again the joint grow smaller, thinner, nearly roundish or ovate and sometimes Hightly trilobed, sometimes not. The midrib is very indistinct. The joints in the lower part are mostly about 2—3 mm long and 3—4 mm broad, higher up 5—6 mm long and 4—5 mm broad; the uppermost are about 4 mm high and 3 mm broad. Area of distribution: Indian Sea, Easter Island. Fam. Phyllosiphonaceae. Ostreobium Born. et Flah. O. Quekettii Born. et Flah. Bornet, E. et Cu. FLAHAULT, Sur quelques plantes vivant dans le teste calc. des Mollusques (Bull. Soc. bot. Fr., T. 36, 1889). This species was found in a small Sfzrordzs, the shell being quite green- coloured by the occurrence of the alga. Area of distribution: Europe, North America etc. most likely cosmo- politan. MARINE ALGAE FROM EASTER ISLAND 255 Phaeophyceae. Fam. Ectocarpaceae. Ectocarpus Lyngb. E. Chnoosporae nov. spec. — Fig. 6. Thallus pulvinulos parvos ca. 17/2 mm latos formans, e filis repentibus et filis erectis constructus. Fila repentia irregulariter ramosa et plus minus cohe- rentia e cellulis oblongis ca. 15—20y. latis et 24—28 y. longis composita. Fila ted Ty ay —s Ty ceo a = Ko = any sf! iar em CF rae ar. > = Fig. 6. Ectocarpus Chnoosporae nov. spec., parts of plants, c. 79/1. erecta ca. I mm alta, in parte basali e cellulis 8—10 p latis et 25—30 yp. longis in superiori parte e longioribus formata, simplicia aut in inferiori parte ramosa; rami aut longi, erecti, aut breves sporangia plurilocularia gerunt. Sporangia subfusiformia, ca. 100 », longa et 25 wv. lata. On the thallus of Chnoospora fastigiata J. Ag. var. pacifica J. Ag. a small brown alga was found which seems to me most naturally to come near the crustshaped Ectocarpfi quoted by OLTMANNS, Morphologie u. Biologie der Meeres- algen, 2. edition, p. 10, mentioned from the posthumous manuscript of KUCKUCK. As examples of the gradual reduction of the thallus of Actocarpus he men- tions two species, Lctocarpus faeroensis and spectosus, originally described by me under Myrionema in accordance with the at that time recent work of SAUVAGEAU on the Myrionemaceae. Later, these species were referred to various genera by various investigators; now KUCKUCK simply has included them in £ctocarpus. 256 F. BORGESEN The basal part of the plant consists of a monostromatic tissue composed of irregularly curved and ramified creeping filaments growing gradually more or less tightly together. The cells of the basal layer are about 15—20 yp, thick and 24--28» long. From these cells erect filaments arise. These are mostly simple having a zone of growth somewhat above their base. Here the cells are short with intercalary divisions. Upwards they gradually increase in length, the chromatophores become less developed and the ends of the filaments become hairlike, gradually dying off. These assimilating filaments measure about 8—10yp, at their base, the length of the cells being 25—30. In the zone of divisions the cells are about 16 » wide and mostly shorter than long. Keeping about the same width the upper cells attain length of about 65 ». The filaments are up to 1 mm long. Most of the assimilating filaments are simple but sometimes carry a single equally developed branch near their base; in the upper part they are always undivided. Near the base they also carry short filaments bearing terminal plurilocular sporangia. These filaments consist mostly of a single or of two cells. Now and then, too, a plurilocular sporangium is borne on a short filament arising directly from the basal filaments. In one specimen (Fig. 6 b) two small sessile gametangia were observed at some distance from the base and one sporangium on each of the filaments, somewhat recalling the genus Gowonema Kuck. und Skottsb.*. The plurilocular sporangia are ovate-spindleshaped, about 100 ». long and 25 broad. In the fresh divided cells, the chromatophores form several roundish discs, in the older cells oblong bodies or short staffs. This new “cfocarpus evidently comes near Compsonema fasciculatum and C. coniferum, described by SETCHELL and GARDNER in Phycological Contribu- tions II to IV (University of California Publications, vol. 7, 1922), but the present plant is much larger and also shows several differences in the shape of sporangia, assimilatory filaments etc. Area of distribution: Endemic. E. breviarticulatus J. Ag. — Fig. 7. I. AGarpu, Nya alger fran Mexico (Ofversigt af K. Vetensk.-Akad. Férhandl., 1547, Pp. 7). BORGESEN, F., Marine ‘Algae ‘of the D. W. I, vol. I, p. 173s ee. carpus hamatus Cr. in Mazé et ScHRAMM, Essai de classification des algues de la Guadeloupe, 2° Edit. 1870—77, p. 111; Vickers, A., Phycologia Barbadensis, part I1, pl. 29. The specimens found quite agree with the West Indian plant. The filaments are about 30 » thick. The length of the cells is rather variable. Near the summit of the filaments and generally in the periphery of the tufts the cells mostly are about as long as broad, while in the main fila- ments in the interior of the tufts the cells are often more than twice as long as broad. 1 SKOTTSBERG, C., Bot. Ergebnisse d. schwed. Exp. nach Patagonien und dem Feuerlande. VII. Marine Algae. 1. Phaeophyceae, p. 9. MARINE ALGAE FROM EASTER ISLAND 2hy The plurilocular sporangia (Fig. 7 b) are roundish-quadrangular or obovate. Commonly they are about 40 » long and 35 u. broad as is also the case in the West Indian form, the measure given by me (lI. c.) drawn from a rather large sporangium. So far one is able to judge from material kept so long in formaline, there appears to be several shorter or longer ribbon-like chromatophores, irregularly bent and with branches of different length in the young parts of the filaments, while in the older ones the chromatophores are more like small roundish or somewhat irregularly shaped discs. From Mrs. GEpr’s description and figures! of Ectocarpus spongiosus Dickie® I feel convinced that this species is the same as /. breviarticulatus, originally described by I. AGARDH in 1847 on specimens collected by LIEBMANN at St. Augustin, Mexico. Area of distribution: Pacific coast of Mexico, West Indies, Indian Ocean, Easter Island. Fig. 7. Ectocarpus breviarticulatus J. Ag. Fig. 8. ctocarpus sp. a filament with pluril- a sterile filaments, c. 45/1; b (c. 78/1) and ocular sporangium, c, 4/1; b the sporangium, ce (c. ?%/1) pluriocular sporangia. 6. E. spec. -— Fig. 8. The figure represents all the material found, and I would not have men- tioned the plant if it had not shown so much likeness to another »Lctocarpus spec.» mentioned by Mme. WEBER in the »Phaeophyceae of the Siboga», p. 131. 1 Barton, Eruet S., List of Marine Algae collected at the Maldive and Laccadive Islands by I. S. Gardiner (Journ. Linn. Soc.-Bot., vol. 35, p. 479)- 2 Dickiz, G., On the Algae of Mauritius (Journ. Linn. Soc., Bot., vol. 14, 1875, p- 191). 17 — 2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 258 F. BORGESEN The plant was attached to Vztophyllum spec. From the decumbent short filament, the long erect one arises. At the base the filament was 12, thick, somewhat higher up 19 », a litle above the sporangium 24. Along the intercalary growth-zone the width is about 23; from here the filament evenly tapers into a pseudo-hair like prolongation with cells about 15 » thick and So p. long. . The plurilocular sporangium is ovate fusiform, 80 p. long and 24 y. broad. It was placed upon the basal cell of a short branch. In the plant of Mme. WEBER the sporangia are sometimes sessile, sometimes placed upon a short pedicel. Fam. Ralfsiaceae. Mesospora Weber van Bosse. M. Van-Bosseae nov. spec.’ — Fig. 9. Thallus 2—3 cm et ultra altus e disco basali et filis erectis constructus. Discus basalis e filis repentibus coherentibus subdichotomo divisis compositus unde fila erecta oriuntur. Fila erecta in parte basali e cellulis latioribus quam altis et plus minus cohaerentibus stratum subparenchymaticum formantibus, in parte superiori libera, subclavata e cellulis cylindricis ca. 24 », longis et 8—II p. latis, ad apicem versus cellulis gradatim latioribus ad 16». latas composita. Sporangia plurilocularia in superiori parte filorum formata. Through the great kindness of Mme. WEBER I have been able to compare my plant with original material from the Siboga-Expedition. By means of this and the detailed description of Mme. WEBER I have convinced myself that the plant from Easter Island, even if it agrees with the Malayan one in many respects, nevertheless differs essentially in others, making it necessary to regard it as a new species. The plant grows on stones to which it adheres firmly with its whole lower surface. It forms dark brown or nearly black crusts as much as 2—3 cm in diameter or even more. The basal part of our plant is formed of several brown layers of old decayed tissue, often including blue green algae, in transverse section presenting a picture very like that found in fig. 43 of Mme. WEBER. I have succeeded in finding a small marginal portion, represented in fig. gb. It is composed of creeping, congenital filaments growing in length by means of a top-cell. By transverse walls these filaments are divided into more or less isodiametric cells. The filaments are about 10—15 y. thick, the top-cell often 204. or more. The filaments are radiant to all sides and by and by, as the disc expands and the width of the outermost ends of the filaments increases the cells show longitudinal walls, the filaments thus becoming dichotomously divided. At a very early stage erect filaments arise from every cell of the basal stratum. These erect filament grow vertically, standing quite close together. 1 | have the great pleasure to name this species in honour of Mme. Dr. A. WEBER VAN Bosse, the indefatigable explorer of the algal flora’ of the Malayan Archipelago. MARINE ALGAE FROM EASTER ISLAND 259 In the basal part the cells are broader than long and more or less coherent, forming a nearly parenchymatic tissue (Fig. 9 a). While this horizontal layer in Mme. WEBER’s plant is composed only of 1—4 layers of cells, ten or even more are not uncommon in my plant. The cells are considerably broader than long, ca. 18 ». broad and only 7 p. high. The filaments composed of these cells are now and then dichotomously divided; generally they are vertical, but now and then parts of the tissue occur with curved filaments. Above this tissue the filaments become thinner and at the same time mutually free in their whole length. In this respect our plant differs essentially = Aw f = . 0 4 O GAY Ui } a WG oe ny OS 0 BO RANA wd des -Y BAe WooK 1 idl Oo Bos woo ‘sie U5 Gh sm econ theless OF We ood Vibe le sia 19. 'a8 Beek) See SSse O| Ob C) i! = Oa) mee, O ici) 5 So PR erbagaaass) aS) \\ l . CY A Oy (9 a) 2 as) @ an CLO yo KX Me: Wi, DR YOK 7 my = NS wy, / JY : P. SPO Fig. 9. Mesosfora Van-Bosseae nov. spec. a transverse section of thallus, c. 7/1; b part of margin seen from above, c. 1/1; ¢ assimilating filaments and hairs, c. 1/1; d formation of plurilocular sporangia, c. 7°/1. from that of Mme. WEBER, who describes her plant in the following way: »Les filaments verticaux, claviformes qui la composent ont de 10—20 cellules, ils sont libres dans la partie basale et mediane, mais adherent fortement entr’eux au sommet, couvert d’une couche mucilagineuse assez €paisse». In our plant as explained above the basal parts of the filaments adhere rather firmly while upwards the filaments become quite free, only imbedded in mucilage. At the base of the free part of the filaments the cells are cylindrical and as much as 24 ». long and 8—11 yu. wide; higher up the cells grow shorter, mostly broader than long, the width of the filaments being about 16 uw. The filaments consist of about 20—30 cells. Thus, they are a good deal longer than in Mesospora Schmidti; furthermore the shape of the cells is 260 F. BORGESEN different, being nearly cylindrical with thin walls in our plant, while, in Mme. WEBER’s plant, they are barrelshaped and thickwalled. Now and then groups of hairs occur (Fig. 9 c); the hairs are about 10 p, thick at the base and composed of cells longer than broad; their growth-zone lies a little below the surface of the thallus and consists of quite short cells, densely filled with protoplasm and chromatophores; thence the cells quickly become long and colourless. At their base the hairs are covered by a single or some few sheaths. Unilocular sporangia were not present. On the other hand I have found some structures (Fig. 9 d) in the upper end of the filaments, similar to those observed by Mme. WEBER who regards them to be plurilocular sporangia. In my material they were not so regularly divided as in the Indian plant. They very much reminded of those found by me in Ralfsta expansa.’ In spite of a diligent search I have not succeeded in finding emptied plurilocular sporangia, but nevertheless I feel convinced that we have to do with such organs. By reason of the thick walls in MWesospora Schmidt a bursting of the wall takes place at the formation of the plurilocular sporangia as is easily seen in the drawings of Mme. WEBER, in our plant with its thin walls such a bursting is not visible. The cells contain an irregularly lobed and bent plate-like chromatophore in which a single or two refractive pyrenoid-like bodies are present. In one of the specimens, sometimes in almost every filament, a cell with homogeneous, yellow brown contents was found. Area of distribution: Endemic. Ralfsia Berk. R. expansa J. Ag. AGARDH, J., Spec. Alg., vol. 1, p. 63. BORGESEN, F., ‘Two crustaceous brown algae from the Danish West Indies (Nuova Notarisia, Serie 23, 1912, p. 123). The specimens found are not quite typically developed and I have hesitated to decide whether they are most naturally referable to R. verrucosa or to RX. expansa. It must be remembered that REINKE, in »Algenfloray p. 48, mentions that he has found a marked bilaterality in specimens of A. verrucosa from Cherbourg. Now, a transverse section of one of the specimens found mostly showed no bilaterality at all while in the other specimen contained in the collection this character vas better developed, being often in good accordance with the figure of Mme. WEBER. As compared with my West Indian specimens the Pacific ones had upon the whole a much thinner thallus. In one of the specimens unilocular sporangia were present; they were about 24 u. broad and 65 » long, in shape and size approaching those I have found in specimens from St. Thomas (I. c. fig. 2a). At the base of the spo- rangia a small cell was present just as in the West Indian plant; as I have ‘ Compare my figure 148c in »Marine Algae of the D. W. I», vol. I, p 1gt. MARINE ALGAE FROM EASTER ISLAND 261 pointed out (I. c.) this pedicel is not mentioned in the description of KUCKUCK ! nor figured in REINKE's Atlas, pl. 5—6; but it is present in HARVEY’s figure in »Phycologia Brit.», pl. 98; this character is, perhaps, not a reliable one. Area of distribution: West Indies, Malayan Archipelago, Easter Island. o Fam. Elachistaceae. Elachista(?) spec. — Fig. ro, On an old basal part of a Sargassum(?) a small brown alga was found which might perhaps be referred to the genus L/achista as it shows a good Fig. 10. “/achista(?) spec. Fragments with assimilatory filaments, mature and empty pluri- locular sporangia; in b the basal part of a hair. C, 7%. deal of likeness to Elachista rosariotdes and L. pusilla described by SKOTTS- BERG in Bot. Ergebn. d. schwed. Exp. nach Patagonien und dem Feuerland, 1907-9, VIII. Marine Algae, 1. Phaeophyceae, p. 21— 23. The plant forms small tufts. The basal part is endophytic, consisting of thin-walled, more or less barrel-shaped cells, about 15 » wide and of almost the same length. From the basal part arise hairs, long assimilating filaments, and shorter fertile branches. The assimilating filaments are rather thin at their base, about 4—7 u., with cells up to 30 long. Upwards the thickness increases to about 12—-18 »; at the same time the cells are proportionally shorter as they mostly keep the 1 Kuckuck, P., Bemerkungen zur marinen Algenvegetation von Helgoland, I, p. 244. 262 F. BORGESEN same length as found in the thinner part of the filament. In the young yet growing filaments the zone of cell division lies a little above their base, where shorter cells with rich contents are observed. The cells contain a number of roundish, oval, discformed chromatophores. The length of the vegetative fila- ments is about 300 w. - The fructiferous filaments are branched from their base and composed of very irregular thin-walled cells. These cells are more or less pear-shaped, thin below and broadly rounded above, where often several branches are given off. The cells are about 8—12y. broad and 20. long. From the uppermost of these cells the plurilocular sporangia arise. The sporangia are almost cylindrical, uniseriate, about 8 y, thick and up to 200, generally 150». long. The remains of empty sporangia are long preserved. When a sporangium is emptied a new one is formed at the side of the other one, and one cell may often bear 5—6 or more emptied sporangia besides a sporiferous one. Unilocular sporangia were not observed. Hairs occur, but were not common. As mentioned above the present plant seems to show some likeness to the Elachista-forms, described by SKOTTSBERG.’ Also the Easter Island plant is more or less endophytic and shows no difference between assimilating filaments and paraphyses, contrary to £. fucecola. Particularly it seems to agree with a plant from Station 17 b of SKOTTSBERG (fig. 10 c) but differs by the narrower sporangia and the more cylindrical vegetative cells. Further, the assimilating filaments in the plant from the Easter Island are shorter. It cannot be denied that the present plant shows a striking likeness to Myriectis moniliformis (Fosl.) Kylin, and highly favours the supposition of SKOTTSBERG that all these forms, even if they do not belong to one genus, must be regarded as nearly related to each other. Fam. Encoeliaceae. Colpomenia Derb. et Sol. C. sinuosa (Roth) Derb. et Sol. Derpks, A., and A. I. I. SoLttrr, Mémoire sur quelques points de la Physio- logie des Algues, p. 11 (here incorrectly called szwata); BORGrSEN, Mar. Algae of the D. W. L, vol. I, p. 176. — Ulva sinuosa Roth, Catalecta Botanica, III, p. 327, tab. XII. Some few large specimens were present in the collection. They were fertile with plurilocular sporangia in irregular groups scattered over the surface of the thallus, as in the West Indian specimens. Still, there are some points of difference. For instance, I have not been able to find any club-shaped para- physes so nicely developed in the West Indian form (comp. my fig. 138) and described by MITCHELL in MURRAY’s Phycological Memoirs, p. 53. Further, according to MITCHELL and my own observations the plurilocular sporangia are formed round the cryptostomata, while in the material from i Professor SkorrsBErG kindly sent me some slides of these species so I have been able to compare my plant with the authentic material. MARINE ALGAE FROM EASTER ISLAND 263 Ivaster Island the sori do not seem to be restricted to the vicinity of these but are formed everywhere on the surface of the thallus. Otherwise, the Pacific specimens agree very well with West Indian ones. Area of distribution: Widely spread in all warmer seas. Hydroclathrus Bory. H. cancellatus Bory. Bory, Dict. class. VII, p. 419. Harvey, Phycologia Australica, pl. 98; Nereis, p. 120, tab. IX A. ‘THURET et Bornet, Etudes phycologiques, 1878, p. 12. Mut- CHELL, M. in Murray, Phycol. Memoirs, p. 53. The specimens found agree very well with West Indian plants. The thallus consists of a medulla of large roundish-polygonal cells, and a small- celled cortex. Where two strings of the network come into contact the epider- mal cells grow out in a rhizoid-like fashion, uniting in this way different parts of the net work of the thallus. Compare my fig. 139, |. c. The material was sterile. Area of distribution: Seems to occur in all warm seas. Chnoospora J. Ag. C. fastigiata J. Ag. var. pacifica J. Ag. — igs T« 12) AGarou, J... Spec: Alg:; vol. 1, 1348, p. 171. — Chnoospora pacifica J. Ag., Nya alger fran Mexico (Ofvers. K. Vetensk.-Akad. Férh., 1847, p. 7). Some small specimens of a brown alga, richly provided with cryptostomata, from which long hairs protrude giving the plant a nearly moniliform appea- rance, seem to be referable to the genus Chvoospora as a small form of Chu. fastigiata var. pacifica. The largest specimen was about 6 cm high, the others 4 to 5. The plant is fastened to the substratum by means of a rather large irregularly shaped disc from which the erect shoots arise. The ramification is mostly dichotomous, some- times tri-polychotomous but rather irregular, with the whole branch system more or less fastigiate, even if a tendency to lie in the same plane is rather obvious. In the basal part the branches are about I mm a thick, rarely more, higher up thinner. The thallus is a Figo tr. °Chnbdspora' fastiviata little compressed, the transverse section being oval. J. Ag. Part of a plant, c. */:. 264 F. BORGESEN Scattered over the whole surface there are cryptostomata, generally rather shallow depressions, from the bottom of which hairs grow out to form rather long and dense, brushlike tufts. The blunt summit of the thallus has no hair- groups contrary to SKOTTSBERG’s Cladochroa. A transverse section of the thallus (Fig. 12 d) exhibits a cortical layer of oblong or short cylindrical cells, having nearly the same aspect on a longitudinal section; the parenchymatic tissue inside consists of rather thickwalled cells, oblong to subcylindrical in longitudinal, more spherical in transverse sections. Ae a Fig. 12. Chnoospora fastigiata J. Ag. a Transverse section of the thallus with plurilocular sporangia and hairs, c. #%/1; b longitudinal section of apex of a branch, c. */1; ¢ longitudinal section of thallus, c. '4°/1; d transverse section of thallus, c. 1/1. In the growth-point no marked top cell is present. As figure 12 b shows, the longitudinal growth takes place by means of a series of cells at the apex of the filaments, the lower part of these cells being gradually cut off. The figure of the growth point in our plant is very like those of Scy/o- thamnus australis and Cotlodesme bulligera (comp. OLTMANNS, Morphologie und Biologie der Algen, vol. I, 1922, p. 62) as drawn by KUCKUCK and’ found among his posthumous prepatory notes to the great work on the Phaeophyceae that was never completed, much to the regret of all algologists. According to OLTMANNS the notes of KUCKUCK regarding this matter are quite frag- mentary. MARINE ALGAE FROM EASTER ISLAND 26 unl As to the fructification this has been cleared up long ago by the late Mrs. Gepe.' The plurilocular sporangia (Fig. 12 a) are formed around the trichostomata; the surface cells grow out to cylindrical or somewhat clavate bodies which are divided by transverse walls into small cells in which the gametes are formed. Area of distribution: Chnoospora fastigiata has been found in several localities in the Pacific and on the coast of Venezuela in the Atlantic Ocean. Fam. Sphacelariaceae. Sphacelaria spec. Only some few sterile filaments, even without propagula, were observed; they cannot be determined. Fam. Dictyotaceae. Dictyopteris Lamx. D. repens (Okamura). — Fig. 13. Flaliseris repens Okamura, List of Marine Algae collected in Caroline and Marianne Islands, 1915 (Bot. Magaz., vol. 30, 1916, p. 8, pl. 1, figs. 7—18). This pretty little plant was found creeping upon the base of Ga/axaura paschalis and other algae. The specimens found seem to correspond very well with the description of OKa- MURA. The thallus consists of two layers of cells except in the midrib (Fig. 13 a); a sclerenchymatic rib along the margin of the thallus, as found in D. delicatula, is not present here. Besides this difference it further differs from the West Indian plant, to which it other- wise shows a great resem- Fig. 13. Déctyopteris repens (Okam.) Cross sections of blance, by the fact that the thallus, a through the midrib, b through margin. c. 7/1. hair-groups, well developed and regularly scattered in the West Indian plant, are mostly wanting in the Pacific one or, when present, poorly developed and placed without order. Rhizoids like those in the W. I. plant are frequent; they grow out from the midrib and from the margin of the thallus. The rhizoids have a long stalk composed of cells about 100 ». long and 33 yu. broad being somewhat narrowed at the cross walls. They end in a small coralliform disc. Unfortunately all the material was sterile. Area of distribution: Pacific Ocean 1 Barton, ETHEL S., On the fruit of Chnoospora fastigiata J: Ag. (Journ. Linn. Soc., Bot. vol. 33, 1897/3, p. 507, pl. 28). 266 F. BORGESEN Zonaria Drapern. Z. variegata (Lamx.) Mert. — Fig. 14. MERTENS in Martius, Icones plant. cryptog , p. 6, tab. 11, fig. 11. RICHARDs, H. M., Notes on Zonaria variegata Lamx. (Proc. of the Amer. Acad. of Arts and Sciences, 1890). SauvaGcEau, C., Observations sur quelques Dictyot. et sur un Aglaozonia nouveau (Bull. de la Station biol. d’Arcachon, 8 année, 1904—05). Borcesen, F., Marine Alg. D. W. L, vol. I, p. 197. — Dzictyota variegata Lawx., Essai, p. 57, tab. V, figs. 7—9. Gymnosorus variegatus (Lamx.) J. Ag., Analecta algol., cont. 1, ‘p. a,70s94- - - The specimens collected agree very well with those from the West Indies. They are 7—8 cm in diameter or even more; when dried they have a dark nearly black-brown colour and their consistency is tough and rather stiff. (QS) SOO = au) BOOSIE SU cs Ss Ge eae) Bac Gams) ES) PPOs eT eeae) i tt li i Gil at | ui Fig. 14. Zonaria variegata (Lamx.) Mert. a, b transverse sections, c. '7/1: a radial, b tangential; c upper surface of thallus, seen from above, c. 7/1. The thallus is about 150—200 y, thick. When examined from above the epidermal cells are seen to be arranged very regularly in rows (Fig: 14.c): The marginal rows consist of rather. long rectangular cells, but these cells become soon divided by longitudinal and transverse walls-into smaller cells, each row gradually getting composed. of from two to four rows of small cells. These cells have rounded corners and are filled with chromatophores. Eo MARINE ALGAE FROM EASTER ISLAND 207 The epidermal layer on the lower side of the thallus is also composed of long rows of cells, but here the cells mostly remain undivided or they become divided now and then by a longitudinal wall into two cells, and many of the cells are not divided at all. The rows of rectangular cells, longer than broad, are upon the whole much more distinct here than on the upper side. Fig. 14a shows a radial transverse section of the thallus. The large cells in the centre are about 60p high and 80, broad. On the upper side 3—4 or even more flat cells are cut of. The uppermost of these cells are divided into smaller cells, forming the epidermal layer. From the lower side of the large cells also 3—4 cells are cut off; the lowermost, especially the epidermal ones, are again divided into smaller cells though fewer than in the epidermal layer above. A tangential section of the thallus (Fig. 14 b) shows that the width of. the cells in each layer is only about half of that found in the radial section, and the vertical walls are seen to be more or less distinctly undulated. From the epidermal cells below numerous rhizoids grow out having the same moniliform appearance as those in the West Indian plant, and terminated by a coralliform disc. Scattered hairs occur in small groups on the upper side. The specimens are sterile. Ralfsia ceylanica Harv., described and figured by Mrs. GEpp', was later regarded by Mrs. and Mr. GEpp? as a creeping form of Zouaria variegata, After having seen the rather thick form from Easter Island, I feel inclined to adopt this view. In an earlier paper® I referred it to Aglaozonia canartensis Sauvag., which is, in several respects, very similar. Area of distribution: Seems to occur in all warmer seas. Z. crenata J. Ag. J. AGaRDH, Till Algernes Systematik (Lunds Univers. Arsskr., I. IX, 1872, p. 48). A single large specimen is contained in the collection. As expressed by J. AGARDH (I. c.) this species takes an intermediar position between Zonarza flava and Z. Turneriana, a view confirmed by the Easter Island specimen. The specimen was sterile. Area of distribution: West Australia, Malayan Archipelago etc. Fam. Fucaceae. Sargassum Ag. S. nov. spec., to be described by L. G. SJOSTEDT in a separate paper. Very common, forming extensive associations in the lower litoral region. Area of distribution: Endemic. _ ! Barron, E. S., List of Marine Algae collected at the Maldive and Laccadive Islands by 6 ey Gardiner (Journ. nn, Soc., Bot., vol. 35). << °2 Marine Algae and Marihe Phanerogams of the »Sealark» Expedition (Transactions ot the Linnean Soc. of London, 2. Ser., Zool , vol. XII, part 4, 1909). : .' 8 BORGESEN, F., Two eeustaceads eae algae from the Danish West Indies (Nuova Notarisia, Serie 23, 1912). 268 F. BORGESEN Rhodophyceae. Fam. Bangiaceae. Goniotrichum Kiitz. G. elegans (Chauv.) Le Jolis. Le Joris, Alg. mar. Cherb., p. 103. Rosenvince, L. Ko_pEruP, The marine of Algae of Denmark, Part I, rg09, p.75, where literature, synonyms etc. are also quoted. Of this species some smaller specimens are found growing epiphytic upon Fler postphonia tenella. The specimens seem to agree very well with the description and figures of ROSENVINGE. In a few of the filaments several rows of cells occurred quite in accordance with the figures D. and E. of this author. Area of distribution: Europe, Mediterranean Sea, Maroc, West Indies, Peru, Easter Island. Erythrotrichia Areschoug. E. carnea (Dillw.) J. Ag. J. Acarpu, ‘Till Algernes Systematik, VI, Ulvaceae (Lunds Univ. Arsskrift, vol. XIX, 1883, p. 15). Rosenvince, Mar. Algae of Denmark, part 1, 1909, p. 67. — Conferva carnea Dillwyn, Brit. Conf., 1809, pl. 84. Conferva ceramicola l.yngb., Hydrophytol., 1819, p. 144, pl. 48 D. 4angia ceramicola Chauvin, Recherch. sur Vorg. de plus. genr. d’Algues, Caen 1842, p. 29—30; Harvey, Phycol. Brit., pl. 317. Erythrotrichia ceramicola Aresch., Phyc. Scand., 1850, p. 210; Le Jouis, Alg. mar. Cherb., 1880, p. 103, pl. 3, figs. 1—2; BeERTHOLD, G., Die Bangiaceen des.Golfes von Neapel, 1882, p. 25. Some small specimens are found epiphytic on Fanza teneWla and Callitham- nition paschale. The plants seem to be in good accordance with the description of ROSENVINGE, |. c. The filaments are about 24 p thick. Their base was quite like the figure of ROSENVINGE (I. c. fig. 8). The cells were mostly rather short, rarely longer than wide. The spore-formation takes place in the well-known manner by means of an oblique wall at the upper end of the cells. Area of distribution: Atlantic coast of Europe and North America, Mediterranean Sea, Maroc, West Indies, West coast of North America etc. Erythrocladia Rosenv. E. subintegra Rosenv. ROSENVINGE, L. KoLpERUpP, The marine Algae of Denmark, Part I, tg09, p. 73. Specimens quite like ROSENVINGE’s figure of older plants and young ones like those I have figured (Marine Algae of D. W. L, vol. II, p. 7—9, fig. 3 a, b) were found on Cladophora spec., mentioned above, p. 249. Area of distribution: Denmark, West Indies, etc. Most probably widely spread. MARINE ALGAE FROM EASTER ISLAND 269 E. vagans nov. spec. — Fig. 15. Thallus divaricatus in superficie corticis hospitis (Cyworzella spec.) vage circumrepens, e filamentis irtegulariter ramosis compositus, Ramificatio aut alterna aut secunda, longa parte filamentorum nuda. Cellulis aut cylindricis, ca. 4—5 latis et 30 uw vel ultra longis, aut magis irregulariter formatis, curvatis sinu- osis, raro furcatis. Chromatophora _ parietalia lo- bata. Sporangia fere rotundata, ca. 4p. lata. The plant (Fig. 15) described here was found ona sterile crust of a Cruoriella. It creeps divaricately in the thick epidermal walls of the host forming, when systems of branches from different plants meet, a more or less dense network above the cells of the host. The filaments are irregularly ramified sending out a side branch now and then at both sides (Fig. 15 c). The branches are sometimes alternate, sometimes more secund; often a branch is borne by every cell in the filament, often, too, long series of cells are destitute of branches. The filaments have apical growth; when the apical cell has reached some length it is divided by a transverse wall into two cells. The cells sometimes are quite or almost cylindrical, about 4—5 u. broad and often more than 30u long, but oftener very irregular in shape, curved, sinuated, even sub- Fig. 15. Erythrocladia vagans nov. spec. a and b furcate. In the cells a well developed pee one ete — ee pee Den ca a ments with sporangia, c. 49/1. parietal, lobed chromatophore is present, now and then leaving a space free; no pyrenoids were observed with certainty. The sporangia (Fig. 15 c,d,e) are cut off from the vegetative cells by means of an oblique curved wall, in accordance with the description of ROSEN- VINGE. As the figure shows, the formation of the sporangia takes place both in the apical cell and in the intercalary cells; in one case the outgrowth of a sinuated cell had become transformed into a sporangium (Fig. 15 d). The spores are roundish, or somewhat flattened on one side and measure about 4 p. 270 F. BORGESEN in diameter; the colour is denser and more deeply red than in the sterile cells, and no chromatophore could be distinguished. The dried plant has a fine rose colour. Area of distribution: Endemic. E. Laurenciae nov. spec. — Fig. 16. Thallus suborbicularis, e filamentis irregulariter radiantibus et ramosis compositus, in circuitu thalli inter se liberis, in media parte plus minus con- natis. Ramificatio aut alterna, aut secunda sat irregularis. Cellulae oblongae-subcylindricae, 4—5 p. latae et 8—10—12 p. longae. Fig. 16. Exrythrocladia Laurenciae nov. spec. Parts of thallus; in a the tissue-of the host is visible, mC seo a In the thick epidermal walls of Lawrencia claviformz7s another representative of the genus L7ythrocladia was found. The plant has long ramose filaments with apical growth and these fila- ments radiate from the centre in all directions covering a more or less extensive area, probably as much as I mm across. The ramification is rather irregular, lateral or alternate. In some specimens it happens that the alternate branching is so regular that the branch with its branchlets gets a pennate appearance. In the centre of the plant the filaments gradually combine to form a more or less coherent layer (Fig. 16 a). The cells are oblong-subcylindrical, often a little curved so that the fila- ments get sinuated; cells 4—5 4. wide and generally 2—2'/2 times as long. In the older parts of the thallus the cells get proportionally shorter. and thicker, 8—10 yp. wide and about as long or a little more. They contain an irregularly lobed parietal chromatophore with one or sometimes two pyrenoids. In some instances a roundish cell were found cut off from mother cell by an oblique wall. Such cells having more dense and homogeneous contents MARINE ALGAE FROM. EASTER ISLAND 271 than the vegetative cells I regard them as asexual sporangia. They are about 5p in diameter. E. Laurenciae comes near the preceding one but differs by its shorter and proportionally thicker cells, by its richer ramification and by the development of a coherent layer in the centre of the thallus. Area of distribution: Endemic. Fam. Helminthocladiaceae. Acrochaetium Naeg. A. (Chantransia) moniliforme Rosenv. — Fig. 17 a. ROSENVINGE, L. Ko tperup, Mar. Algae of Denmark, Part I, p. 99. The enormous distance between Easter Island and the original locality certainly aroused some doubt as to the identity; still, the few specimens observed seem to answer so well to the description and figures of ROSENVINGE that it seems natural to bring them to his species. From the basal spore, that does not differ much from other cells, several filaments arise. The cells are swollen in their middle, about 10—11 p, long and 8—gp wide. In one of the plants a long hair was found. The cell contents was very dense; the pyrenoids situated a little above the centre of the cells suggest a stellate chromatophore. The plants are epiphytic on Daectyopteris repens. Regarding the affinities between this species and Acrochaetium catenulainn Howe! or A. crassifes Borgs.?, compare HOWE’s and my own remarks Il, cc. Area of distribution: Denmark, Easter Island. A, discoideum nov. spec. — Fig. 17 b—d. Thallus epiphyticus, caespitosus, usque ad 300 p, altus. Pars basalis disci- formis, unistratosa substrato (9: foliis Savgassz) firme adhaerens, e filis repentibus ramosis, initio inter se discretis, postea plus minus confluentibus constructa. Cellulae filorum repentium irregulariter curvatae aut plus minus tumidae, lat. 3—4 », long. g—12 py, diametro triplo fere longiores. Fila erecta, e disco basali orta, simplicia, stricta, usque ad 200—300 y. et ultra alta, superne gradatim in pseudopila transformata, e cellulis cylidricis ca. 6—8 ». latis et (in parte basali) ca. 15 y. longis composita. Sporangia oblonge-ovata aut in disco sessilia aut pedicellata, ca. 11—13 p. longa et 6—8 y. lata. This species was found on a leaf of a dried specimen of Sargassum, whose surface it nearly covers. The basal layer spreads widely over the surface of the substratum to which it is firmly attached. A single plant may reach more 1 Howe, M. A., Mar. Algae of Peru, p. 84. 2 BorGESEN, F., Marine Algae of the D. W. I., vol. II, p. 20. 272 F. BORGESEN than !/2 mm in diameter, perhaps even more, but it is difficult to tell the exact size as the plants gradually get fused together. . The basal layer is composed of irregularly ramose filaments which are free at the margin, otherwise forming a monostromatic disc. The disc cells measure 3—4 . across and 9-12». in length; often they are curved and bent. From this basal layer erect long assimilating filaments and sporangia arise. The assimilating filaments are simple, straight and composed of cells about 6—8y, wide and 15». long in the lower part; upwards the cells grow longer Fig. 17. a Acrochaetium moniliforme Rosenv., a single plant with sporangia (two emptied) and a hair, c. 1. b—d A. discoideum nov. spec.: b, ¢ sections through plants, in ¢ the superficial cells of the host; d part of horizontal disc; all c. 4/1. and become almost colourless. The assimilating filaments attain a length of about 200—300 1... The sporangia are sometimes sessile, sometimes pedicellate. They are oblong-ovate, about 11—13 » long and 6—8 » across. Regarding the cell contents this was in a bad condition, the material having been dried, but the chromatophore seems to be parietal with a parietal pyrenoid. This species must be compared with Chantransia leptonema and Ch. re- ducta of ROSENVINGE, both exhibiting about the same mode of growth but differing materially in other respects from A. dzscordeum. Area of distribution: Endemic. MARINE ALGAE FROM EASTER ISLAND 2 ~ Ww A. Ralfsiae nov. spec. — Fig. 18, Io. Thallus caespitulosus, '/2—1 mm altus, e filis endophyticis robustioribus moniliformibus et filis erectis liberis tenuioribus compositus. Fila endophytica inter assimilatores hospitis (Rad/siae expansae) immersa, ramosa, plus minus aggregata, e cellulis subglobosis-ovalibus, long. 20—22 p, lat. 16—20 », composita. Fila libera erecta versus apicem gradatim attenuata, e cellulis cylindricis, 7 p. latis et 18—20—27 py. longis composita, statim a basi sparse et irregulariter ramosa. Rami aut longiores pluricellulares, aut unicellulares sporangiferi. Sporangia oblonge-lanceolata, g—12. lata et 17—20 y, longa. SS SS 1) OSG TCH PO865 big. 18. . If we compare this description with that Fig. 31. Rhizophyllis pacifica nov. spec. a—c transverse sections of thallus; d marginal portion e cross section of margin. ac. 9/1, b,c /1, d 1/1, e 77/1. given above, there is a considerable difference, as I have not observed any branches from the tube of the long cells, nor is it arranged in such a way that it can be called a »Centralachse». The parenchymatic tissue gradually passes into the cortical layer above. The latter is composed of small rounded cells arranged close together in short rows of 2—3 cells which become smaller toward the surface. Seen from above they form a very dense epidermal layer composed of quite small cells, only 4—5y. in diameter and with very thick walls. The epidermal cells of the lower surface are larger, polygonal and c. 40—50 wu. across. Scattered in the cortical layer we finally come across the large characteristic gland-cells (Fig. 30). They are oval-pyriform, often even somewhat lageniform, c. 35». wide and 75 » long, with their broadly rounded base sunk deep down into the parenchymatic tissue, penetrating with their attenuated summit between the cortical cells to just below the upper surface of thallus, leaving a small 284 F. BORGESEN porus open above their mouth-like upper end. When the thallus is seen from above the gland cells are easily observable as'clear bodies between the dark cortical cells. They are uniformly scattered over the thallus, and contain a clear mucilage. The tetrasporangia (Figs. 30, 31) are formed in large sori-on the surface of the thallus. The sporangia are elongate-ovoid, often slightly tapering toward the base, and transversly divided by oblique, often somewhat curved walls into four spores. The sporangia measure c. 40 X 17 w.. As the plant was preserved in alcohol, nothing is known about the colour in a fresh stete. Neither cystocarps nor antheridia were found. According to the descriptions of SCHMITZ, HAUCK and DE TONI and to my observations on dried material, X/. Sguamariae bears considerable resemblance to the new species in its anatomical structure. On the other hand, characters like the irregular crust-shaped thallus, the (presumably) marginal growth etc. remove Kh. pacifica from the Mediterranean species. The generic position of the former remains uncertain as long as we have not seen a more complete material, including sexual plants. Area of distribution: Endemic. Fam. Squamariaceae. By Mme. Dr. A. WEBER VAN BOSSE. Cruoriopsis Duf. C. de Zwaanii Web. v. B. WEBER VAN Bossk, Liste des Algues du Siboga, Rhodophyceae, prem. partie LOZ, sp 14207. In all essential points the alga of Easter Island resembles Cr. de Zwaaniz, but I have seen only tetrasporic plants. It may be that plants bearing cystocarps are different from Cy. de Zwaanit. Area of distribution: Island Nias, west coast of Sumatra, Easter Island. Ethelia W. v. B. E. pacifica nov. spec. Thallo substrato affixo, + 3004. crasso, constante e mesothallo et peri- thallo superiore et inferiore. Mesothallo constante e filamentis ramosis, flabelli- formibus. Perithallo superiore constante e filamentis obliquis in parte basal, terminatis in partem superiorem satis latam erectam. Perithallo inferiore solum filamentis obliquis munito, quorum pluria exeunt in rhizoidea. Cellulis perithalli superioris + 3—6 yp. latis et 8—9 yp. altis, cellulis perithalli inferioris 5—9 pv. latis et 18, altis. Organis fructificationis non visis. MARINE ALGAE FROM EASTER ISLAND 285 Thallus adhering firmly to the substratum + 300 y, thick, consisting of a mesothallus and a perithallus superior and inferior. Mesothallus consisting of flabella-like branched filaments. Perithallus superior consisting of oblique fila- ments in the basal part, ending in a pretty broad straight upper part. Perithallus inferior having only oblique filaments, many of these running out in rhizoids. Cells of perithallus superior + 3--6y broad and 8—g vp. high. Cells of perithallus inferior 5—9 uy. broad and 18 yp. high. Organs of fructification not observed. As long as the organs of fructification are unknown it will remain doubtful whether an alga, with a midlayer of cells giving off ascending and descending filaments, is a representative of the genus L¢/elia (the type of whichis £. /os- fiev) or a representative of the genera Cruoriclla or Peyssonnela and differing from the types of these genera no more than Aalfsza deusta differs from Ralfsia clavata.. Ethelia Fosltet has a mesotballus with an extremely well defined perithallus superior and inferior and nemathecia with paraphyses that have top- cells adhering together and covering almost the tetrasporangium in a young state. The new alga from Easter Island does not give any light on this matter as none of my many slides showed signs of fructification. Its anatomical structure mostly resembles Ethelia (Peyssonnelia) biradiata from the Indian Ocean, It differs from this alga by its mode of growth, for /. facifica adheres firmly to the substratum whereas “£. dzvadiata may be very easily detached and seems even here and there to lie loose on the substratum while growing. The thallus of £. pacifica adheres so firmly that I found it impossible to make a good slide through the thallus, unless I cut the substratum at the same time. The decalcifying of the substratum, consisting mostly of worm tubes, required much time as the tubes were of different consistence. The use of various acids had also a destructive influence on the contents of the cells and propitiated the swelling of the membranes, circumstances that were unfavourable to the making of good slides. Another difference between the two algae lies in the size of the filaments. The ascending filaments of /. d¢radiata are + 8». broad and are 2—3 times as high as broad. Those of £. pacifica are broad 3—6y, and 8—gu. high. The filaments of the perithallus inferior are in both algae longer and broader than the ascending filaments. The colour was gone in the specimens of £. pactfica that I had for examination. £&. dzradiata has a bright red colour. Area of distribution: Endemic. Fam. -Corallinaceae, Melobesicaec, par Mme PAUL LEMOINE. M. SKOTTSBERG a recueilli quelques espéces intéressantes dans cette ile isolée d’ou jusqu’ici non ne connaissait que deux especes: Porolithon onkodes (Heyd.) Fosl. et Porolithon praetextatum Fosl.; il a dailleurs retrouvé la premicre de ces deux especes, associée a d’autres espéces pacifiques littorales. 1 conf. ReEINKE, Algenflora der Westl. Ostsee deutschen Anteils, 1889, p. 48. 286 F. BORGESEN Lithothamnion (Phil.) Fosl. L. siamense Fosl. -— Fig. 32a. M. Fosuiz in Flora of Koh Chang (Bot. Tidsskr. XXIV, 1901, p. 19 (non f. simulans)); Id. in WEBER et Fosuiz, The Corallinaceae of Siboga Expedition, Siboga Expeditie LXI, 1904, p-1o, fig. 3, Bl ie 1—o- N’ayant jamais vu déchantillon de ZL. szamense j'ai quelque hésitation a rapporter a cette espece un échantillon de I'Ile de Paques dont les caracteres paraissent cependant coincider parfaitement avec la description de M. FOsLig; cependant les figures I, 2 qui représentent le type de l’espece en donnent une idée confuse. L’échantillon de l’Ile de Paques mesure 2 a 3 cm; il est formé de lamelles extremement minces (50 a 70y, d€paisseur apres décalcification) et fragiles, de taille variable, irrégulicrement superposées et imbriquées de maniére a constituer un complexe de crottes d’une épaisseur de 2 a 5 mm; ces lamelles, finement stri¢es sur leur face inférieure, sont lisses sur leur face supérieure; la plupart sont plus ou moins horizontales, mais souvent aussi elles sont contournées ou recourbées, ou s’enroulent en forme de tiges creuses; elles donnent également naissance a de petites excroissances; enfin on remarque aussi des lamelles disposées verticalement. Cet échantillon parait correspondre a une variété plus résistance que ceux décrits par M. FOSLIE; cet auteur a en effet observé des crottes de 30 a 604 d’épaisseur seulement qui, en se superposant n’atteignaient que 300 pu. d’epaisseur. Pour les autres caractéres sa description s’applique a |’échantillon en question, et les divergences qu’on peut relever sont de faible importance. Chacune des feuilles qui constitue l’échantillon, de I’Ile de Paques est constituée par l’hypothalle et le périthalle; dans les Jamelles disposées horizontale- ment le périthalle est plus développé que l’hypothalle; l’inverse se produit dans les parties ot les lamelles sont recourbées et presque libres (fig. 32a). L’hypo- thalle est formé de cellules courtes de 5 a 15 y de long et 5 a 8y de large; dans certaines lamelles les cellules ne dépassent pas 7 y. de longueur; les cellules hypothalliennes sont rectangulaires a angles arrondis. Le perithalle est formé de files laches dont les cellules mesurent 3 a gu X 3 a 7 Ub. Je crois avoir observé les deux sortes de conceptacles: les conceptacles a sporanges ont un toit légerement bombé, leur diameétre est de 280 a 480 1; je nai vu ni les pores du toit du conceptacle, ni les spores. Les conceptacles a cystocarpes ont un toit de forme convexe percé d'un pore au sommet; leur diametre varie de 300 a 500». L. siamense se rapproche par sa structure et la fragilité de ses crodtes de Lithothamnium mesomorphum Fos\. de \ Atlantique (Bahamas, Bermudes, Antilles). Répartition géographique: ZL. stamense a été découvert dans le Golfe de Siam, puis retrouvé ensuite en différentes régions du Pacifique: Sumbawa (Est de Java), Archipel Sulu, Ile Nusa-Laut, Ile Kei; il a été recueilli a des profondeurs variant de 5 a 69 metres. A I'Ile de Paques il a été recueilli dans la zone littorale a Hanga Piko. MARINE ALGAE FROM EASTER ISLAND 287 Lithophyllum (Phil.) Fosl. Deux des especes de Lithophyllum de |\'lle de Paques n’ont aucunement la structure typique de ce genre, telle que je l’ai caractérisée au point de vue anatomique, mais elles en possedent les organes reproducteurs; aussi je les laisse provisoirement dans ce genre, dans lequel elles constituent une section spéciale, ainsi que quelques autres espeéces deja signalces dans différentes régions du globe. L.(?) samoense Fosl. — Fig. 32 b. M. Fostie: Algologiske notiser I, D. K. N. Vid. Selsk. Skrifter, 1906, no. 2, p. 20. M. FosLiE n’a jamais figuré cette espcce, quil avait tout d’abord réunie au L. decipiens Fosl. En comparaison de l’échantillon de l'Ile de Paques, j'ai étudié un échantillon de l’Herbier BORNET, provenant de Tahiti, dont M. FOsLig fait d’ailleurs mention loc. cit. L. samoense movtre en coupe verticale un hypothalle peu développé, formé de files horizontales dont les cellules mesurent 6 a 17 ». de longueur et 3 a7 u. de large, quelquefois 9. Le périthalle est formé de tres petites cellules carrées ou rectangulaires de 3 a 5 py, atteignant 7 et 8p de longueur et 3 a 7 yu, de large. L’espéce forme souvent plusieurs thalles superposés. Jai noté, pour les conceptacles les dimensions de 125 a 280, apres décalcification, tandis que FOSLIE indique 100 a 200; ces conceptacles convexes deviennent ensuite déprimés ainsi que le montre l’échantillon de | Herbier BORNET. JL. samoense a été recueilli sur un caillon, associé a Por. onkodes et Lithophyllum rasile, et il recouvre MWelobesia (Litholepis) accola; \\ forme une croute a surface granuleux avec une bordure lisérée de blanc; la limite de chacune des crottes qui vit sur le méme support est marquée par un rebord crénelé. Répartition géographique: Tahiti (Herbier BORNET, Museum d'Histoire Naturelle de Paris); Samoa, Ile Savaii (RECHINGER ex FOSLIE). L.(?) myriocarpum (Fosl.) —- Fig. 32 c. Lithothamnion myriocarpum Foslie, On some Lithoth. in Det K. Norske Vidensk. Selsk. Skr., Trondhjem 1897, no. 1, p. 19. Gonzolithon myriocarpum Forslie, Siboga Expeditie 1904, p. 45, Pl. IX fig. 6, 7; Idem in Alg. Not. III, Det K. Norske Vidensk. Selsk. Skr. 1906, no. 8 (1907), p. 14 (f. confragosa); Idem Alg. Not. VI, Ibid. 1909 no. 2, p.9. Non Avelobesia myriocarpa Crouan, = M.(Pliostroma) zonalis (Crouan) Fos}. L. myrtocarpum forme des crottes d’aspect peu caractéristique, assez fréquentes, semble-t-il dans le Pacifique. En coupe l’hypothalle montre des cellules plus ou moins rectangulaires de 7 a 15 de longueur et 3 a 7 uw de largeur; le périthalle est formé de cellules de forme assez variable; les cellules d’une méme file communiquent entre elles par un pore brillant, et souvent par une partie retrécie en forme de petit canal; le méme caractére s’observe dans le Lithothamnium calcareum a Europe. Répartition géographique: Mer Rouge, Indes, Célebes, Iles Chagos, Seychelles, N"* Guinée, Lucipara, Dammer, Mangareva, Tahiti etc. 288 F. BORGESEN bs es ~w fa) ‘S Rap peepee @ d as po mace esces ARN Si 5 SB Fig. 32. a coupe verticale du thalle de Lithothamnium siamense,; h hypothalle, # périthalle; b coupe du thalle de Lithophyllum(?) samoense; c coupe d’un fragment de crotte de L.(?) myriocarpum,; a Lithophyllum (Derm.) rasile recouvrant Melobesia (L.) accola dont les cellules sont en grisé; e cellules de Melobesta (Litholepis) accola, a de Tahiti, 6 de PIle de Paques; f coupe d’un thalle de MWelobesia paschalis; g cellules de Afelobesia paschalis vues de dessus. L. (Dermatolithon) rasile Fosl. — Fig. 32 d. M. Fosuie: Alg. Not. III, Det K. Norske Vidensk. Selsk. Skr., Trondhjem 1906, no. 8 (1907), p. 34 et Alg. Not. VI, Ibid. 1909, no. 2, p. 57. Cette petite espéce est représentée dans la collection de I'Ile de Paques par un petit thalle d’un rose vif, de 10 mm X 12 mm qui recouvre AZelobesza (Litho- lepis) accola; sa bordure, lisérée de blanc, montre quelques stries concentriques peu marquées; a la surface de Lithophyllum(?) samoense on observe également quelques trés petits thalles trés jeunes qui appartiennent a la méme espece. M. FosLig indique pour cette espéce une variation assez considérable dans la longueur des cellules hypothalliennes (14 a 54). En effet dans une coupe dans laquelle l’hypothalle était seulement recouvert par une rangée de cellules corticales, les cellules hypothalliennes mesuraient 20 a 60 y.; au contraire dans d'autres coupes (fig. 32d) j'ai observé lhypothalle 2, formé de cellules de 15 u. environ de hauteur, mais il était surmonté de 3 a 4 rangées péerithalliennes; celles-ci sont formées de cellules de 8 a 12 y. de hauteur; la derniere rangée est recouverte de cellules corticales; d’aprés la description de M. Fosig, les cellules périthalliennes peuvent également subir une assez grande variation de taille et pourraient atteindre 36 w. Répartition géographique: Océan Indien: Laquedives; Pacifique: Timor (échantillon récolté par M™’ WEBER VAN BosSE, nommé L. papzllosum f.? par FosLiE in Siboga Expeditie 1904, p. 63); Tahiti (leg. SEURAT, Herbier du Museum de Paris). MARINE ALGAE FROM EASTER ISLAND 289 Melobesia (Lamx.) Fosl. M. (Litholepis) accola (Fosl.) Lem. — Fig. 32 e. On sait que M. FOSLIE a crée Litholefis pour des Melobesza dont le thalle est formé d'une seule rangée de cellules, et dans lesquelles les thalles se superposent de telle sorte qu’en coupe on observe souvent plusieurs rangées. Dans l’échantillon de l'Ile de Paques j’ai observé jusqu’a 5 a 6 rangées super- posées, toutes semblables (fig. 32 e, b); les cellules sont soit rectangulaires, soit allongées verticalement, souvent aussi de forme assez irrégulicre; cependant leur dimension ne varie que de 5 a 10. pour la hauteur et 5 a 12, pour la largeur. M. FOSLIE avait indiqué la dimension de 14 a 30. pour les cellules de M. accola; mais j'ai étudié un eéchantillon de Tahiti qui fait partie de la collec- tion du Museum d'Histoire Naturelle et qui a été déterminé par lui, et dans la coupe obtenue j'ai observé une rangée de cellules de 5 a 12y, de hauteur et 7 a 10. de largeur (fig. 32 e, a); l'espece de Tahiti est donc bien semblable a celle de l’Ile de Paques. L’aspect de cette petite espece est curieux: les crottes sont tres minces et sont discontinues; elles sont pourvues de tres nombreux conceptacles de 100 a 180, environ, percés d’un pore. Lorsqu’on observe le thalle de dessus, les cellules, rectangulaires-arrondies, mesurent 7 a 104. 5 a 74, disposées en files. L’un des échantillons de l’Ile de Paques provient de Hanga Piko, région littorale; l'autre sans indication précise partage le meme support que les autres especes étudiées, et est recouvert en certains points par L. raszle et L. sa- MOeNse. Répartition géographique: Tahiti, Hao (leg. SEURAT, Herbier du Museum d'Histoire Naturelle de Paris), Ile de Paques. M. paschalis nov. sp. — Fig. 32 f, g. La seule espéce é€piphyte recueillie par M. SKOTTSBERG me parait étre une espece nouvelle; elle forme sur Zonaria variegata une croute extremement mince, a peine calcifiée, de 3 cm environ, de couleur grise sans aucune bordure. A la loupe on remarque que les conceptacles se trouvent au centre de petits épaississements qui forment comme de minuscules crotttes secondaires qui se détachent facilement. En coupe on constate que la partie la plus mince du thalle est formée d'une seule rangée de cellules; celles-ci ont une forme assez variable et souvent elles ne se touchent pas; leur hauteur est d’environ 5 29 uw, leur largeur 5.a15 p. Ces cellules, vues de dessus, sont encore plus irréguliéres dans leur forme; et variables dans leur taille; certaines sont transformées en hétérocystes (fig. 32 g). Si on fait une coupe dans une région fructifiée du thalle, on peut suivre le processus de l’épaississement du thalle; j’ai figuré (fig. 32 f) l’aspect de la 19 — 2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 290 F. BORGESEN coupe telle que je l’ai obtenue; on remarquera l’irrégularité de forme des cellules, leur manque de cohésion; dans les parties formées de 2 rangees il y a souvent un espace entre les deux rangées qui paraissent alors indépendantes l'une de autre. Dans les parties les plus épaisses, j'ai observé 5 cellules superposees, toutes de taille A peu pres semblable, 3a 6p, de hauteur; l’épaisseur maximum du thalle est de 20 a 25 p. Les conceptacles, de tres petite taille, mesurent environ 160 y, de diametre; l'un d’eux paraissait contenir des bispores de 30 a 39. 15 a 204, et une tétraspore de 50. X 30 p. Melobesia farinosa était jusqu’ici la seule espece pourvue d’hétérocystes; mais le thalle vu de dessus est bien différent de celui de Melobesia paschals, les cellules sont de forme beaucoup plus constante et elles sont alignées en files réguliéres, méme dans la var. So/mszana ou le tissu est beaucoup plus lache et lacuneux que dans le type. Des différences fondamentales existent aussi dans la dimension des con- ceptacles, et la structure de l’espece vue en coupe. Porolithon Fosl. P. onkodes (Heydr.) Fosl. Voir la bibliographie dans M™¢ P. Lemoine, Annales Inst. Océanogr. Monaco, tf wietasc, 1 (LOK). pant p ; \ ae) Hf Li } {\ y \ WY Pi f Wy A Wp 4) f ) f i Oe in/4 W f ff VY, \! B Wy * i aN I x Ma A » ANGE, ee WAY 0 yy, Gi RN \ , HY WA , OS se. Fig. 35. Callithamnion paschale nov. spec. a part of Q specimen, c. “/1; b of o specimen, c. 18/1; © of tetrasporic plant, c. 15/1; d base of a plant, c. %/1. C. paschale recalls scopulorum as described by the author in »Botany of the Feerdes», Part. I], p. 377. But C. scopulorum differs, among other things, in being often bi-tripinate, the cells in the main stem are proportionally much longer, the branches taper much more against their apex, the tetrasporangia are more spherical, etc. C.-Pennula Grunow from St. Paul’s Island, Novara-Exp. p. 60, pl. 6 fig. 1, with which our new species might perhaps also be compared, is a much larger and more branched plant with a dense nest of basal rhizoids. | Area of distribution: Endemic. 296 F. BORGESEN Crouania J. Ag. C. attenuata (Bonnem.) J. Ag. AGARDH, J., Algae mediterr., 1842, p. 83. Some fragments of a very small form were found. GRUNOW in >Alg. Novara», p. 62 mentions having found a very delicate form at Tahiti. Area of distribution: Mediterranean Sea, warmer shores of the Atlantic and Pacific Oceans. Carpoblepharis Kiitz. C. Schmitziana (Reinb.) Okamura. — Fig. 36. Oxamura, K., On Microcladia and Carpoblepharis (Bot. Mag., vol. XIV, 1900, p- 8, pl. I, figs. 14—17. Icones of Jap. Alg., vol. I, pl- I, figs. 5—18). — Gloto- thamnion Schmitsianum Reinbold, Hedwigia, 1895, Bd. 34, p. 205, pl. Ill. — Rezn- boldiella Schmitziana (Reinb.) De-Toni, Syllog. Alg., vol. IV, sect. III, p. 1498. On a stone with several other small algae some few bits of a very delicate species were found, which I feel convinced belongs to C. Schmztzzana, first described by REINBOLD, even if the Easter Island plant seems to be much smaller and differs a little in its habit from REINBOLD’s figure. The plant grew on a Cruoriella to which it is fixed by means of numerous rhizoids growing out mostly in bundles from cortical cells. The decumbent rhizome-like parts of the thallus are about 60» thick. From this now and then branches are given off; some of these soon stop their growth, become more or less vertical, and reach a length of up to 200—300 »,; but most branches Fig. 36. Carpoblepharis Schmitz. Show an indefinite growth like the mother filament, tana (Reinb.) Okam., fragments fixing themselves to the substratum. oes g Sex The erect branches become more or less flattened with a broad obtuse tip; they sometimes are as much as 50». wide, but often only half that size. Longitudinal growth is performed by a broad top-cell. The segments cut off often remain distinct long after they have become divided into smaller cells. The main shoots are covered with a dense, untransparent cortex. The material is sterile. I have followed OKAMURA in referring this species to Carpoblepharis. It seems to be a highly variable plant, more so if the small reduced forms, referred here with some doubt by Mme. WEBER’, really belong to the same species. The Malayan plant has the creeping rhizome-like base reduced to a mono- siphonous filament, a rather remarkable fact. Area of distribution: Japan, Malayan Archip., Easter Island. 1 Liste des Algues du Siboga III, Rhodophyceae II, p. 336, figs. 126—7. MARINE ALGAE FROM EASTER ISLAND 297 Ceramothamnion Richards. C. Codii Richards. RicHaRDs, Ceramothamnion Codii, A new rhodophyceous alga (Bull. Torrey bot. Club, vol. 28, p. 57). Some few small loose-lying fragments have been found only which I think are referable to this plant. What species the host plant has been I can not say; no Codium was among the material handed over to me. Dr. H. E. PETERSEN kindly inspected the material and confirmed the determination. Area of distribution: Bermuda, Easter Island. Ceramium (Roth) Lyngb. By Dr. H. E. PETERSEN. ©] cerociatum’ Collins*& Hervey. — Fig.737- Cottins & Hervey, The alge of Bermuda, 1917, p. 144. In referring the specimens from Easter Island to this species I base this upon the figures of Mme. WEBER, Liste des algues du Siboga, III, Rhodo- phyceae, p. 331. Fig. 37. Ceramium cruciatum Collins & Hervey. a top of filament, c. ®/1; b part of filament lower down, c. 749/1; ¢ tetrasporangia, c. 74°/1. According to CoLLins & HERYEY Ceramium cructatum has »cortication consisting of cells elongate in the direction of the filament». Mme. WEBER having seen an original specimen of this species now gives a figure of a cortical belt, in which the cells are not arranged in longitudinal rows as in the figures 208 F. BORGESEN of COLLINS and Hervey. As the specimens from Easter Island (Fig. 37 b) regarding the shape of the cortical cells agree with the statement of Mme. WEBER I think I need not lay stress upon the remarks of COLLINS and HERVEY as to his point. Specimens with cystocarps and tetraspores (Fig. 37 c) were present. In the tetrasporic specimens the tips of the filaments were flattened in a peculiar way (Fig. 37 a). Area of distribution: Bermuda, Celebes, Easter Island. C. Skottsbergii H. E. Petersen, nov. spec. — Fig. 38. In the material a small Cevamzum is found characterized by very narrow cortical belts in which several pluricellular spinelike hairs are present. I consider this plant to be a new species. Here the diagnosis: Fig. 38. Ceramium Skottsbergii nov. spec. Part of filament, showing spinose cortical belts, c. 1°°/1. Fronde repente, irregulariter ramosa; zonis semper discretis, paucis cellulis constantibus, altitudine zonarum 15 —25 »., seepe duobus cellulis formata. Cellulis axialibus usque ad 80 p, altis; 7o—75 wp latis. Pilis verticillatis e cellulis zonarum formatis, articulatis, ex 3 cellulis compositis, c. 30—35 ». longis, c. 8», latis, obtusis. Organa fructificationis non visa. Area of distribution: Endemic. Fam. Rhodomelaceae. Laurencia Lamour. L. claviformis nov. spec. — Fig. 39. Planta caespitosa, ca. 11/,—2 cm alta, e ramis basalibus repentibus et ramis erectis composita. Rami basales rhizoideis robustis saxo adfixi. Rami MARINE ALGAE FROM EASTER ISLAND 299 erecti, teretes, clavati, in inferiori parte ca. I mm lati, in superiori ca. 2 mm, apice late rotundato, aut simplices, aut ramosi, ramulis clavatis, sparsis, paucis, irregulariter exeuntibus, interdum ramosis, ramellos in superiore parte gerentibus. Tetrasporangia sparsa in superiore parte ramorum orta. To judge from the specimens collected the plant forms low dense tufts, ca. 1'/2—2 cm high, on rocks between tide marks. The decumbent creeping filaments, about 1—1'/s mm thick, are fastened to the substra- tum by means of short, thick, vigorous hapters (Fig. 39 a). From these prostrate filaments erect ones grow up. These measure at their base about I mm in diam., gradually thickened upwards, attaining twice their basal diameter and slightly clavate. Of the erect filaments the more robust ones are branched in their upper half. They carry some few lateral branches, especially near their upper end, where several may be clustered together. The branches grow out to all sides without any order, they are clavate like the primary branch and sometimes carry sub- terminal branchlets. Tetrasporic plants only were gathered. The tetraspores are found in the cortical |. oe eal oh ee eee : Fig.39. Laurencta claviformis nov. spec. layer, scattered over the surface, especially ~~ Parts of plants, a c. 2/1, bc. /1. in the upper end of the branches. As to the limitation of the species in this troublesome genus I refer to my remarks in »The marine Algae of the D. W. I», vol. II, p. 244. I regret that, in describing a new species, I add to the difficulties. Nevertheless I think this is the best way to take until the very desirable revision of the whole genus has been undertaken. The present plant surely comes near a group of species (f. i. L. perforata, radicans, vaga, decumbens etc., comp. KUYZING, Tab. Phycol., vol. XV, pl. 49—51) about which J. AGARDH, in »Epicrisis», p. 649 says: »nec mihi patet quomodo distinguantur». Area of distribution: Endemic. Chondria Ag. Subgenus Coelochondria Falkenb. Ch. repens nov. spec. — Figs. 40, 41. Thallus teres, minutus, ca. I—2 mm altus et 200—400 », latus, in aliis algis epiphyticus, partim e ramis repentibus, rhizoideis brevibus robustis adfixis, par- tim e ramis curtis, erectis constructus. 300 F. BORGESEN Fig. 40. Chondria repens nov. spec., part of specimen, c. 1°/1. Rami repentes ramosi, ramis aut decumbentibus, repentibus, axi primario similibus, aut erectis, curtis, obovatis, simplicibus, ca. '/3—1 mm altis et 300— 350u. latis, tetrasporangiis instructl. Tetrasporangia in superiori parte ramorum sparsa, ca. 80—90 y, lata. Of this little plant some very few fragments are found only, the longest specimen being about 1 cm long. The terete thallus has creeping basal filaments fixed to other algae by means of short thick hapters formed by a bundle of rhizoids. From this basal part shorter or longer branches are given off. Some of these branches may grow out to long shoots like the mother branch fixing themselves to the substratum in a similar way but most of the branches remain short, erect and become fertile. Gradually they become more or less distinctly clavate and produce tetrasporangia in their upper end. The ripe tetrasporangia are about 80—90 y. in diameter. The cortical layer is rather thick and the central strand generally not visible. Seen from above the surface cells are roundish to polygonal with rather thick walls. A transverse section (Fig. 41) shows that the thallus is composed of a parenchymatic tissue, the central axis not being especially distinct. In » Algues du Siboga», HI Rhodophyceae, p. 349 Mme. WEBER has described a small species Chondria minutula, of nearly the same size and habit, but ei seins otherwise widely different, belonging as it does to spec. Cross section of Fatt, the group Euchondria Falkenb. Cc. 12°/1, Area of distribution: Endemic. MARINE ALGAE FROM EASTER ISLAND 301 Polysiphonia Grey. Pe spec: A small antheridial plant with four pericentral cells. From a decumbent filament fixed to the substratum by means of rhizoids erect filaments arise. The filaments are about 50 thick and the length of the cells about 80. The specimen bears considerable likeness to ASKENASY’s’ figure of an antheridial plant of Polysiphonia mollis Hook f. et Harv. Herposiphonia Naeg. H. tenella (C. Ag.) Naeg. — Figs. 42, 43. NAce., C., Herposiphonia (in ScHLEIDEN und NAGELI, Zeitschr. f. wiss. Bot., 3—4 Heft, 1846, p. 238, tab. VIII). FAaLKENBeRG, P., Rhodomelaceen, p. 304. BOrGESEN, F., Marine Algae D. W. I., vol. I, pp. 286 and 472. In referring the plant from Easter Island to this species it must be pointed out that before we know the male plant no exact determination can be made; eigemiy remarks |. c., _p., 470. =—-=. i (es Bo: Fig. 42. Herposiphonia tenella (C. Ag.) Naeg., part of specimen growing on Valonia. C. **/s. Sterile specimens were found on Valonia ventricosa. They had 5—8 peri- central cells. The creeping, prostrate main filaments are fixed to the wall of the host by means of vigorous rhizoids quite in the same way as found in the West Indian plant. ! Asxenasy, E., Ueber einige australische Meeresalgen. Flora, 1894, p. 13, pl. 4, fig. 22. 302 F. BORGESEN The mutual arrangement of the branchlets and branches is seen in Fig. 42. This shows that the branchlets are, as a rule, developed from every fourth segment; the next has a branch and the branches alternate on the right and left side of the thallus. But it happens that the branchlets are developed from every third or rarely from every second segment. The branchlets are of rather variable length; in some specimens I have counted up to 50 segments. The segments are longest in the middle of the branchlets, shorter near the base and summit. The diameter of the branchlets is about 50. The tricho- blasts are as a rule poorly developed. On Zonaria variegata | found some few fragments of an other form, both cystocarpic and tetrasporic, but unfortunately no male plant. Of the female plant Fig. 43 gives an illustration; it is much more robust than the sterile one from Valonza. It also differs from this by the scarce development of branches, which generally are arranged without any order. Gone Fig. 43. Herposiphonia tenella(C. Ag.) Naeg. 9. a part of creeping plant, c. /1; b growing apex Of Same, ic:.'"2/a; 1c youns cysrocarp,| c. 170/:1: The trichoblasts are well developed in the female plant. The cystocarps are formed in the second segment of the trichoblast, the lowermost becoming polysiphonous. The form of the young cystocarp (Fig. 43 c) very much resembles what I have found in Lophostphonia cristata (comp. my fig. 432 1. c.). The trichogyne is thick and comparatively short. The ripe cystocarp is urn-shaped with a short broad neck; it is about 550 u. long and 460 y, broad. The tetrasporic specimens have longer and much more slender branchlets than the female plant; in the mutual arrangement of the branchlets and branches they agreed with the plant found on Valonza. The tetrasporic branchlets are about 60». thick and 1,5 mm long. ; Area of distribution: Mediterranean Sea, Morocco, West Indies, Malayan Archipelago, Easter Island. A small piece of a plant with 11 pericentral cells may perhaps belong to Hf. subdisticha. Okamura (Bot. Mag. XII, 1899, p. 11, pl. I figs. 12—14). It resembles the figures quoted, but being sterile and lacking growth points it is insufficient for a safe determination. MARINE ALGAE FROM EASTER ISLAND 303 Dipterosiphonia Schmitz. & Fallenb. D. dendritica (Ag.) Falkenb. — Fig. 44. FALKENBERG, P., Rhodomelaceen, p. 324. BoRGESEN, Fo Mar... Alo. D. W,'T., p...302. — Hutchinsia dendritica Ag., Systema, p. 146; Spec- ies Alg., vol. Il, p. 104. — Polysiphonia dendri- foal no:,) Spec. -Alg:,,vol.-Ily pars. 3,.. p+. 91.6: Some small sterile fragments were found on Laurencia claviformts. In one respect it differs essentially from the West Indian plant, as described by me |. c.; in this, the branchlets with definite growth are always undivided, while, in the Easter Island form, they often bear a few spine like side branches. These are mostly situated above the middle of the branchlets and their number rarely surpasses two or three. They form acute angles with the main branch. Otherwise, the Easter Island plant seems to agree perfectly with the West Indian one. Like this it has five pericentral cells, three on the upper side, two below, and it is attached to the host plant by meams of vigorous hapters, Hig oi bar sananeenten Gb Dipiere- which pierce the sides of the host plant like siphonia denaritica(Ag.) Falkenb. C. ®/1. claws. Area of distribution: Brazil, West Indies, Australia, Easter Island. Dasya C. Ag. D. villosa Harv. Harvey, W. H., Algze of Tasmania, London Journal of Botany, vol. III, 1844, p- 433- The specimens had been preserved in formaline and were in a state of dissolution, falling to pieces at the least touch. Nevertheless I think the determination is right, the plants being in good accordance with the descrip- tion of HARVEY and with KU1TziNnG’s Figure; I have also been able to compare them with a specimen of HaARVEY’s Australian Alga, No. 218 from Van Diemen’s Land. When young the stichidia are rather short and thick, ovate to lanceolate, when older they become longer and subcylindrical with attenuated summit, but perhaps the great length is due to the bad preservation, all the tetraspores having fallen out and the cells getting more or less separated. Area of distribution: Tasmania, Easter Island. 304 F. BORGESEN Falkenbergia Schmitz. F. rufolanosa (Harv.) Schmitz. in ENGLER u. PRANTL, Natirl. Pflanzen- familien, p. 479. FALKENBERG, P., Rhodomelaceen, p. 690. — VFolysiphonia rufolanosa Harvey, Marine Botany of W. Austr., no. 87 in Trans. Acad. vol. XXII; Ktrzine, Tab., Phycol., vol. 14, pl. 54. Only a few small pieces of this plant were found, too small in fact fora safe determination. In their general appearance the specimens resemble the figure of KUTzING; this is, however, as pointed out by FALKENBERG, not very accurate. The thallus was about 4o yp. thick and the length of the cells amounts to 30 u. On the other hand, it cannot be denied that this plant seems to concord with the form of Falkenbergia Hildebrandit which 1 discovered in the West Indies (I. c., p. 331). According to my note, these species most probably are mere forms of the same species, a conclusion to which YENDO’ also has arrived, having been able to compare his plant with the original specimen in Herb. HaRvEY in Dublin and referring to my figures and description of the West Indian plant. 7 Area of distribution: Australia, Japan, Malayan Archipelago, and most probably, West Indies etc. Fam, Delesseriaceae. Nitophyllum Grev. N. spec. — Fig. 45. A small antheridial plant of a Vetophyllum was found mixed up ina Galaxaura tuft. The height of the whole plant found was 6 mm only. It has no veins, is monostromatic and belongs to the group Leptostroma of J. AGARDH. The growth takes place by means of a well developed top-cell. The antheridial sori are situated in the middle of leaf-like thallus-segments. They are found on both sides of the thallus. The cells of the thallus are polygonal with thick walls; their diameter is about 40 1. To the group Leftostroma J. Agardh refers 3 species and in »Sylloge Algarum» DE-TONI mentions a few additional species, g. 45. Witophyllum spec., o spe- 1 YENDO, Notes on Algae new to Japan, IV (Bot. cimens) Ca 22) Mag., vol. XXX, 1916, p. 63). Fig. MARINE ALGAE FROM EASTER ISLAND 395 all from the European Atlantic coast and the Mediterranean Sea. Most likely therefore the present plant is the representative of a new species. In order to describe it we must know at least also the tetrasporic plant, as the male plant in Netophyllum is often very much reduced in size. Gymnogongrus Mart. G. aequicrassus nov. spec. — Figs. 46, 47. Frons caespitosa, e filis decumbentibus, repentibus et filis erectis composita. Fila decumbentia irregulariter ramosa, aggregata et inter se contexta. Fila erecta teretiuscula, interdum simplicia saepe di- vel trichotoma, ad apicem versus non attenuata, apex late rotundatus. Cystocarpia singulas apicibus frondis immersa, rotundata, superne in partem sterilem attenuatam exeuntia. ® DOLD. Fig. 46. Gyminogongrus aequicrassus NOV. Spec. Fig. 47. Gymnogongrus aegui- Caf crassus nov. spec. Length sec- tion of thallus, c. 49/1. The plant grows gregariously in dense tufts about 2—3 cm high upon rocks between tide marks. The tuftlike growth originates from the numerous erect filaments growing out from the prostrate creeping ones. The latter filaments are irregularly bent and branched, gradually more or less united into an irregularly lobed disc. The erect filaments are nearly terete or somewhat compressed, the transverse section being mostly oval. The filaments are a few times dichotomously, rarely tri- chotomously forked, but quite simple ones also occur. The surface of the 20 —.2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. Il. 306 F. BORGESEN thallus is uneven and irregularly waved. The tips of the shoots are broadly rounded. The angle between the branches approaches 90°. A few cystocarps occur; these are found near the end of the branches and are almost globular. Above the cystocarp a short sterile narrow prolonga- tion is observed. The plant is of a rigid cartilaginous or corneous texture. On a longitudinal section of the thallus (Fig. 47) we find that the sub- cylindrical cells in the centre are arranged more or less in rows; outwards the cells grow shorter with very thick walls and gradually pass into the cortical layer, composed of short rows of small oval cells — upon the whole a picture quite like that of Chondrus crispus as given by KYLIN in Studien tber die Entwicklungsgeschichte der Florideen (K. Svenska Vetenskapsakad. Handlingar. Bd: 163, 1923). The new species appears to be most nearly related to the small species G. densus and G. pygmaeus. Area of distribution: Endemic. Fam. Sphaerococcaceae. Hypnea Lamx. H. Esperi Bory. — Fig. 48. Bory, Voyage de la Coquille, p. 157. Kwtrzinc, Spec. Alg., p. 759; Tabul. Phycol., vol. 18; plz 26; 7a, b, €: The thallus forms small entangled masses among other algae. The filaments are irregularly ramified, often antlerlike in shape, the summit of the filaments often being somewhat curved with short, spine-like side-branches gradually growing out from the convex side. The branches are sometimes short, spinelike, sometimes long and of indefinite growth. The tetra- sporangia are formed in the branchlets. When a branch comes near a fa- vourable substratum short thick discoid hapters are formed by means of which the plant is fastened to other algae. In the same way the branches frequently anastomose. The terete thallus is 4o0—500 thick. The specimens found seem to be in good accordance with the fig. of KUTZING in Tab. Phycologicae (I. c.). Big, 48.) Biypnea-Esperi Bory! Patt of spect This plant is closely related to the men with a tetrasporangium sorus. C. 1. form I called Hypuea spinella in my West MARINE ALGAE FROM EASTER ISLAND 307 Indian work (p. 384, fig. 369). The form figured is somewhat more densely branched and the thallus slightly coarser (500—600 .), but otherwise they seem to agree so well with each other that I am inclined to regard them as forms of the same species. ‘ HI. Esperi differs from HH. pannosa J. Ag., collected on the Pacific coast of Mexico (San Augustin, LIEBMANN) by its more slender and less ramose thallus, HZ. pannosa being described as »densissime intricato decomposito- ramosissima>». Area of distribution: Brazil, Chile, Australia, Easter Island. Fam. Rhodymeniaceae. Chrysymenia J. Ag. C. Skottsbergii nov. spec. — Figs. 49, 50. E callo radicali expanso frondes erectae, plerumque I—2 cm altae, oriuntur, caespitem densum et expansum formantes. Caules crassius filiformes, in parte basali robustiores (ca. 1 mm lati) su- perne gradatim tenuiores et ramosi, 5—6 vel plures ramos vesicaeformes gerentes. Vesiculae subglobosae-pyriformes, 4—5 mm latae et 5—6 mm longae. Fig. 49. Chrysymenia Skottsbergii nov. spec., parts of specimens. C. 2,5/r. This fine plant forms low, ca. r—2 cm high, dense tufts on Lithothamnion to which it is fastened by means a large irregular disc. From this the erect main filaments (stems) arise. These are mostly undivided in their lower half, irregularly ramified to all sides in the upper, the branches and, if these are branched again, the branchlets all ending in a vesicle. In this way each main branch may carry up to ten or even more vesicles. The stem is thickest, c. 1 mm, in the basal part, tapering gradually upwards. The vesicles are spherical-pyriform, about 4—5 mm broad and 5—6 mm long. The stem is solid. A transverse section (Fig. 50 a) shows a cortical layer consisting of quite smal roundish thick-walled cells arranged in more or less 308 F, BORGESEN distinct rows; inwards the cells gradually increase in size forming an even transi- tion to the medullary tissue. On the cross section this is composed of roundish cells. A longitudinal section (Fig. 50 b) of the stem shows that the cells of the cortical layer are of nearly the same shape: roundish to oval. The cells of the medullary tissue on the other hand are long, subcylindrical with oblique cross walls. Between the long cells short ones, often arranged in rows, are interposed here and there. Fig. 50. Chrysymenia Skottsbergii nov. spec. a cross, b length section of stipes; ¢ cross section through wall of vesicle; d inner surface of wall of vesicle with a cluster of glands; e two- celled! glands All c. 2°°/1. A cross section of the vesicles (Fig. 50 c) shows that the wall consists of a single layer (rarely two) of large rounded-rectangular cells, which are covered by a cortical layer of small cells. Opposite the centre of the large cells one or two cortical layers are present, while opposite the cross walls separating the large cells, where more space is left, we find this space occupied by a cell larger than the cortical ones. Seen from above the large cells in the wall of the bladder are polygonal with rounded corners (Fig. 50d). The cavity is filled with mucilage. On the inner side of the wall, facing the cavity, we find the glands characteristic of Chrysymenta, These occur in groups (Fig. 50d), but not in great number. MARINE ALGAE FROM EASTER ISLAND 3°9 On the other hand the glands themselves are rather large and each group mostly consists of a good number of glands (10—20 or more). They are gathered upon a smaller cell situated among the large wall cells; sometimes, too, this small cell carries another smaller one upon which the glands are borne. The glands are oblong, about 20 y. broad and 40 yu, long. In a few cases I have found the glands divided into two cells (Fig. 50 e). All the material was sterile. Of the known species of Chrysymenia, the present one is most closely related to C. Uvaria (L.) J. Ag. (esp. the Mediterranean form), pyrzformis Borgs. and mzcrophysa Hauck. It differs from C. Uvaria in growing less high, by the much smaller number of vesicles on each main stem, by the larger, shortly pedicellate vesicles, further by the gland-cells being larger and arranged in dense groups, while in C. Uvarza these cells are small, rounded and scattered singly over the wall (comp. my fig. 388 in Mar. Algae D. W.L, vol. II, p. 403). From C. pyriformis it also differs in the lower, more condensed habit, in the irregular branching and shortly pedicellate vesicles, while, in the former, generally also in C. Uvaria, the sessile vesicles form a raceme along the main stem and branches. C. pyriformis has larger and more pyriform vesicles. The cortical layer in the vesicles is thicker and more firmly built in C. Skottsbergzt. The glands are grouped in the same manner in both, but they are larger in the latter and more numerous to each group. The quite robust stem is also more strongly built in the new species. It should be remembered, that this was collected in tide-pools, while C. pyrzformzs is a sublitoral species. For more details I refer to my description of the latter. Finally, C. mzcrophysa is a much smaller plant, mostly with a single terminal vesicle, less often with one or two lateral branches in addition. For a detailed description see P. KUCKUCK, Beitr. zur Kenntnis der Meeresalgen. 13 (Wiss. Meeresuntersuchungen, N. F. Bd. V, Abt. Helgoland, 1912). Area of distribution: Endemic. Champia Desv. Ck. sp. In a glass tube with various algae a small fragment of a Champza was found. Had it come from the West Indies I would have named it Cz. parvula without hesitation, but as this species has not been reported from the Pacific with certainty, I prefer to leave it unnamed, the more as it is sterile, the arrangement of the tetrasporangia remaining unknown. Ae Lap MEU A RRO rE Er hag | tia: Keren ipsa) (ye Bey vbhe® Vide Oa e ality SBS Meet it iolligay rae CMI eM iy ed ae niente 2 a ee agp bs Sib ttireenallld Wahi ew peas & wed VG By Pa i ey AE Rei he Use Se IAD HF le Pale nih yy OF fine Seva Ay Or or Sigal io! ae aoe | Nieiae 14) » Le adel, (ERO Le i RIAGs wee Ceo Sarat Gee serie! and ; é tsa Sik PUT Hee iy ay i é aya ayes, Vi ' 27 7 (es ey: ae Lh es Pe 3 Pr) fwd j “~ 4 P viaevie: & ’ { eHih ; 1ef te Oc Ti hal (eat es ‘a P Sey ann i viLiein I ve eill hae i” ‘i: is i the lp | raps sear i + k i p i + sat) ' = hi { =s j é OR ESL SIPS IGE | gst Faded ; i { i U 4% oii) uy; t age ia? i¥ ‘ 17 ' ; r\t U / iy Ay k $ \ Sehehs ri i j } 34 i i { i pa! t 4 Ht ee (é { j ane ; ay i j Pyrenulaceae Astrotheliaceae Sphaerophoraceae Arthoniaceae Graphidaceae Chiodectonaceae Dirinaceae Lecanactidaceae Byssocaulon . ‘Thelotremaceae Diploschistaceae Gyalectaceae Coenogoniaceae Collemaceae . Pannariaceae Stictaceae . Peltigeraceae Lecideaceae . Phyllopsoraceae Cladoniaceae Acarosporaceae Pertusariaceae Lecanoraceae Parmeliaceae Caloplacaceae Theloschistaceae . | 14 Arten _ . Se N rl Mm Num Hi DOWD CONW NN NY He BR H&B HF HO ex OWWM » DIE FLECHTEN DER JUAN FERNANDEZ-INSELN Ba | Dueiiaecteiiets .: wonlh woqeanls,. Bares Ariens arr eee ‘ 8 Arten MTOR ACSF.) 1 501.) cali ehe can aa) re en Fe » ELCINEMOICIENES »4) (ais vo) beudy ee steed coe Nach der systematischen Zugehorigkeit des Algenkomponenten ergibt sich folgende Gruppierung: a.) Flechten mit Pleurococcustypus MRD 4 . 1og Arten, rund 60% b.) Flechten mit blauen Gonidien, beziehungsweise deren phylogenetische Deszendenten. .......... . 48 Arten, rund 25% c.) Flechten mit Trentepohliagonidien . . ..... . . . 28 Arten, rund 15 % Ein Vergleich dieser Zahlen mit denjenigen, welche ich ftir die Samoa- inseln gab’, zeigt uns, dass die perzentuelle Beteiligung der Cyanophili fast dieselbe ist, dass hingegen die Arten mit Trentepohliagonidien den Archi- lichenen gegeniiber stark abfallen. Ich glaube diesen Umstand damit begriin- den zu konnen, dass die Samoainseln ihrer geographischen Lage nach den tropischem Gebiete angehdren, dessen Klima das Auftreten von Flechten mit Trentepohliagonidien begiinstigt. Bemerkenswert bleibt der hohe Perzentsatz der Cyanophili, welcher eine Eigentiimlichkeit vulkanischer Inzeln zu sein scheint. An der Zusammensetzung der Flechtenflora von Juan Fernandez nehmen Teil Elemente des chilenischen Festlandes und der subantarktischen Region Amerikas, hinzu kommen dann Arten, welche im Tropicum oder Subtropicum weiter verbreitet sind, und auch Elemente der kalteren und gemiassigten Gebiete, welche namentlich in den hdheren Bergen Masafueras auftreten. Verhaltnis- massig gross ist die Zahl der Endemismen, doch ist dabei zu bedenken, dass bei einer besseren lichenologischen Erforschung der Nachbargebiete, insbe- sondere des chilenischen Festlandes, flr manche der bisher endemisch ange- sehenen Arten eine weitere geographische Verbreitung festgestellt sein wird. Die ftir Juan Fernandez bisher noch nicht verzeichneten Flechten sind durch ein Sternchen (*) kenntlich gemacht. Literatur. 1. Crompste, I. C., The Lichens of the »Challenger» Expedition. — Juan Fernan- dez. (Journ. Linn. Soc. London, Bot., vol. XVI, 1877, p. 223—224.) 2. Gay, C., Historia fisica y politica de Chile. (Botan., vol. VIII, 1852, Lichenes Pp. 53—228.) 3. Hue, A. M., Lichenes extra-europaei a pluribus collectoribus ad Museum Pa- risiense missi. (Nouv. Archiv. du Muséum, ser. 3, vol. X, 1898, p. 213— 2502°Ser.4, VOL, 18905. p:. 27-—220, taba — Wis ser. 4, vol. Ile roca mp 49—122, tab. I—V et ser. 4, vol. III, 1901, p. 21—108, tab. I—VI.) 4. —-—, Lichenes morphologice et anatomice descripsit. (Nouv. Archiv. du Mu- seu, ser: 4, ‘vol: (MII, 1907) p:°237 2725 ser.) 4,7 vol. XK, 1908p. 169— 2242 Ser. 5, VOlul a 1909,' ps (Ini 006 ¥ ser: is) vols I jroio, p17 ebivet ser. 5, vol: IV; *xqr2)ap. 1—§2.) 1 Die Flechten der Samoa-Inseln. (Denkschrift. Akad. Wiss. Wien, math.-naturw. Klasse, Band CXXXI, 1907, p. 2.) 318 A. ZAHLBRUCKNER Hur, A. M., Lichenum generis Crocynia Mass. plerasque species jutxa arche- typa specierum morphologice et anatomice descripsit. (Mémoir. Soc. Nation. Scienc. Nat. Cherbourg, vol. XXXVII, 1909, p. 223—254.) 6. Jonow, F., Estudios sobre la Flora de las Islas de Juan Fernandez. (Santjago de Chile, 1896, Lichenes p. 196—203.) Montacne, C., Prodromus Florae Fernandezianae, sistens enumerationem plan- tarum cellularum, quas in insula Juan Fernandez a Cl. Bertero collectas describi edique curavit. (Annal. Scienc. Nat., Botan., ser. 2, vol. IV, 1835. Lichenes p. 86—94.) 8. Miurrer, I. (Arc.), Lichenologische Beitrige. XXVII. (Flora, vol. CXXI, 1888.) Nyianper, W., Synopsis methodica Lichenum (Parisiis. Vol. I, 1859—60, vol. II, 1863.) 10. Report of the Scientific Results of the Voyage of H. M.S. Challenger. Botan. vol; I'(1885, Lichenes p. 19—21.) 11. ‘laytor, Tu., New Lichens, principally from the Herbarium of Sir William I. Hooker. (London Journ. of Botan., vol. VI, 1847, p. 148—197.) 12. ZAHLBRUCKNER, A., Flechten in »Botanische Ergebnisse der Schwedischen Ex- pedition nach Patagonien und dem Feuerland 1917—1919>. (Kgl. Svensk. Vetensk.-Akad. Handl., Vol. LVII, ser. 6, 1917, p. 1—81.) Verrucariaceae. Verrucaria (Wigg.) A. Zahlbr. *V. microspora Nyl. Nvt:'in Annal. ‘Scienc, Nat:,* Bot., “ser. 4, vol) Ill,/ 1855, p. 175 "et Bxpor Synopt. Pyrenocarp., 1858, p. 29; A. ZautBr., Catal. Lich. Univers., vol. I, 1921, Dpi.O7 Das Lager der vorliegenden Stiicke ist sehr dunkel, zumeist braunlich- schwarz; die Apothezien sind zahlreich und dicht gedrangt; die Sporen 9—TII p. lang und 6—7 uw. breit. Masatierra: Bahia del Padre, auf Strandfelsen (C. und I. SKOTTSBERG). Microglaena Korb. *M. fernandeziana A. Zahlbr. nov. spec. Thallus crustaceus, uniformis, maculas sat parvas, usque 12 mm latas formans, substratum obducens, tenuis, vix 0,2 mm crassus, sordidulenti-alutaceus, opacus, KHO et Ca Cl,O, non tinctus, minute areolatus, areolis ad 0,2 mm latis, subangulosis, planis, fissuris tenuissimis separatis, in margine linea obscuriore non cinctus, sorediis et isidiis non praeditus, superne strato cor- ticali, ex hyphis intricatis formato obductus; stratum gonidiale sat angustum, gonidiis cystococcoideis, globosis, glomeratis, laete viridibus, usque 10 y. latis; medulla alba, I—. Apothecia dispersa, sessilia, semigloboso-obconica, ad 0,5 mm lata et totidem alta, thallo extus concoloria, vertice nigro, leviter deplanato vel modice DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 31g convexo, poro tenui pertusa; excipulum subglobosum, angustum, pallidum, molle, ex hyphis tangentialibus et conglutinatis formatum, involucrello non cinctum, extus a thallo vestitum; hymenium decolor, nec inspersum nec guttulatum, hymenogonidiis nullis, I lutescens; paraphyses persistentes, capillares, ramosae et subintricatae, eseptatae, ad apicem non latiores; asci convergentes, oblongi vel oblongo-clavati, ad apicem rotundati et ibidem membrana paulum crassiore cincti, 8 spori; sperae in ascis bi- vel subuniseriales, verticales vel obliquae, primum decolores, mox autem dilute olivaceae vel olivaceo-fumosae, denum fuscae, ovales vel ovali-ellipsoideae, crebre murali-divisae, cellulis minutis, sub- cubicis, 34—38 yu. longae et 13—16 yp. latae, halone non circumdatae, I subau- rantiaco-lutescentes. Pyenoconidia non visa. Masatierra: Bahia del Padre, auf Strandklippen (C. und I. SKOTTSBERG). Santa Clara: Morro de los Alelies, auf hartem Gestein (C. und I. SKOTTSBERG). Dermatocarpaceae. Normandina (Nyl.) Wain. *N. pulchella Ny]. Nyt. in Annal. Scienc. Nat., Bot., scr. 4, vol. XV, 1861, p. 382; A. ZAHLBR., Catal. Lich. Univers., vol. I, 1921, p. 198. — Verrucaria pulchella Borr. apud Hook. et SoweErB., Suppl. Engl. Bot., vol. I, 1831, tab. 2602, fig. 1. Masatierra: Portezuelo, 400—500 m, iiber Flechten und Moosen, steril (C. und I. SKOTTSBERG). Pyrenulaceae. Arthopyrenia Mass. Conspectus specierum: A. Sporae persistenter biloculares. a. Loculi sporarum distincte inaequales A. adnexa var. leptosperma. b. Loculi sporarum aequales A. cinchonae. B. Sporae demum 4 loculares, ad latera lyrato-incisae A. planorots. *A. cinchonae Miill. Arg. Mitty. Arc. in Flora, vol. LXVI, 1883, p. 287 et in Mémoir. Suc. Phys. et Hist. Nat. Geneve, vol. XXX, no. 3, 1888, p. 26; WaiIn., Etud. Lich. Brésil, vol. II, 1890, p.. 233; A. Zauter., Catal. Lich. Univers., vol. 1,.1921;'p, 302..— ,Verru- carta cinchonae Ach., Synops. Lich., 1814, p. go. Masatierra: Cordon Salsipuedes, bei 615 m, auf Dendroserzs-Zweigen (C. und I. SKOTTSBERG). A. ZAHLBRUCKNER ie) N (@) #4. adnexa Mill. Arg. Muir. Arc. in Flora, vol. LXVI, 18383, p. 304; A. ZAHLBR., Catal. Lich. Univers., vol. I, 1921, p. 312. var. leptosperma A. Zahlbr. nov. var. Sporae angustiores ut in typo brasiliensi, 20—22 », longae et 7—9 y. latae. Masafuera: Quebrada de la Loberia, bei 300 m, auf alter Lumaborke (C. und I. SKOTTSBERG). A. planorbis Mill. Arg. Mtr. Arc. in Mémoir. Soc.. Phys. et Hist. Nat. Genéve, -voll. XXX, no. 3, 1888, p. 27; A: Zanusr., Catal. Lich. Univers., vol. I. 1921, Pp.) 3\O.ssmhertee caria planorbis Ach., Synops. Lich., 1814, p. 92; Mont. in Annal. Scienc. Nat., Bot. Ser. 2, vol: 1V, “1835, (p: 193. Masatierra: Auf glatten Rinden in schattigen Bergwaldern (BERTERO no. 1618). Coccotrema Mill. Arg. C. Cucurbitula Mull. Arg. Mix. Arc. in Mission Scientif. Cap Horn, vol. V, 1889, p.171;-A. ZAHLBR., Catal. Lich. Univers., vol. I, 1922, p. 359. — fertusaria Cucurbitula Mont. apud Gay, Hist. Fisic. y Polite Chile, Botan., vol. VIII, 1852, p. 200. var. laevigata A. Zahlbr. A. Zautsr., Catal. Lich. Univers., vol. I, 1922, p. 360. — Fertusarza Cucur- bitula var. luevigata Nyl. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 160. Auf glatten Rinden, ohne nahere Standortsangabe (Herbar. Paris). Porina Mill. Arg. Conspectus specierum: A. Apothecia a thallo libera a. Apothecia nigra, 0,2—0,3 mm lata P. fernandeztana. b. Apothecia rufofusca, vix 0,1 mm lata P. rufocarpella. B. Apothecia a thallo crasse obducta P. depressula. *P. fernandeziana A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, substrato arcte adnatus, effusus, tenuis, ad 0,1 mm crassus, bene limitatus, sed linea obscuriore non cinctus, versus partem marginalem thalli tamen hinc inde lineis obscuris parce decus- satus, cervinus, fere opacus, KHO —, Ca Cl,0, —, continuis et inaequalis, passim, DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 321 sed rarius, valde tenuiter subareolato-rimosus, areolis minutis, sorediis et isidiis nullis; fere homoeomericus, hyphae thalli latiusculae, toruloso-inaequales et septatae, leptodermaticae, subintricatae; gonidia chroolepoidea, cellulis breviter concatenatis, rotundatis, 7—12 y. latis, contentu viridiluteo, membrana _ tenui cinctis. Apothecia sat crebra, dispersa, rarius approximata, parva, 0.z—0O,3 mm lata, nigra, opaca, madefacta nitidula, a thallo non vestita, plus minus convexa, semiglobosa vel etiam paulum deplanata, poro demum valde tenui pertusa; excipulum angustum, pallidum, fusco-lutescens, ex hyphis tenuissimis, tangen- tialibus et dense conglutinatis formatum, 35—50 . crassum, involucrello fuli- gineo, semigloboso, ad basin truncato, usque ad basin excipuli producto et ibidem passim parum inflexo, arcte adhaerente obductum; nucleus conico- globosus, ad basin plus minus deplanatus; hymenium decolor, purum, I lu- tescens, imprimis asci; paraphyses capillares, strictiusculae, simplices, eseptatae, ad apicem non crassiores; asci anguste rapiformes, versus apicem angustata, in ipso vertice utplurimum retusuli, recti vel leviter curvuli, 8 spori, 80—85 u. longi et 8—9g y. lati; sporae in ascis biseriales, decolores, subcylindrico-fusifor- mes, ad spices rotundatae, rectae, 4 loculares, ad septa valde tenuia non con- strictae, membrana tenui cinctae, 22—28 yp, longae et + 3.5 wv. latae. Conceptacula pycnoconidiorum minuta, vix 0,1 mm lata, nigra, convexa; perifulerium fuligineum, dimidiatum; fulcra exobasidialia; pycnoconidia oblonga, parva 3,5—4 w longa et ad I ». lata. Masatierra: Cordén Salsipuedes, bei 465 m, auf vulkanischem Gestein (C. und I. SKOTTSBERG). Gehort in den Formenkreis der P. chlorotica (Ach.) Mill. Arg. und unter- scheidet sich von den iibrigen Arten desselben durch das dickliche, glatte Lager, gréssere Apothezien und schmalere Sporen. *P, rufocarpella A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, tenuissimus, effusus, cervinus vel umbrino-cervinus, opacus, KHO —, Ca Cl,O, —, continuus, laevigatus, in mar- gine linea obscuriore non limitatus, soredils et isidiis destitutus, fere homoeo- mericus, gonidiis chroolepoideis, cellulis concatenatis, rotundatis vel subirre- gularibus, 8—toy, latis, contentu viridi-lutescente, membrana crassiuscula cinctis. Apothecia dispersa, exigua, vix 0,1 mm lata, convexa vel fere semiglobosa, rufescentia, opaca, poro haud conspicuo, a thallo omnino libera; excipulum integrum, globosum, angustum et decolor, involucrello sub lente subaurantiaco, KHO rufescente, filamentoso-subcelluloso, molli, dimidiato, excipulo adhaerenti praeditum; hymenium decolor, purum, I lutescens; paraphyses capillares, li- berae, simplices, eseptatae, ad apicem non latiores; asci fusiformes, recti vel curvuli, ad apicem angustati et in ipso apice rotundato-retusi, 75— 78 p. longi et 8—10 », lati, 8 spori; sporae in ascis biseriales, decolores, oblongo-fusiformes vel oblongo-dactyloideae, rectae, triseptatae, cellulis aequalibus, septis tenuibus, membrana tenui cinctae, 19—22 p. longae et + 3,5 wu. latae. Pycnoconidia non visa. Masatierra: Cordén Salsipuedes, 465 m, auf einem etwas iiberhangenden vulkanischen Block (C. und I. SKOTTSBERG). 2I—2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 322 A. ZAHLBRUCKNER Habituell der vorigen, mit ihr an demselben Standort lebenden Art sehr ahnlich, aber mit ihr durch die kleinen, nicht schwarzen Apothezien nicht zu verwechseln. Ahnliche Convergenzerscheinungen lassen sich auch bei anderen Gattungen (z. B. Buellia) auf Juan Fernandez beobachten. *P, depressula A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, subtartareus, tenuis, O,12—0,18 p. crassus, argillaceo-cinerascens, madefactus argillaceus, opacus, KHO sanguineo- sordidescens, Ca Cl,O, non tinctus, effusus, subcontinuus vel irregulariter, mi- nute et tenuiter rimulosus, parum inaequalis, sorediis et isidiis nullis prae- ditus, linea obscuriore non cinctus, superne strato corticali tenui, subdecolore, ex hyphis intricatis formato obductus, caeterum fere homoeomericus; gonidia chroolepoidea, crebra. Apothecia verruciformia, dispersa vel approximata, sessilia, placentiformia, ad basin bene constricta, ad verticem deplanata, sessilia, thallo concolora, usque I mm lata, a thallo crassiuscule cincta (ibidem KHO sanguinea), in parte verticali excipuli partem apicalem e plano convexam, sordide fuscescentem et + pruinosulam, madefactam rufam vel rufo-nigricantem, usque 0,5 mm latam denudentia; excipulum integrum, inferne paulum angustius, rufescenti-ochraceum vel rufescenti-lutescens, KHO rufescens, circa ostiolun plus minus nigricans, ex hyphis tangentialibus, tenuibus et dense contextis formatum, globosum, superne in collum breve, sat abrupte angustatatum abiens; hymenium decolor, purum, I vix lutescens; paraphyses capillares, densae, strictae, liberae, simplices, escep- tatae, ad apicem non latiores; periphyses filiformes, tenuies, densae, strictae, collum intus obducentes; asci convergentes, oblongo- vel subcylindrico clavati, rectiusculi, ad apicem rotundati et membrana passim bene incressata cincti, 8 spori; sporae in ascis biseriales, rarius subuniseriales, decolores, fusiformes vel dactyloideo-fusiformes, utrinque acutatae, rectae vel subrectae, 10—12 locu- lares, septis tenuibus et membrana tenui, 52—64 y. longae et 7,5—9 yp. latae. Pycnoconidia ignota. Masatierra: Quebrada Monte Maderugo, 390 m, auf einer Felswand (C. et I. SKOTTSBERG). Masafuera: Quebrada de las Casas, auf einer feuchten, schattigen Fels- wand (C. und I. SKOTTSBERG). Was die dussere Erscheinung betrifft, so herrscht zwischen der obigen Art und der in Brasilien vorkommenden /orina exserta Mill. Arg. eine ge- wisse Ubereinstimmung, es ist mir wahrscheinlich, dass beide Arten in verwandt- schaftlicher Beziehung stehen. Indes ist die Gestaltung der Fruchtwarzen und der Sporen nicht in Einklang zu bringen. Pyrenula Ach. Conspectus specierum: A. Apothecia extus a thallo obducta et tautum vertice libera P. aspistea. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 323 B: Apothecia nuda, nigra a. Sporae 16—18 x 6—8 p P. mamillana. b. Sporae 19—23 X 8—II yp. P. Kunthi. P, aspistea Ach. Acu. in Gesellsch. Naturf. Freund. Berlin. Magazin, vol. VI, 1814, p. 17, tab. Il, fig. 8; A. Zantsr., Catal. Lich. Univers., vol. I, 1922, p. 423. — Verrucaria aspistea Afz. apud Ach., Method. Lich., 1803, p. 121; Mont. in Annal. Scienc. Nat. bot., ser. 2; vol. IV, 1835, p. 93: Masatierra: Auf Baumrinden, ohne nahere Standortsangabe (BERTERO no. 411); Cordon Salsipuedes, auf Drzmys-Zweigen (C. und I. SKOTTSBERG). *P, mamillana Trevis. Trevis., Conspect. Verruc., 1860, p. 13; A. ZAHLBR., Catal. Lich. Univers., vol. I, 1922, p. 436. — Verrucaria mamillana Ach., Method. Lich., 1803, p. 279. Masatierra: Cordon Chifladores, 350 m, auf Zweigen der Rodzusonza thu- rifera; am Fusse eines Abhanges von El Yunque, bei 400 m, auf Myrceugenia- Zweigen; Portezuelo, bei 600 m, auf Stammen der Rodznsonta thurifera; Que- brada seca, 435 m, auf Dendroser7s‘Rinde; Cordén Salsipuedes, 615 m, auf Dendroseris-Rinde (C. et I. SKOTTSBERG). =P, Kunthii Fée. Fé, Suppl. Essai Crypt. Ecorc. Offic., 1837, p. 80, tab. XLI, fig. 18; A. ZAHLBR., Catal. Lich. Univers., vol. I, 1922, p. 434. — Verrucaria Kunthit Fée, Essai Crypt. Ecorc. Offic., 1824, p. 88, tab. XXXIV, fig. 4. Masatierra: Cordén Chifladores, auf A/yrceugenia-Rinde; Portezuelo, beim SELKIRK-Denkmal 590 m, auf Berberzs-Zweigen (C. et I. SKOTTSBERG). Masafuera: Quebrada del Mono, auf der Rinde der Myrceugenta Schulzet (C. et I. SKOTTSBERG). Astrotheliaceae. Pyrenastrum Eschw. P. chilense Mont. Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 2045 Jonow, Estud. Flor. Juan Fernand., 1896, p. 201. — Pyrenastrum americanum Mont. in Annal. Scienc. Nat., ser. 2, vol. IV, 1835, p. 94 (non Sprgl.). Masatierra: auf Fagara-Rinde, ohne nahere Standortsangabe (BERTERO no. 1622). 324 A. ZAHLBRUCKNER Sphaerophoraceae. Sphaerophorus Pers. S. melanocarpus DC. DC.. apud Lam. et DC., Flor. Frang., edit. 2, vol.- V1, 1805,.p.,ayeceee ZAHLBR., Catal. Lich. Univers., vol. I, 1922, p. 693. — Lzchen melanocarpus Sw., Nova Gener. et Spec. Plant., 1788, p. 147. — Sphaerophoron comspressum Ach., Method. Lich., 1803, p. 135. Masatierra: Auf Baumstammen, ohne nahere Standortsangabe (BERTERO), auf dem Riicken zwischen Quebrada Piedra agujereada und Quebrada Laura, 625 m, am Fusse eines Drimys Stammes; Portezuelo, bei 500 m, auf Baum- stammen; Felskamm unw. Tres Puntas, 300-350 m (C. et I. SKOTTSBERG). Masafuera: Las Torres, bei 1,200 m; Los Inocentes, siiddstliche Ab- hange, 950 m, im Drzmys-Wald, am Fuss der Baumstamme (C. et I. SkorTs- BERG). Arthoniaceae. Arthonia Ach. Conspectus specierum: A. Apothecia lirellina, elongata. a. Sporae biloculares; loculi sporarum subaequales A. cytist var. meridionalts. b. Sporae triloculares; loculorum apicalis loculis reliquis latior et major I. Sporae 11—15 p. longae et 4—6 », latae A. subnebulosa. II. Sporae 15—17 y. longae et 5—6 y, latae A. berberina. B. Apotheria rotundata vel angulosa; sporae 4-—6 loculares, loculus spo- rarum supremus maximus A. complanata. *A. cytisi Mass. — Taf. 24, Fig. 1. Mass. in Nuov. Annal. Sc. Nat. Bologna, vol. VII, 1835, p. 216; A. ZAHLBR., Catal. Lich. Univ., vol. II, 1922, p. 36. — Arthonia excipienda Nyl., Lich. Scandin, LOO tp. 201. var meridionalis A. Zahlbr. nov. var. — Taf. 24, Fig. 3—4. Lirellae tenuiores et praesertim magis ramosae et plus minus radiatae. Caeterum cum planta europaea congruens. Masatierra: Portezuelo de Villagra, 590 m, auf den Zweigen der Ber- beris corymbosa (C. et I. SKOTTSBERG). *A. subnebulosa A. Zahlbr. nov. spec. Thallus epiphloeodes, tenuissimus, substratum arcte obducens, effusus, in margine late ochraceo lutescens et nitidulus, in parte centrali magis albescens, DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 325 subnitidus vel fere opacus, KHO —, Ca Cl,0, —, laevigatus, continuus, in ambitu linea tenui et nigricante cinctus, molliusculus, sorediis et isidiis desti- tutus, homoeomericus; gonidia chroolepoidea, cellulis concatenatis, plus minus rotundatis, 6-—10 y. latis, contentu dilute viridi-lutescente impletis. Apothecia sessilia, depressa, maculiformisubstellata, ramis brevibus, ut- plurimum confluentibus, I—1,3 mm latis et paulum longioribus, obscure fusca, epruinosa, emarginata, demum elabentia; hymenium angustum, superne rufo- fuscum et KHO paulum in olivaceum vergens, caeterum decolor et purum, | cupreum; hypothecium decolor; paraphyses tenues, eseptatae, intricatae; asci pyriformi-subglobosae, ad basin abrupte cuneati, 28—30 p, longi et 18—2o0 up, lati, 8 spori; sporae in ascis 3—4 seriales, ex incolore mox fuscidulae et demum fusco-obscuratae, solaeformes, rectae vel curvulae, triloculares, loculus apicalis reliquis latior et major, membrana tenui cinctae, II—1I5 y. longae et 4—6 u, latae. Pycnoconidia non visa. Masatierra: Quebrada Gutierrez, auf Fwanza-Rinde (C. et I. SKOTTS- BERG). Die neue Art geh6drt zu jenen Arthonien, welche konstant dreizellige Spo- ren besitzen. Die Sporen der bisher bekannten Arten dieses Formenkreises sind durchwegs grosser. Benachbarte Arten diirften Avthonia ephelodes Nyl. und Arthonia scitula Krph. sein. *A. berberina A. Zahlbr. nov. spec. — Taf. 24, Fig. 2. Thallus epiphloeodes, tenuis, uniformis, flavido-albescens, nitidulus et laevi- gatus, circa apothecia albidior, opacus et leviter verruculoso-inaequalis, KHO —, Ca Cl,0, —, effusus, continuus, sorediis et isidiis destitutus, in margine passim linea nigricante et tenui cinctus; gonidia chroolepoidea. Apothecia sessilia, immarginata, fusco-nigricantia, epruinosa, madefacte plus minus in sanguineo-fuscum vergentia, irregulariter stellato-radiata, ramis sat brevibus, angustis, ad apicem rotundatis vel retusatis, convexiusculis, usque I mm longis; hymenium superne rufo-fuscum, KHO nigricans, non inspersum, caeterum decolor, purum, 70—80 ». altum, I e coeruleo cupreo-sordidescens; hypothecium tenue, pallidum, flavescens, molle, ex hyphis intricatis formatum; paraphyses tenuies, ramosae et intricatae, parum distinctae, eseptatae; asci late pyriformes, ovali- vel subgloboso-clavati, ad apicem rotundati et membrana crassa cincti, 40 —48 w. longi et 16 —18 », lati, 8 spori; sporae in ascis triseriales, decolores, ellipsoideae vel oblongo-subcuneatae, utrinque rotundatae, rectae, triseptatae, loculus apicalis reliquis multum major et paulum latior, septis te- nuibus, ad septa non constrictae, 15—17 uw. longae et 5—6 uy, latae, I cupreae. Conceptacula pycnoconidiorum minuta, punctiformia, vertice nigro, vix prominulo; perifulcrium dimidiatum, sub lente obscure fuscum; fulcra exo- basidialia; pycnoconidia bacillaria, recta vel subrecta, 3,5—4 yp. longa at ad 0,6 p. lata. Masatierra: Portezuelo, beim SELRIRK-Denkmal, 590 m, auf Aerberzs- Zweiglein (C. et I. SKOTTSBERG). Die Art gehort in den Artenkreis der Avthonia varia (Ach.) Nyl. und 326 A. ZAHLBRUCKNER weicht durch die Gesamtheit der Merkmale von den wbrigen Gliedern der Gruppe ab. *A,. complanata Feée. Fir, Essai Crypt. Ecorc. Officin., 1824, p. 54; A. Zanrer., Catal. Lich. Uni- vers:, vol. dij«1922, 9p: 22 Masatierra: Cordén Salsipuedes, auf Drzmys-Zweigen (C. und I. SKOTTS- BERG). Graphidaceae. Graphis (Adans.) Ach. *G. intricata Fee. Fir, Essai Crypt. Ecorc. Officin., 1824, p. 42, tab. IX, fig. 3; A. ZaHLBr.’ Catal. Lich, Univ., vols-ll, 1923; 5p: 313: Masatierra: Quebrada seca, 435 m, auf /agara-Rinde (C. und I. SKOTTs- BERG). Die Stiicke von Juan Fernandez zeigen nur seltener sternformige Lirellen, diese sind zumeist einfach oder sparlich verzweigt, doch bildet FEE auch solche ab. Die Sporen fand ich 8—g-zellig, 36—4o p. lang und 8,5—9 u. breit. Das Excipulum ist kohlig, geschlossen, am Grunde abgestutzt und nur etwas dicker; die Lippen sind zugespitzt und neigen gegen einander. *G. Dumastii Sprgl. SPRGL., Syst. Veget., vol. IV, pars 1, 1827, p. 254;:A. Zautsr, Catal.- Lich. Univers., vol. II, 1923, p. 302. — Mssurina Dumastii Fée, Essai Crypt. Ecore. Offhein: 18240p5 XC, tabs al shieie7 etisp-45o, tab AScVa, (fig. 24: Masatierra: Quebrada seca, 435 m, auf der Rinde der Dendroseris micrantha (C. und I, SKOTTSBERG). Lirellae labiis thallinis integris, superne conniventibus, strato corticali an- gusto, I2—16 », crasso, decolore, subchondroideo, ex hyphis dense conglutinatis longitudinalibus formato obductae, gonidia pauca et medullam subcineream includentes; excipulum rudimentarium, anguste flabellatum, sordide fuscidulum, parti superiori hymenii adnatum; hypothecium valde angustum, decolor; hy- menium decolor, purum, superne non obscuratum, I vix lutescens; paraphyses strictae, filiformes, contextae, simplices, eseptatae; asci anguste clavati, 8 spori; sporae in ascis subuniseriales, decolores, late ellipsoideae, apicibus obtuse rotun- datis, rectae, 4 loculares, loculis lentiformibus, 21—24 ». longae et 10—I3 yp. latae, I pallide aurantiaco-cupreae. MONTAGNE gibt fiir Masatierra als auf Rinden lebend noch Graphis scripia (L.) Ach. an. Mir ist diese Flechte in der SKOTTSBERG’schen Aufsammlung DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 327 nicht aufgefallen. Ich halte ihr Vorkommen nicht fiir wahrscheinlich, obwohl sie auch fiir Chili angegeben wird. Die alteren Angaben dieser Flechte beziehen sich vielfach auf andere Arten der Gattung Grapjiis. Phaeographina Mill. Arg. P. scalpturata Mill. Arg. Mtr. Arc. in Flora, vol. LXV, 1882, p. 399 et in Mémoir. Soc. Phys. et Hist. Nat. Genéve, vol. XXIX, no. 8, 1887, p. 48; A. ZAHLBR. in Sitzungsber. Akad. Wiss. Wien, math.-naturw. Kl., vol. CXI, Abt. 1, 1g02, p. 388, tab. II, fig. 15—16. — Graphtis scalpturata Ach., Synops. Lich., 1814, p. 86; Monr. in Annal. science. Nat., Bot., ser. 2, vol. IV, 1835, p. 93. Masatierra: Auf Baumrinden, ohne nahere Standortsangabe (BERTERO no. 1612). Ich fand diese in den subtropischen und tropischen Regionen lebende Flechte in der SKOTTSBERG’schen Aufsammlung nicht. Die Angabe ware zu iiberpriifen. Chiodectonaceae. Enterostigma. E. Skottsbergii A. Zahlbr. A. ZAHLBR. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. g. Masafuera: auf dem Hochplateau, 1,100—1,200 m, auf Lava (C. et I. SKOTTSBERG). Dirinaceae. Dirina Fr. *D. limitata Ny]. Nyx. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 158. Masatierra: Bahia del Padre, auf Strandfelsen (C. et I. SKOTTSBERG). Lecanactidaceae. Schismatomma Fw. et Korb. *S. accedens A. Zahlbr. ts comb. Lecidea accedens Ny. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 163 et vol. XI, 1869, p. 225, not. — TZoninia accedens A. Zahlbr. in Beiheft. zum Bot. Centralbl., vol. XIX, 2. Abt., 1905, p. 77. 328 A. ZAHLBRUCKNER Thallus epilithicus, crassiusculus, bullatus, bullato-verrucosus vel subcere- brinus, verrucae ad ambitum thalli subdispersae, mox autem confluentes et thallum expansum formantes, rufescenti-fuscescens vel badius, rarius primum testaceo-fuscescens, fere opacus, KHO —, Ca Cl,O, —, in superficie laevigatus vel toruloso-subplicatus, subtus niger vel nigrescens, sorediis et isidiis nullis, hypothallo nigricante, ex hyphis formato laxiuscule retiformi-contextis, fuscis vel nigricantibus, usque 3,5 y. crassis, ramosis; superne et ad latere verrucarum distincte, subtus minus distincte corticatus; cortex ad ambitum rufo-fuscus, caeterum decolor, chondroideus, 44— 90 y. crassus, ex hyphis intricatis et dense contextis formatur, non inspersus; medulla crassa, solida, tartarea, alba vel dilute flavescenti-albida, KHO, Ca Cl,O, et KHO + CaCl,O, non tingitur, I lutescit, ex hyphis inspersis, intricatis, ad 2 yp. crassis, sat leptodermaticitis for- mata; stratum gonidiale infra corticem superiorem situm, continuum, gonidiis chroolepoideis, cellulis rotundatis vel irregularibus, increbre concatenatis, con- tentu pallide viridescente, membrana tenui cinctis, 8—16 yw, longis. Apothecia e verrucis erumpentia, mox alte sessilia vel subpodicellata, dispersa, lecanorina, rotunda vel rotundata, demum sinuoso-flexuosa, ad basin bene constricta, usque 25 mm lata; receptaculum extus laevigatum, corticatum, medullam sat amplam et gonidia includens; discus dense caesio-pruinosus, albus, opacus, KHO —, CaCl,0O, —, planus vel convexiusculus, demum (ut videtur morbose) pruina destitutus et niger; margo tenuis, integer vel subinteger, thallo concolor vel paulum_ pallidior, leviter prominulus, persistens; hypothecium obscure rufo-fuscum vel nigricans, ad latera sat tenue, infra hymenium crassiu- sculum, ex hyphis plus minus perpendicularibus, versus hymenium non bene limitatis et passim in cum penetrantibus formatum; hymenium superne crassius- cule pulverulento-inspersum, sordidum, caeterum fere decolor vel dilute fuscescens, spumoso-inspersum, 170—190 y, altum, I vinose cupreum vel pro parte aeru- ginoso-obscuratum; paraphyses filiformes, 1,;—1,8 p crassae, leviter flexuosae, ramosae et subintricatim connexae, eseptatae, ad apicem non latiores; asci hymenio subaequilongi, facile liberi, clavati, ad basin caudato angustati, ad apicem rotundati et membrana incrassata cincti, 8 spori; sporae in ascis biseriales, subdigitiformes, ellipsoideo-oblongae vel oblongo-fusiformes, utrinque rotundati, rectae, triseptatae, septis tenuibus, 20—25 py. longae et 5,5—8 wy. latae. Pycnoconidia non visa. Masatierra: Cordon Salsipuedes (C. und I. SKOTTSBERG). Ohne ein Urstiick der NYLANDER’schen Art gesehen zu haben, bin ich bei den augenfalligen habituellen Merkmalen dieser Flechte sicher, sie richtig gedeutet zu haben. NYLANDER stellte die Flechte zu den Zondnzen, in welcher Auffassung ich ihm friiher folgte, da ich keine Gelegenheit hatte, sie naher zu studieren. Ubrigens sagt NYLANDER selbst, dass er seine Beschreibung nach einem kimmerlichen Exemplar entwarf, sonst ware ihm der Gonidientypus sicherlich nicht fremd geblieben. Von den Paraphysen wird ferner gesagt, dass sie »graciles» waren, was die Frage, ob in einfach oder verzweigt, ein- fach, wie bei TZonzuza, oder verdstelt und mehr weniger netzartig-verbunden, wir bei Schzsmatomma seien, offen Jasst. Auch das bei den Schzsamatommen und Lecanactis-Arten so oft zu beobachtende leichte Loslésen der Schlauche von DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 329 den Paraphysen wird nicht erwahnt. Alle diese Merkmale aber sind fir die generische Einreihung der Flechte massgebend. Habituell dhnlich ist die ebenfalls auf Juan Fernandez vorkommende Toninia bullata, sie kann aber mit dieser Flechte wegen des glanzenden Thallus und der schwarzen Apothezien schon ausserlich nicht verwechselt wer- den; bei mikroskopischer Untersuchung treten die trennenden Merkmale noch mehr hervor. [(Chrysothricaceae. Chrysothrix noli tangere Mont. in Annal. Scienc. Nat., Bot., ser. 3, vol. XVIII, 1852, p. 313 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 213; Mass. in J. R. Istit. Veneto, ser. 3, vol. V, 1860, p. 499, tab. III; Senft in Ber. Deutsch. Bot. Gesellsch., vol. XXXIV, 1916, p. 592, tab. XVII. — Cilicia nol tangere Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. Il, 1834, p. 375, tab. XVI, ne 2 ét vol. IV, 1835; p: 94: Zweifelhaft. MontaGNne schreibt, Prodr., 1835 p. 94: Crescit ad ramulos ar- borum et in Usnea ceratina: Sterilis lecta. Specimina chilensia sola fructifera vidi. — BertERo coll. n. 176. — In Annal. 1834 werden nur Bertero’s Exemplare aus Chile erwihnt: diese sind fertil, die insularen sollen aber steril sein. Wahrscheinlich stammen aber auch die sterilen Stiicke aus Chile, jedenfalls die auf Uswea wach- senden, aber wohl auch die anderen. Néaheres bei Ramalina und Usnea.] Byssocaulon Mont. B. niveum Mont. — Taf. 25, Fig. 2. Mont. in Annal. Scienc. Nat., Bot, ser. 2, vol. III, 1835, p. 355; Born. in Annal. Scienc. Nat., Bot., ser. 5, vol. XVII, 1873, p. 60, tab. VII, fig. 5; BonnreR in Journ. de Bot., vol. I, 1887, p. 3, fig. 6; Nvyv., Lich. Japon., 1890. p. 59. — Parmelia gossypina var. filamentosa Mont. in Annal. Scienc. Nat., Bot., ser. 3, vol. XVI, 1851, p. 50 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 144; Jonow, Estud. Flor. Juan Fernandez, 1896, p. 201. — Syssocaulon filamen- tosum Ny\. in Mémoir. Soc. Scienc. Natur. Cherbourg, vol. V, 1857, p. 119. — Cro- cynia nivea Hue in Mémoir. Soc. Nation. Scienc. Natur. Cherbourg, vol. XXXVH, 1909, Pp. 244. Masatierra: auf Baumrinden in schattigen Waldern der hdchsten Berge (BERTERO); Cordén Central, 500 m, auf feuchten Felsen uber verwitterten Moosen (C. und I. SKOTTSBERG); Pangal, bei 600 m, tiber Moosen (C. und I. SKOTTSBERG); am Wege bei Portezuelo, auf Baumrinden (C. und I. SKOTTSBERG). Thelotremaceae. Ocellularia Sprgl. #Q. (sect. Ascédium) subdenticulata A. Zahlbr. nov. spec. — Taf. 24, Fig. 6. Thallus substratum obducens, tenuis, submembranaceus, albus, opacus, KHO —, Ca CI,O, —, continuus, plus minus aequalis, sorediis et isidiis destitutus, haud bene limitatus et linea obscuriore non cinctus. 330 A. ZAHLBRUCKNER Apothecia dispersa vel approximata, sessilia, depresso-subglobosa, ad basin distincte constricta, 0,6—O,8 mm lata et fere totidem alta; discus immersus, pallide fusco-rufescenticarneus, opacus, epruinosus, angustus et demum hymenio elapso dilatatus; receptaculum rufescenti-aJbidum, circa discum album, marginem annulatum, demum crenulatum vel subdenticulatum formans; excipulum duplex, interius decolor, angustum, ex hyphis tangentialibus, tenuibus et dense conglu- tinatis formatum, exterius sat crassum, obscure rufofuscum, KHO sanguineo- rufescens; hymenium decolor, purum, I lutescens, non columellatum; paraphyses capillares, densae, strictae, simplices, eseptatae, ad apicem non latiores; asci saccato-clavati, recti, ad apicem rotundati et membrana incrassata cincti, 4—8 spori, I cuprei; sporae in ascis 2—3 seriales, decolores, fusiformes, plus minus vermiculares, 16—20 loculares, loculis depresso-lentiformibus, centralibus re- liquis paulum latioribus, 50—80 yu, longae et 7—9 ». latae, I violascentes. Pycnoconidia non visa. Masafuera: Nordlicher Teil der hohen Kammes, auf bemoosten Dzck- sonza-Stammen (C. und I. SKOTTSBERG). Thelotrema (Ach.) Mill. Arg. *T. lepadinum Ach. Masatierra: Cordén Chifladores, auf Baumrinden (C. et I. SKOTTSBERG); Quebrada Gutierrez, auf Fuanza-Rinde (C. et I. SKOTTSBERG); Quebrada seca, 435 m, auf der Rinde von Coprvosma und Fagara (C. et I. SKOTTSBERG). Masafuera: Quebrada de la Loberia, im Walde auf Lama-Rinde, 800 m (Cz let J.’ SKOTTSBERG). Stimmt mit der europaischen Pflanze vollig tiberein. Auf diese Flechte diirfte sich jenes Zkelostrema beziehen, welche CROMBIE (Journ. Linn. Soc. London, Bot., vol. XVI, p. 223) ohne nahere Bestimmung der Art fiir Juan Fernandez angibt. Diploschistaceae. ~ Diphloschistes Norm. A. Apothecia pyrenocarpoidea, disco minuto et margine radiato-plicato; thallus KHO D. actinostomus. B. Apothecia lecanorina, disco lato; thallus KHO lutescens D. scruposus vat. arenarius. D. actinostomus A. Zahlbr. A. Zautpr. in Hedwigia, vol. XXXI, 1892, p. 34; A. L. Smiru, Monogr. Brit. Lich., vol. I, 1918, p. 384; Wain. in Acta Acad. Scient. Fennic., ser. A, vol. VI, no. 7, 1915, p. 132. — Urceolaria actinostoma Pers. apud Acu., Lichenogr. Univ., 1810, p. 288. — Verrucaria actinostoma Ach., Lichenogr. Univ., 1810, p. 288; Mon, in, Annal.. Scienc. Nat., Bot., sex. 2, vol. 1V, 1835, p. 93. Masatierra: Auf Felsen in den Bergen (BERTERO no. 374). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 331 D. scruposus Norm. Norm. in Nyt Magaz. for Naturv., vol. VII, 1853, p. 232. — Urceolaria scruposa Ach., Method. Lich., 1803, p. 147; Croms. in Journ. Linn. Soc. London, Bot., vol. XVI, 1877, p. 223; JoHow, Estud. Flor. Juan Fernandez, 1896, p. 200. Masatierra: ohne nahere Standortsangabe (MOSELEY). var. arenarius Mill. Arg. Mirr. Arc. in Hedwigia, vol. XXXIII, 1892, p. 196; Stnr. in Verhandl. zool.-bot. Gesellsch. Wien, vol. LXI, 1911, p..45. — Urceolaria scruposa var. arenaria Senaer., Lich. Helvet. Spicil:, sect. 2, 1826, p: 75. Masatierra: Puerto Frances, auf der Erde (C. und I. SKOTTSBERG); Cor- don Centinela, auf losem Tuff (C. und I. SKOTTSBERG); Strandgebiet zwischen der Kolonie und Pangal, auf der Erde (C. und I. SKOTTSBERG). Gyalectaceae. Microphiale (Stzbg.) A. Zahlbr. *M. lutea A. Zahlbr. A. Zautpr. in Sitzungsber. K. Akad. Wiss. Wien, math.-naturw. Classe, vol. CXI, 1. Abt., 1902, p. 393. — Lichen luteus Dicks., Fasc. Plant. Crypt. Brit., vol. I, Epos, p. 11, tab. IT, fig. 6. Receptaculum crassiuscule corticatum, cortice chondroideo, extus anguste ochraceo-flavescente, intus hyalino, ex hyphis radiantibus, tenuibus et intricatis formato, medullam et gonidia increbra, in parte inferiore sita includente; hypothecium lutescens, molle, ex hyphis intricatis formatur; hymenium superne anguste dilute sordidescens, caeterum decolor et purum, 70—8o ». altum, I e coeruleo aeruginoso-sordidulum; paraphyses strictae, filiformes, simplices, esep- tatae, ad apicem non latiores; asci oblongo-clavati, hymenio subaequilongi, ad apicem angustato rotundati et membrana vix crassiore praediti, 8 sport; sporae in ascis plus minus biseriales, decolores, fusiformi-oblongae, ad apices rotun- datae, rectae, uniseptatae, septo et membrana tenui, cellulis acqualibus, 1o—13 p. longae et 3,5—4 yw. latae. Masatierra: auf dem Riicken oberhalb Pangal, 400 m, auf Coprosma py- vifolium und Quebrada seca, 435 m, auf Coprosma triflorum (C. und I. SKOTTS- BERG). Masafuera: Quebrada del Mono, 475 m, auf Myrceugenia Schulzei (C. und I. SKOTTSBERG). Gyalecta Ach. *G, jenensis A. Zahlbr. nov. comb. Lichen jenensis Batsch, Elench. Fungor., Cent. prima, 1786, p. 219, fig. 153. — Lichen cupularis Hedw., Descript. et Adumbr. Muscor. Frondos., vol. H, 1789, 332 A. ZAHLBRUCKNER p- 58, tab. XX, fig. B. — Gyalecta cupularis Schaer., Lich. Helvet. Spicil., sect. 2, Lo20,0D» 7.0. Masatierra: Bahia del Padre, auf Strandklippen und auf der Steilwand siidlich von Tres Puntas, 380 m (C. und I. SKOTTSBERG). Pachyphiale Lonnr. P. cornea Poetsch et Schiederm. Portscu und Scureperm., Aufzahl. samenl. Pflanzen., 1872, p. 222. — Lzchen corneus With., A Botan. Arrang. Brit. Plants, vol. II, 1776, p. 712. — JLecidea car- neola Ach., Lichenogr. Univ., 1810, p. 194, tab. Il, fig. 7. — Szatora carneola Fr. in Kgl. .Vetensk.-Akad. Nya Handl., 1822, p. 267; Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p:,03 et -apud (Gay, Hist. Kisic. y Polit. @huvlemiagien vol. VIII, 1852, p. 174; JoHow, Estud. Flora Juan Fernandez, 1896, p. 201. Masatierra: auf Baumrinden (BERTERO no. 20). Unter den von SKOTTSBERG gesammelten Flechten fand ich diese Art nicht. ’ Coenogoniaceae, Coenogonium Ebrbg. *C. velutinum A. Zahlbr. nov. spec. Thallus glauco-cinerascens, opacus, substrato omnius adnotus, aequaliter velutino-pannosus, pulvinulos deplanatos, rotundatos, usque 4 cm latos et demum plus minus confluentes, caeterum bene limitatos formans; filamenta 7,5—9,5 p. crassa, septata, cellulis cylindricis, 18—25 p longis, ad septa non constrictis, membrana et septo mediocri cinctis, increbre ramosa, extus hyphis_ verti- calibus. Apothecia biatorina, sessilia, dispersa vel approximata, usque I,2 mm lata, ceraceo aurantiaca, epruinosa, e concaviusculo subplana vel vix convexa; margo disco pallidior, albidellus, tenuis, integer, persistens; excipulum integrum ad latera hymenii angustius quam infra hypothecium, ex hyphis radiantibus, dense conglutinatis et crebre septatis formatum, paraplectenchymaticum, I dilute coerulescens; hypothecium pallidum, flavescens, ex hyphis valde tenuibus et intricatis contextum; hymenium fere decolor, tantum superne dilute flavens, purum, 100-110 », altum, I e coeruleo mox rufescenti-obscuratum; paraphyses graciles, strictae, simplices, eseptatae, ad apicem haud latiores; asci copiosi, angusti, subcylindrici et paulum clavati, hymenio subaequilongi, 8 spori; sporae in ascis subuniseriales, decolores, ellipsoideae vel ellipsoideo-fusiformes, versus apices angustatae, uniseptatae, membrana et septo tenui, ad septa non con- strictae, rectae, 9,5;—11 » longae et 3—3,5 p. longae. Pycnoconidia non visa. Masatierra: an den siidlichen Abhaingen des El! Yunque, 500 m, auf Drimys-Stammen (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 333 Die Lagerfaden, Gestalt und Grodsse der Sporen sind wie bei Coenogonium complexum Nyl., aber die Paraphysenende sind nie kopfformig, ausserdem ist das Lager gleichmassig ausgebreitet und tragt keine zottigen Biischel. Racodium Fr. R. rupestre Pers. Cystocoleus rupestris ‘Thw.; Jonow, Estud. Flora Juan Fernand., 1896, p. 203. Masatierra: ohne naheren Standort,.auf vulkanischem Gestein (JOHOW.) Collemaceae. Lemmopsis (Wain.) A. Zahlbr. *L. polychidioides A. Zahlbr. nov. spec. — Taf. 25, Fig. 7. Thallus minute filamentose et irregulariter ramosus, ramis densis, filifor- mibus, cylindricis, vix 0,1 mm latis, sat brevibus,;decumbentibus, rigidiusculis et implexis, luride fuscescentibus, madefactis obscurioribus, homoeomericus, omnim paraplectenchymaticus, cellulis marginalibus angulosis, internis magis rotundatis, omnibus valde leptodermaticis, 6—11 p. latis, ad ambitum thalli extus strato tenui, amorpho, lutescente passim tectis; gonidia nostocacea, concatenata, catenis ut plurimum abbreviatis, cellulis rotundis, 5—6 yp. latis. Apothecia biatorina, sessilia, 1—1,5 mm lata, rufa, opaca, e concaviusculo demum convexa; margo disco paulum dilutior, tenuis, integer; receptaculum ex hyphis radiantibus, septatis (cellulis cylindricis) et leptodermaticis formatum; excipulum dimidiatum, tantum ad latera hymenii evolutum et leviter infra hy- menium productum, ex hyphis tangentialibus, tenuibus, dense contextis com- positum; hypothecium sat crassum, fere hymenio subaequilatum, subdecolor (pallide lutescens), ex hyphis intricatis formatum; hymenium superne anguste rufofuscum, caeterum decolor et purum, 100-110 y. altum, I e coeruleo cupreo- luridum; paraphyses filiformes, conglutinatae, simplices, eseptatae, ad apicem clavatae; asci cylindrico-clavati, hymenio subaequilongi, ad apicem rotundati et membrana modice incrassata cincti, 8 spori; sporae in ascis uni- vel subbiseriales, simplices, decolores, ovales vel late ellipsoideae, utrinque acutatae, membrana tenui cinctae, 12—15 p. longae et 7,5—8 yp. latae. Pycnoconidia, ignota. Masatierra: Cordén Chifladores, auf morschen. Baumzweigen (C. und I. SKOTTSBERG). Im Habitus gleicht die neue Art sehr dem Polychidium muscicolum, nur die Apothezien sind heller und zeigen das Rotbraun der Pannarza-Apothezien. Das Lager ist anatomisch gebaut wie die Arten der Sekzion Homodium der Gattung Leptogium. Lemmopsis Arnuldiana (Hepp) A. Zahlbr. besitzt eine obere und untere Rinde aus eckigen Zellen gebildet, das Innere ist homoemerisch, wie bei Collema; diese Art entspricht demnach in ihrem anatomische Bau der Sekzion Euleptogium. Somit sehen wir derzeit bei Lemmopszs zwei anato- 334 A. ZAHLBRUCKNER mische Typen inbezug des Lagers und es lassen sich darauf in analoger Weise zwei Sekzionen: Lulemmopsis A. Zahlbr. nov. sect. und Homodiopsis A. Zahlbr. nov. sect. begriinden. Physma Mass. *P, chilense Hue. — Taf. 25, Fig. 4. HveE in Bull. Soc. Linn. Normand., ser. 5, vol. IX, 1906, p. 126. — Physma tricolor A. Zahlbr. in Annal. Mycolog., vol. VI, 1908, p. 132. — Lrcuter, Plant. Chilens. exsicc. no. 637! Masatierra: Portezuelo de Villagra, bei 500 m, auf Astchen der Fagara Mayu, fruchtend (C. und I. SKOTTSBERG). Leptogium (Ach.) S. Gray. A. Thallus subtus nudus, rhizinis nullis. a. Apothecia pustulis thalli inflatis, excavatis immersa; margo thal- linus plicato-rugosus L. phyllocarpum. b. Apothecia peltato-sessilia, lecanorina; margo + integer. ga. Thallus tenuiter isidiosus L. caestum. fo 6. Thallus nudus, isidiis destitutus. I. Stratum medullare tenue, haud distincte mucosum. 1. Thallus plumbens vel coerulescens JL. molluccanum. 2. Thallus atroviridis L. moluccanum f£. marianum. II. Stratum medullare bene evolutum et bene mucosum L. tremellotdes. B. Thallus subtus dense albo-rhizinosus L. Menziesiz. C. Thallus utrinqgue tomentosus L. callithamnium. *=L. moluccanum Wain. Watn., Etud. Lich. Brésil, vol. I, 1890 p. 223; A. ZaHLBR. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 12. — Collema moluccanum Pers. apud Gaunpicu., Voy. Uranie, Bot., 1826, p. 203. Haufig auf Juan Fernandez. Masatierra: Quebrada Damajuana, 248 m, auf Baumrinden im Walde, fruchtend; Bahia Cumberland, Cordén Central, 530 m, auf Denxdroseris, fruch- tend, Valle Colonial, Quebrada del Portezuelo, 349 m, im Walde auf Rinden, fruchtend; Quebrada seca, auf Dendroseris micrantha; Quebrada Monte Ma- derugo, auf Rinden; Cordén Salsipuedes, auf Wargyricarpus, fruchtend; Riicken zwischen Quebrada Laura und Quebrada Piedra agujereada, 625 m, auf Rin- den, fruchtend (C. und I. SKOTTSBERG). Masafuera: Quebrada de la Loberia, 250—300 m, auf Baumrinden und Steinen im Walde, fruchtend (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 335 f. marianum Wain. Warn., Efud. Lich. Brésil, vol. I, 1890, p. 224. —- Collema marianum Pers. apud Gavupicu., Voy. Uranie, Bot., 1826, p. 203; Monr. in Annal. Scienc. Nat., Bot. ser. ‘2; vol. IV,°1835, p: 93 Masatierra: Auf Baumrinden (BERTERO no. 1645). — Vielleicht nur der Typus der Art. L. tremelloides S. Gray. S. Gray, A Natur. Arrang. Brit. Plants, vol. I, 1821, p. 400; Watn., Etud. Lich. Brésil, vol. I, 1890, p. 224 et in Annal. Acad. Scient. Fennic., ser. A, vol. VE no. 7, 1915, p. 107. — Lichen tremelloides Linn. f., Suppl. Spec. Plant., 178r, Pp. 450. Masatierra: Quebrada seca, 435 m, auf J/yrceugenia-Rinde, fruchtend (C. und I. SKOTTSBERG). var. azureum Nyl. Ny. Synops. Lich., vol. I, 1858, p. 125. — Lzchen azureus Sw. apud Acu., Lichenogr. Suec. Prodr., 1798, p. 137. — Collema azureum Ach., Lichenogr. Uni- vers., 1810, p. 664; Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. 1V, 1835, p. 93. — Leptegium azureum Mont. apud Wess, Hist. Natur. Iles Canar., vol. II, 2 part., 1840, p. 129; Croms. in Journ. Linn. Soc. London, Bot., vol. XVI, 1877, p. 223 et in Rep. Voy. Challenger, Bot., 1835, p. 89. Masatierra: Auf Baumrinden, ohne nahere Standortsangabe (BERTERO no. 1640; MOSELEY). *L. caesium Wain. Wain., Etud. Lich. Brésil, vol. I, 1890, p. 225. — Collema tremelloides var. caestum Ach., Lichenogr. Univers., 1810, p. 656. Masatierra: Quebrada seca, 435 m, auf Rinden, steril und Quebrada Juanango, 250 m, auf Rinden, fruchtend (C. und I SKOTTSBERG). Masafuera: Quebrada de la Loberia, 280 m, auf glatten Baumrinden, steril (C. und I. SKOTTSBERG). L. phyllocarpum Mont. Mont. in Annal. Scienc. Nat., Bot., ser. 3, vol. X, 1848, p. 134; WAIN., Etud. Rich? Brésil;; voli. I) 1890); p. 1230. — Gall phyllocarpum Pers. apud Gaupicu., Voy. Uranie, Bot., 1826, p. 204; Mont. in Annal. Scienc. Nat., Bot., semf2iivek PVE LOS. Ds 03. Masatierra: Auf Baumzweigen in den feuchteren und kihleren Gebirgs- waldern (BERTERO no. 1644). Cordén Centinela, 530—600 m, auf Baumasten, fruchtend und Portezuelo de Villagra, bei 500 m, auf /agara Asten, fruchtend (C. und I. SKOTTSBERG). 226 A. ZAHLBRUCKNER 33 *J7|, Menziesii Mont. Mont. in Annal. Scienc. Nat., Bot., ser. 3, vol. XVIII, 1852,"p. 313-et apud Gay, Hist. Fisic.y Polit.,;Chile, Bot., vol. VII, 1852,;p. 223, tab. XIIL) fieagee Nyt., Synops. Lich., vol. I, 1858, p. 128; Hur in Nouv. Archiv. du Muséum, ser. vol. X, 1898, p. 229. — Lichen Menziesti Sm. apud Acu., Method. Lich., 1803, 4 Por p: 22. Masafuera: im Hochlande, bei 1,200 m, steril (C. und I. SKOTTSBERG). L. callithamnium Ny]. Nv. Synops. Lich., vol. I, 1858, p. 126; Warn, in Annal. Acad. Scient Fennic., ser. A, vol. VI, 1915, p. 105. — Stecta callithamnia ‘Yayl\. in Hoox., Lon- don Journ. of Bot., vol. VI, 1847, p. 183. — Stetina callithamnia Mill. Arg. in Flora, vol. LXXI, 1888, p. 164; StzBer. in Flora, vol. LXXXI, 1895, p. 137. Masatierra: ohne Standortsangabe (BERTERO no. 163g!); Cordén Cen- tinela, 530 m, auf Baumrinden, steril und Quebrada seca, 435 m, auf Baum- rinde, steril (C. und I. SKOTTSBERG). Den anatomischen Bau des Lagers fand ich mit der Beschreibung WAINIO’s vollig tibereinstimmend; dieser und die Gestalt der Sporen sind fiir die gene- riche Zuweisung ausschlaggebend. Pannariaceae. Parmeliella Mill. Arg. A. Thallus caespitoso-polyphyllus ut in Dermatocarpone aquatico, livido- fuscescens, species saxicola P. symptychta- B. Thallus subcrustoso-squamulosus, squamis planis et adpressis, pallidus; species corticola P. nigrocincta. P. nigrocincta Mill. Arg. Mitty. Arc. in Flora, vol. LXIV, 1881, p. 86. — FParmelia nigrocincta Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 91 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 142; JoHow, Estud. Flora Juan Fernandez, 1896, p. 200. — Pannaria nigrocincta Nyl. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1355, p. 182 et Synops. Lich., vol. Il, 1863, p. 39; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. X, (1908) 1909, p. 204. Masatierra: ohne Standortsangabe, auf Rinden (BERTERO); Cordén Chifla- dores, 350 m, auf Rodbinsonia thurifera, auf dem Riicken iiber Pangal, 400 m, und Quebrada seca, 435 m, auf Coprosma (C. und I. SKOTTSBERG). Masafuera: Quebrada Loberia, 300 m, auf Baumrinden (C. und I. SKOTTS- BERG). Parmeliella(?) symptychia A. Zahlbr. nov. comb. Lannaria symptychia Tuck. in Proceed. Americ. Acad. Acts and Scienc., vol. XII, 1877, p. 168 et Synops. North Americ. Lich., vol. II, 1888, p. 144. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 337 Auf Felsen, ohne nahere Standortsangabe (HILL). Ich sah diese Flechte nicht und bin tiber ihre Gattungszugehorigkeit nicht sicher. Pannaria Del. A. Thallus omnino paraplectenchymaticus P. fuegiensis. B. Thallus stratosus,. a. Thallus laciniato—incisus; apothecia rufa, sporae 17—30 p. longae P. rubiginosa. b. Thallus fere crustaceus, in margine minute lobulatus; apothecia alutacea; sporae usque 14 p. longae P. hilarts. *P. fuegiensis A. Zahlbr. A. Zaucpr. in Kgl. Svensk. Vet.-Akad. Handl., vol. LVII, no. 6. 1917, p. 13. Masatierra: Quebrada Juanango, auf abgestorbenen Baumstammen (C. und I. SKOTTSBERG). *P, hilaris A. Zahlbr. nov. spec. Thallus in hypothallo nigro, tenui primum rosulas parvas, plus minus ro- tundatas, subplacodiformes, hypothallo arcte adpressas format, rosulae demum confluent et in thallum subcrustaceum, usque 5 cm latum abeunt, thallus valde tenuis, 0,;—0,8 mm crassus, cinereo-caesius, opacus, KHO —, CaCl,O, —, centro utplurimum pulverulento-granulosus, subareolatim rimosus, (areolis parvis) et in ambitu effiguratus, lobi marginales rotundati, incisi vel inciso-crenati, arcte adpressi, ab hypothallo cincti; soredia et isidia desunt; subtus non rhizi- nosus; superne corticatus, cortex decolor, 24—28 yp. crassus, grosse paraplecten- chymaticus, cellulis I—3 superpositis, 6—18 yp, latis, valde leptodermaticis, su- perne strato tenui amorpho tectis; stratum gonidiale latum, gonidiis olivaceis vel olivaceo-aeruginosis, rotundatis, 3,5—4 p. latis, hyphis intricatis, leptoder- maticis, usque 1,8 p. crassis; stratum medullare ex hyphis plus minus horizon- talibus, densis formatur. Apothecia lecanorina, adpressa, parva, usque 0,8 mm lata, rotunda, plus minus approximata; discus alutaceus, opacus, epruinosus, subplanus; margo thallinus tenuis, thallo concolor, leviter prominulus, primum subgranulosus, demum integer, paraplectenchymaticus, hyphis flabellatim dispositis, leptoder- maticis, stratum gonidiale includens; hymenium decolor, purum, I20—145 ». altum, I e coeruleo aeruginoso-sordidum; hypothecium angustum, decolor, ex hyphis intricatis formatum; paraphyses filiformes, conglutinatae, simplices, esep- tatae, ad apicem non latiores; asci anguste clavati, ad apicem rotundati et membrana modice incrassata cincti, 8-spori; sporae in ascis subbiseriales, de- colores, simplices, ovali-ellipsoideae, membrana tenui et laevi cinctae, 9—14 yp. longae et 6- 7,5 y. latae. Pycnoconidia non visa. in 22—2391. The Nat. Hist. of Juan Fernandez and Easter Ish. Vol. 11. 3238 A. ZAHLBRUCKNER Masatierra: Portezuelo, beim SELKIRK-Denkmal, 6co m, auf der Rinde einer Rodinsonia (C. und I. SKOTTSBERG). Habituell kommt diese neue Art noch am meisten der Pannaria nebulosa (Hoffm.) nahe, aber die tbrigen Merkmale sprechen gegen eine nahere Ver- wandtschaft. P. rubiginosa Del. Dev. \im |Dictionn. Class. Hist.) Nat.,-“vol; XIIE, 1828," p. 20; Nye. Synops: Lich,, vol. II; 1863, p. 29, tab. IX, fig. 15. — Lichen rubiginosus Thunbeesapua Acu., Lichenogr. Suec. Prodrom,- 1798, p. o9. — Parmelia rnbiginosa Ach., Method, Lich., | 1803, ip. 212) Mont. in -Annal. Scienc. Nat., Bot., ser. 2) vol. avenue sim go et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 141; JoHow, Estud. Flora Juan Fernandez, 1896, p. 199. Masatierra: Auf Rinde der J/yrceugenza, ohne nahere Standortsangabe (BERTERO no. 1632, 3008); Plazoleta del Yunque, tiber Moosen (C. und I. SKOTTSBERG). *var vulcanica A. Zahlbr. nov. var. Thallo microphyllino et sporis brevioribus differt a typo. Squamulae thalli minutae, incisae, in centro thalli plus minus congestae, ad ambitum thalli dispersae, in hypothallo atro sedentes; sporae 13—16 p. longae et 8—9 yp. latae. Vielleicht handelt es sich um eine eigene Art, doch war das mir vorlie- gende Material zur Entscheidung der Frage nicht ausreichend. Masatierra: Cerro Damajuana, auf losen Steinen im Wald (C. und I. SKOTTSBERG). Massalongia Korb. =M. carnosa Korb. K6rB. Syst. Lich..German., 1855, p. roo; A..L. Smitu, Handb. Brits lich. vol. Ty 1918, p.. 88,° tab: 30. — Lichen carnosus Wicks., Fase. Plant) Cxypepeumee VOLE ER TOO mp 2), etalon Leste n-ne Masafuera: auf Blocken beim Campo Correspondencia, 1150 m (C. und I. SKOTTSBERG). Psoroma Ny. A. Species saxicola, thallus sorediis instructus P. vulcanicum. B. Species ad cortices, ligna vel muscos vigentes; thallus esorediosus. a. Thallus cephalodiis praeditus P. cephalodinume. b. Thallus cephalodiis destitutus. I. Thallus squamosus. I. Thallus luteo-pallens. a. Squamae thalli substrato adpressae P. pholidotum. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 339 6. Squamae thalli + erectiusculae P. dasycladum. 2. Thallus lurido-cervinus P. sphinctrinum. II. Thallus ramoso-divisis P. angustisectum. *P, vulcanicum A. Zahlbr. nov. spec. Thallus epilithicus, maculatim effusus, subtartareus, alutaceus vel pallide cervinus, opacus, KHO vix mutatus, Ca Cl,O, —, versus ambitum squamulosus, squamulls parvis, 0,3—0,8 mm latis, rotundatis, planis, integris vel incisis, plus minus dispersis et hypothallo atro insidentibus, centrum versus granulosus vel subgranulosus, granulis congestis, demum subareolatim fissis, passim sorediosus, sorediis leprosis, depressis, canis, rotundatis, planiusculis, thallum non superan- tibus; superne corticatus, cortice 24—28 yp alto, paraplectenchymatico, cellulis rotundatis, luminibus 5—9 u. latis praeditis, membrana mediocri cinctis, superne strato angusto, fuscescente tecto; caeterum fere homoeomericus, cellulis goni- diorum plus minus globosis, usque 15 yp, latis, contentu laete viridi, membrana mediocri cincti, subtus anguste nigricantifuscus et ex hyphis intricatis formatus. Apothecia lecanorina, sessilia, ad basin breviter constricta, rotundata et dispersa, vel approximata et dein pressione mutua subangulosa, I—1I,z2 mm lata; discus rufus, opacus, planiusculus; margo thallinus integer, primum bene prominulus et crassiusculus, demum plus minus depressus et subcrenulatus, extus corticatus, cortice paraplectenchymatico, superne angustiore, e seriebus cellularum 1—4, rotundatis formato, gonidia copiosa includens; excipulum infra hymenium bene evolutum, ad latere hymenii angustatum, sordide fuscescens, ex hyphis intricatis formatum et maculis minutis rotundatisque praeditum; hy- pothecium decolor, sat tenue, ex hyphis tenuissimis et intricatis formatum, hy- nenium superne dilute fuscescens, caeterum decolor et purum, usque 180 uv. altum, I e sordide coerulescente cupreo-rufescens; paraphyses filiformes, dense contextae, strictae, simplices, eseptatae, ad apicem non latiores; asci anguste clavati, hymenio subaequilongi, ad apicem rotundati et ibidem membrana paulum incrassata cincti, 8-spori; sporae in ascis subuniseriales, decolores, simplices, ovali-ellipsoideae vel ellipsoideae, in uno apice vel utrinque acutatae, membrana leviter granulosa cinctae, 17—21 . longae et 7,5—8,5 u. latae. Pycnoconidia ignota. Masafuera: Heide auf dem Hochland, 1100—1300 m, auf vulkarischem Gestein (C. und I. SKOTTSBERG). Die charakteristischen Merkmale liegen in dem mehr weniger kornigen Lager, dem Vorhandensein der Soredien und in der Farbe und Berandung der Apothezien. *P, cephalodinum A. Zahlbr. nov. spec. — Taf. 24, Fig. 7. Thallus placodinus, substrato adpressus, plagas usque 3,5 cm latas for- mans, lutescenti-pallidus passim albidus, opacus, KHO —, Ca C!,0O, —, tenuis, usque 0.2 mm crassus, lobi marginales continui vel subimbricati, distanter di- visi, in apice paulum latiores, in margine incisi et amguste pallidiores, usque 2 mm lati, planiusculi vel subcanaliculati, superne nudi, in centro lobulis mi- noribus vel hinc inde subsquamulosus, subtus nigrescens, sorediis et isidiis nullis, 340° A. ZAHLBRUCKNER sed in superficie cephalodiis dispersis obsitus, superne corticatus, cortice an- gusto, 18—24 p. crasso, paraplectenchymatico, cellulis superpositis 2—3, lumine plus minus rotundo, parietibus crassiusculis; stratum gonidiale infra corticem situm et eo paulum latius, continuum, cellulis laete viridibus, globosis, g—12 p. latis, membrana distincta cinctis; medulla alba, ex hyphis intricatis, non inspersis formata, hyphis ad 3 pw crassis, leptodermaticis, inferne in stratum angustum nigricantem abeuntibus; cephalodia I—1,8 mm lata, depressa, in ambitu rotun- data vel irregularia, in superficie paulum inaequalia, subfarinosa vel granulosa, coerulescenti-cinerea vel plumbea, ex hyphis dense intricatis, 1,7—1,8 p. latis, leptodermaticis, increbre septatis formata, gonidia concatenata et hinc inde glo- merulosa, pallide olivacea, rotundata vel oblonga, rarius subirregularia, 5— 8 p lata, membrana tenuissima cincta includentia. Apothecia lecanorina, sessilia, usque 2 mm lata, ad basin bene constricta, copiosa et approximata, rotunda vel rotundata; discus in juventute pruinosus, demum nudus, alutaceus vel alutaceo-rufescens, opacus; margo thallinus cras- siusculus, crenulatus, prominulus et plus minus inflexus, extus strato corticali, ex hyphis intricatis et inspersis formato, usque 20 », crasso, ad basin latiore abductus, gonidia copiosa, usque ad verticem) marginis penetrantibus includens; excipulum dimidiatum, decolor, ex hyphis tangentialibus et dense conglutinatis formatum; hymenium decolor, tantum superne anguste fuscescens et inspersum, caeterum purum, I e coeruleo lutescenti-obscuratum; hypothecium lutescens, molle, inferne substipitato-productum et strato medullari thalli superpositum; paraphyses filiformes, ad 2 y, latae, conglutinatae, simplices, eseptatae, ad apices haud latiores; asci oblongo- vel ellipsoideo-clavati, hymenio paulum breviores, ad apicem rotundati et membrana bene incrassata cincti, 8-spori; sporae in ascis biseriales, decolores, simplices, ellipsoideae vel ovali-ellipsoideae, mem- brana tenui et sublaevi (subindistincte granulosa) cinctae, 11—15,5 uw longae et 8,5—9 uw. latae. Pycnoconidia non visa. Masatierra: Cordén Chifladores, 350 m, auf Dyrzmys-Rinde (C. und I. SKOTTSBERG). Das plakodiale Wachstum des Lagers und das Vorhandensein von Zephalo- dien kennzeichnen die Art gut. P. pholidotum Mill. Arg. Murr. Arc. in Flora, vol. LXXI, 1888, p. 45 (excl. syn.). — Parmelia pholi- dota Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 91 et apud Gay, Hist Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 146; Jonow, Estud. Flora Juan Fernandez, 1896, p. 200. — Pannaria pholidota Ny\. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 182, Synops. Lich, vol. Il, 1863, p. 30; Humane ous Archiv., du Muséum, ser. 4, vol. X, (1908) Igu9, p. 17). Masatierra: Auf Drzmys-Rinde, ohne nahere Standortsangabe (BERTERO no. 1623, 1626), Cordén Chifladores, 350 m, auf Stammen der Rodbznsonia thu- rifera; Rabanal, auf Robinsonta gracilis; Valle Colonial, Quebrada seca, 435 m, auf Dendroseris micrantha; Portezuelo, beim SELKIRK-Denkmal, 590 m, auf Robinsonia gracilis; Cordén Salsipuedes, 625 m, auf Margyricarpus (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 341 P. sphinctrinum Nyl. Nyy. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 181 et Synops. Lich., vol. II, 1863, p 24; REINKE in PrRincsu., Jabrb. fiir wiss. Bot., vol. XXVIII, 1895, p. 446, fig. 165 et 166. — Farmelia sphinctrina Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 90 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 139; JoHow, Estud. Flora Juan Fernandez, 18,6, p. 199. — Ffannaria sphinctrina Yue in Bull. Soc. Bot. France, vol. XLVIII, (1901) 1902, p. LVI et in Nouv. Archiv. der Muséum, ser. 4, vol. VIII, 1907, p. 265. Masatierra: auf Rinden, ohne nahere Standortsangabe (Bertero no. 1625, 1630); Cordon Chifladores, 350 m, auf der Rinde der Rodinsonia thurifera und auf Drimys; Cordon Centinela, 530 m, auf Dendroseris micrantha; Portezuelo, auf Rinden beim SELKIRK-Denkmal, 590 m, auf Rodzxsonia gracilis; Quebrada seca, 435 m, auf Dendroserts micrantha; Cordon Salsipuedes, auf Drimys; Quebrada Juanango, 250 m, auf Baumrinden (C. und I. SKOTTSBERG). Masafuera: Cordon del Barril, auf Drzmys-Stammen (C. und I. SKorts- BERG). var. leprolomum Ny]. Nyt., Lich. Nov. Zeland., 1888, p. 52. — Lecanora sphinctrina var. leproloma Nyl. in Compt.-Rend. Séanc. Acad. Paris, vol. LXXXIII, 1876, p. 89. — Pannaria sphinctrina var. leproloma Hue in Bull. Soc. Bot. france, vol. XLVIII, 1901, p. LVI et in Nouv. Archiv. du Muséum, ser. 4, vol. VIII, 1907, p. 269. Masatierra: Cordén Salsipuedes, auf Baumrinden (C. und I. Skorts- BERG). var. crispellum Nyl. Ny. Synops. Lich., vol. II, 1863, p. 25 et Lich. Nov. Zeland., 1888, p. 52. Masatierra: Portezuelo, c. 600 m, auf der Rinde einer Rodzxsonia thu- rifera (C. und I. SKOTTSBERG). *P, dasycladum A. Zahlbr. nov. spec. Thallus squamulosus, squamulis dense contextis, sat parvis, I—I,8 mm latis, erectiusculis et plus minus imbricatis, in ambitu thalli rotundato-flabellatis, incisis et crenatis, centralibus magis rotundatis, omnibus tenuibus (0,25— 0,4 mm crassis), cervinis vel cervino-fuscidulis, opacis, KHO et CaCl,QO, non tinctis, epruinosis, in margine passim subplumbeo-cinerascentibus, subtus albidis et nudis, versus insertionem rhizinis nigricantibus obsitis; sorediis et isidiis nullis; superne corticatus, cortice strato amorpho tenui supertecto, decolore, paraplecten- chymatico, cellulis minutis, lumine ad 2 w lato, rotundatis vel subangulosis, membrana mediocri cinctis, in seriebus superpositis 4—10; stratum gonidiale infra corticem dispositum, subcontinuum, ex hyphis in parte superiore plus minus verticalibus, in parte basali magis intricatis formatum, gonidiis globosis, laete viridibus, 10—24 y. latis; cortex hinc inde etiam latera squamulorum obducit; stratum medullare ex hyphis praesertim longitudinalibus, 3—5 yp. latis, fasciatim- 342 A. ZAHLBRUCKNER subintricatis formata, ad basin squamorum in hyphas coerulescenti-nigricantes, usgue 9 } crassas, sat pachydermaticas et increbre septatas abientibus. Apothecia lecanorina, inter squamulas sessilia, primum cupuliformia, demum plus minus explanata, ad basin bene constricta, usque 3 mm lata, rotunda vel demum subsinuata; discus obscure rufo-fuscus, opacus, epruinosus, e conca- viusculo subplanus vel convexulus, superne inaequalis; margo thallo concolor, sat crassus, prominulus et leviter inflexus, crenulatus vel lobulis. parvis incisis- que obsitus, corticatus, medullam et gonidia includens; excipulum dimidiatum, ad latera hymenii flabellatum, ad verticem dilatatum, ex hyphis tangentialibus, 1,6—1,8 ¥ crassis, pachydermaticis et increbre septatis formatum, superne fuscescens, caeterum decolor, molle; hypothecium decolor, ex hyphis intricatis, sat crassis formatum; hymenium superne rufescenti-fuscescens et strato sat an- gusto, amorpho et decolore obductum, caeterum decolor, purum, I e coerulescente cupreo-sordidulum; paraphyses filiformes, 2—3 crassae, conglutinatae (im- primis in parte superiore), simplices, eseptatae, ad apicem haud latiores; asci cylindrico-clavati, ad apicem rotundati, 8-spori; sporae in ascis uniseriales, de- colores, simplices, oblongae vel late ellipsoideae, utrinque breviter acutatae, mem- brana tenui, extus inaequali cinctae, 14—18 p longae et 8—11 yp, latae. Pycnoconidia non visa. Masatierra: Felskamm unweit Tres Puntas, 350 m, auf Humuserde (C. und I. SKOTTSBERG). Durch die Wachstumsweise des Lagers gegentiber den anderen Arten der Gattung gut gekennzeichnet. *P, angustisectum A. Zahlbr. nov. spec. — Taf. 25, Fig. 3. Thallus substrato latiuscule adhaerens, depressus, fusco- vel lurido-fla- vescens, nitidulus, KHO et CaCl,O, non tinctus, ex initiis verruculosis mox iteratim dichotome et sympodialiter ramosus, ramis angustis et elongatis, usque 3 mm longis, 0,16 —0,25 mm latis, 0,15 —O,r7 mm crassis, subtoruloso-inaequalibus, planiusculis vel convexulis, hinc inde minutis, subcorallinis, ramis ultimis ad apicem rotundatis vel subretusis, omnibus superne laevigatis, inferne paulum pallidioribus, erhizinosis; sorediis et isidiis destitutus; protothallus obscuratus nullus; superne corticatus, cortex paraplectenchymaticus, e seriebus cellularum superpositarum 3—5, subangulosarum vel rotundarum, 8—1q4 p. latarum, mem- brana tenui cinctarum formatus; pars caetera thalli etiam paraplectenchyma- ticus, sed cellulae multum minores, in strato gonidiali ex hyphis perpendicula- ribus, in parte inferiore thalli ex hyphis longitudinalibus, leptodermaticis et crebre septatis formatus; stratum gonidiale infra corticem dispositum, continuum, circa dimidium crassitudinis thalli occupans, gonidiis laete viridibus, globosis, 5—6 yp. latis. Apothecia lecanorina, sessilia, majuscula, usque 5 mm lata, ad basin bre- viter constricta; discus rufus, opacus, epruinosus, e concavo subplanus; margo thallinus primum verruculosus, mox ramis thallinis corallinoideis, brevibus plus minus obsitus; receptaculum extus laeve, corticatum, cortice crasso, paraplecten- chymatico, luminibus cellularum oblongis vel rotundatis, cellulis sat pachyder- maticis, in margine crenulato-erosum et ad basin rhizinosum, rhizinis decolo- tibus, simplicibus, breviusculis, medullam et gonidia includens; excipulum an- DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 343 gustum, bene limitatum, usque ad verticem hymenii assurgens, decolor, ex hyphis tangentialibus et conglutinatis formatum; hypothecium angustum, fusce- scenti lutescens, molle, ex hyphis intricatis formatum; hymenium superne rufum, caeterum decolor, purum, 140—150 yp, altum, I e coeruleo cupreo-obscuratum; paraphyses filiformes, strictae, conglutinatae (imprimis in parte superiore hy- meni), simplices, eseptatae, ad apicem vix crassiores; asci hymenio parum breviores, clavati, recti vel subrecti, ad apicem rotundati et membrana parum incrassata cincti, 8-spori; sporae in ascis uni- vel subbiseriales, decolores, sim- plices, ovales vel ellipsoideo-subfusiformes, membrana tenui, verruculoso-inae- quali cinctae, contentu oleoso, 22— 38 p, longae et 8—15 », latae. Pycnoconidia ignota. Masatierra: Quebrada Piedra agujereada, bei 650 m, iiber Moosen, und auf dem Riicken oberhalb Pangal, 795 m, iiber Moosen und auf dem Erdboden (end I SKOLTSBERG), Die ersten Thallusanfange erinnern an das Lager des Psoroma hypnorum (Hoffm.); die fertige Flechte hat indes mit disser nichts zu tun, so dass man die Annahme, als lage ein abnorm ausgebildetes Lager derselben vor [etwa wie die var. thelephorotdes Th. Fr. der Ochrolechia tartarea (L.)| zuriickweisen muss, denn die anatomischen Verhaltnisse und auch andere Merkmale stellen sich einer solchen Annahme entgegen. Coccocarpia Pers. *C. Gayana (Mont.) Nyl. var. subdivisa A. Zahlbr. nov. var. — Tab. 24, Pig. 5. Thallus superne glaucus, opacus, lobatus, lobis substrato adpressis vel imprimis in ramulis arborum plus minus liberis et adscendentibus; lobi adpressi rotundato-incisi, in margine crenulati, concentrice non striati; lobi liberi + elongati, in margine incisi et crenulati; rhizinae paginis inferioris ad ambitum thalli albae, in centro aeruginoso-nigricantes. Apothecia habitu iis plantae typicae similia, Receptaculum gonidia nulla includens, extus corticatum, cortice decolore, paraplectenchymatico, e seriebus pluribus cellularum superpositarum formato, cellulis rotundis vel subrotundatis, inferne majoribus, ad 12 y, latis, versus ambitum cubicis sensim minoribus, mem- brana mediocri cinctis; excipulum integrum, in margine flabellatum, infra hy- menium saepe angustius, decolor, ex hyphis tangentialibus, dense contextis for- matum; hypothecium dilute luteo-ochraceum, molle, ex hyphis tenuissimis et intricatis formatum; hymenium superne strato tenui et amorpho tectum, decolor, purum, I cupreo-rufum; paraphyses filiformes, simplices, eseptatae, ad apicem non incrassatae, conglutinatae; asci hymenio breviores, clavati, 4-spori; spori in ascis biseriales, decolores, simplices, ovales vel ellipsoideo-ovales, membrana tenui laevique cinctae, contentu oleoso pellucido, 12—14 p, longae et 7—7,5 p. latae. . Masatierra: auf dem Abhange iber Pangal, bei 795 m, auf Baumzweig- lein (C. und I. SKOTTSBERG). 344 A. ZAHLBRUCKNER Stictaceae. Lobaria Schreb. L. crenulata Trevis. Trevis. Lichenoth. Veneta, 1869, no. 75; WaAIN., Etud. Lich. Brésil, vol. I, 1890, p. 197; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. Ill, 1901, p. 35. — Parmelia crenulata Hook. apud KuntH, Synops. Plant. Aequin. Orb. Novi, vol. I, 1822, p. 23; Mont. in Annal. Scienc. Nat., Bot, ser. 2, vol. IV, 1835, p. go et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 130; Jonow, Estud. Flora Juan Fernand., 1896, p. 199. — Stecta crenulata Vel., Hist. Lich. Sucta, 1822, je ezrse UO A CIINAS Hae Leiba Ricasolia crenulata Ny\. in Mémoir. Soc. Scienc. Nat. Cherbourg, vol. V, 1857, p. 104 et Synops. Lich., vol. I, 1860, p. 372. Masatierra: auf Rinden in den Waldern der hoheren Berge (BERTERO no. 1633); Quebrada Juanango, 250 m, auf Baumstammen, reichlich fruchtend (C. und I. SKOTTSBERG). Sticta Schreb. Conspectus specierum: A. Gonidia nostocacea (S/ccfna). a. Thallus subtus cyphellis veris praeditus. I. Thallus monophyllus, saepe lobatus, lobis latis et rotundatis, in margine non isidiosis St. fuliginosa. II. Thallus laciniatus vel lobato-laciniatus, lobis angustioribus et in magrine plus minus coralloideo isidiosis St. Weigelit. b. Thallus subtus pseudocyphellis praeditus. I. Pseudocyphellae albae. 1. Thallus superne maculis vel sorediis albis cinereisve or- natus. g. Thalli lobi in margine nudi, esorediosi SZ. argyracea. 6. Thalli lobi in margine albosorediati St. intricata. 2. Thallus superne sorediis destitutus. x. Lobi thalli angusti, in margine integri, canaliculati St. fragillima. @. Lobi thalli latiores, in margine rotundato-incisi, plani vel subplani, nunquam canaliculati S¢. czmnamomea. II. Pseudocyphellae citrinae vel luteae. 1. Thallus superne aequaliter reticulato-lacunosus St. Berteroana. 2. Thallus superne plus minus laevigatus, nunquam lacunosus. #. Thallus superne plus minus hirsutus vel puberulus. §. Thallus superne hirsutus, cervinus vel rufescens, subtus concolor St. hirsuta. §§. Thallus superne puberulus, murinus, subtus cer- vino-pallidus St. Gutllemint. @. Thallus superne nudus. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 345 §. Sporae biloculares St. carpoloma. §§. Sporae quadriloculares St. Mougeotiana. B. Gonidia palmellacea vel cystococcoidea (Eusticta). a. Thallus subtus cyphellis veris praeditus. I. Thallus stipitatus; medulla KHO non tincta. 1. Lobi thalli plus minus lineares St. lineariloba. 2. Lobi thalli rotundati et latiores St. latifrons. II. Thallus non stipitatus; medulla KHO sanguineo-rubens St. laciniata var. denudata. b. Thallus subtus pseudocyphellis praeditus. I, Pseudocyphellae albae. I. Receptaculum apotheciorum gonidia includens St. Freycinetit. 2. Receptaculum gonidiis nullis St. Richardt. If. Pseudocyphellae luteae vel flavae. I. Receptaculum gonidiis destitutum; sporae incolores St. nitida. 2. Receptaculum gonidia includens; sporae plus minus obscu- ratae, rarius decolores. a. Apothecia marginalia St. aurala. @. Apothecia superficialia St. orygmaea. A. Sect. Stictina (Nyl.) Hue. S. argyracea Del. Dev. Hist. Lich. Sticta, 1822, p. 91, tab. VII, fig. 3; Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 116; Hers. in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXI, afd. III, no. 13, 1896, p. 29; Hue in Nouv. Archiv. du Muséum, ser. 4, vol. III, 1901, p. 87. — Lichen argyraceus Bory apud Del., l. s. c. — Stictina argyracea Nyl|., Synops. Lich., vol. I, 1860, p. 334. — Pseudocyphellaria argyracea Wain. in Hedwigia, vol. XXXVII, 1898, p. (34). Masatierra; Auf Baumrinden, ohne Standortsangabe (BERTERO) Porte- zuelo, 590—600 m, steril (C. und I. SKOTTSBERG). Masafuera: auf dem zentralen Hochplateau, bei 1,200 m, in einer Form mit etwas breiteren Lagerabschnitten (C. und I. SKOTTSBERG). var. *sorediifera Del. Det. Hist. Lich. Sticta, 1822, p. 92, tab. VII, fig. 31. — Stsctna argyracea var. sorediifera Nyl., Synops. Lich., vol. I, 1860, p. 334. — /seudocyphellaria argy- racea vat. sorediifera Hellb. in Bihang till Kgl. Svensk. Vetensk.- Akad. Handl., vol. XXV, afd. Ill, no. 6, 1899, p. 24. Masatierra: Puerto Frances, 500 m, fruchtend (C. und I. SKOTTSBERG). 346 A. ZAHLBRUCKNER var. verrucosa Mont. Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIL, 1852, p. amor Jonow, Estud. Flora Juan Fernand., 1896, p. 198. ‘ Masatierra: Auf Baumrinden in den hoher gelegenen Waldern (BERTERO). Ich sah diese Form nicht; wahrscheinlich diirfte es sich nur um eine krankhafte Bildung handeln. S. intricata Del. Det. Hist.. Lich. Sticta, 1822, p. g6,, tab. Vil, fig. 33. — Sictzaeemrmeaee Nyl., Synops. Lich., vol. I, 1860, p. 334. — Pseudocyphellaria intricata Wain. in Hedwigia, vol. XXXVII, 1898, p. (36). Ohne nahere Standortsangabe, in einer Form mit schwarzlicher Lager- unterseite (nach NYLANDER a. a. O.) *S. fragillima Bab. Bas. apud Hoox., Flora Nov.-Zeland., vol. II, 1855, p. 279. — Stsctina fra- gillima Nyl., Synops. Lich., vol. I, 1860, p. 335; REINKE in Princsu. Jahrb. fiir wiss. Bot., vol. XXVIII, 1895, p. 440, fig. 156, Il; Srzser. in Flora, vol. LXXXI, 1895; p.. £23. Masatierra: auf dem Bergriicken zwischen Quebrada Piedra agujereada und Quebrada Laura, 625 m, tiber Moosen, steril (C. und I. SKOTTSBERG). Masafuera: Quebrada Loberia, 280 m, auf Baumstammen im Walde (C. und I. SKOTTSBERG). var. “linearis A. Zahlbr. nov. comb. Stictina fragillima var.. linearis Mill. Arg.! in Flora, vol. LXXI, 1888, p. 23; STZBGR. in Flora, vol. LXXXI, 1895, .p. 129. Masatierra: auf dem Bergriicken zwischen Quebrada Piedra agujereada und Quebrada Laura, 625 m, auf Baumzweigen; Cordon Central, 450 m, uber Moosen, steril; Portezuelo, 600 m, auf Stammen der Rodinsonia thurifera; Cor- dén Salsipuedes, 500 m, auf Drzmys, steril (C. und I. SKOTTSBERG). [Sticta subvariabilis Ny]. Croms. in Journ; Linn. Soc. London, Eot., vol: XVI, 1877, p:. 223,eh apud Hemot., Rep. Challenger, p. go; Jonow, Estud. Flora Juan Fernand., 1896, p. 199. Masatierra: auf moosigen Baumstammen (MOSELEY). Habituell gleicht die Stcta subvariabilis Nyl. sehr der vorher genannten Art, der einzig greif- bare Unterschied liegt in den Gonidien, welche bei der Ersteren dem Pal- mellaceentypus angehoren (vergl. Stzbgr. in Flora, vol. LXXXI, 1895, p. 114). Ich fand in der reichen Ausbeute SKOTTSBERG's auch nicht ein Exemplar der NYLANDER’schen Art und bezweifle, ob die Pflanze MOSELEY’s dazu geh6rt; ich halte sie eher fiir Stécta fragillima var. linearis. Jedenfalls miisste die von DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 347 MOSELEY gesammelte Flechte auf ihre Gonidien gepriift werden, bevor man Sticta subvariabilis als Birger der Flechtenflora von Juan Fernandez aufnimmt.| *S. cinnamomea A. Rich. A. Ricu., Voyage de Découv. de 1’Astrolabe, Bot., vol. I, 1852, p. 28, tab. VII, fig. 3; Hetrs. in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXI, afd. III, no. 13, 1896, p. 32. — Stectina cinnamomea Miill. Arg. in Flora, vol. LXVI, 1883, p. 22. — Pseudocyphellaria citnnamomea Wain. in Philipp. Journ. Scienc., sect. C., vol. VIII, 1913, p. 120. — Séectina fragillima var. dissimilis Nyl., Synops. Lich., vol. I, 1860, p. 336; StzBer, in Flora, vol. LXXXI, 1895, p. 129. Masatierra: Bahia Cumberland, auf Baumrinde, steril; Portezuelo, 350 — 550 m, auf Baumstémmen, steril; Quebrada Juanango, 250, auf Baumrinden, steril; am Fusse des Berges El Yunque, 2— 300 m (C. und I. SKOTTSBERG). Masafuera: Auf dem Zentralplateau und nodrdlich des HOhenkammes, auf Dicksonia; Cordén Atravesado und Cordon del Barril, bei 1300 m, steril (C. und I. SKOTTSBERG). S. Berteroana Mont. — Taf. 24, Fig. 1. Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. go et apud Gay, Hist. Fisic. y Polit. Chile, Bot, vol. VIII, 1852, p. 125; Jonow, Estud. Flora Juan Fernand., 1896, p. 199; A. ZAHLBR. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 16 — Stctina Berteroana Nyl. apud Srzser. in Flora, vol. LXXXI, 1895, p. 128;.Nyz. in Acta Soc. Scient. Fennic., vol. XX VI, 1900, no. 10, p. 8, not. Apothecia versus centrum thalli sparsa vel approximata, subpedicellata, ad basin parum constricta, rotunda, primum cupuliformia, margine inflexo, crassiusculo, lurido, opaco, subintegro, demum dilatata, usque 3 mm lata, plana vel modice convexa, margine angustiore, tenui et crenulato, demum fere eva- nido; discus in juventute caesio-pruinosulus, demum nudus, fusco niger, opacus; receptacnilum extus ochraceo-pallidum, parum inaequale, corticatum, cortice paraplectenchymatico, cellulis rotundis, sat pachydermaticis, pedicellum obdu- cente et in corticem thalli paginae superioris abeunte, superne in ambitu fimbriato-eroso, in parte basali inaequali et ibidem ciliis increbris, obfuscatis munito, medullam angustam, ex hyphis laxiusculis, sublongitudinalibus et ramosis formatam includens, gonidiis destitutum; hypothecium ochraceo-fuscum, ex hyphis intricatis formatum; hymenium 80—100 yw altum, purum, incolor (excepta parte suprema), I e coeruleo rufo- vel cupreo-fuscescens; paraphyses gelatinose conglutinatae, filiformes, simplices, ad apicem clavatae, capite 5—6 v. crasso, strato tenui, amorpho supertectae; asci hymenio paulum breviores, cla- vati, ad apicem rotundati et membrana paulum incrassata cincti, 8-spori; sporae in ascis biseriales, fuscae, oblongo-ellipsoideae vel subfusiformes, uniseptatae, septo tenui, membrana parum crassiore, 22—29 ». longae et 7—8 uy. latae. Meine friihere Angabe iiber die Pseudozyphellen muss ich dahin richtig- stellen, dass dieselben von gelber Farbe sind. Sie sind sehr klein, etwas er- haben und ausserst sparlich. 348 A. ZAHLBRUCKNER Masatierra: Auf Baumrinden in der Bergregion (BERTERO no. 1662); Cordon Chifladores, 350 m, auf Rodznsonta thurtfera; Cordon Centinela, 530 m, auf Dendroseris micrantha; Quebrada Damajuana, auf losen Steinen im Walde; auf dem Sattel iiber Pangal, 795 m, auf Coprosma pyrifolium, besonders reich fruchtend; im Walde der Bahia Cumberland; auf Baumrinden; Quebrada seca, 435 m, auf Dendroseris micrantha; Cordon Salsipuedes, 615 m, auf Denxdroseris pinnata und Margyricarpus; Quebrada Juanango, 250 m, auf Baumen und Strauchern, schén fruchtend (C. und I? SKOTTSBERG). S. hirsuta Mont. Mont. in Annal Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 88 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 106; Johow, Estud. Flora Juan Fernand.. 1896, p. 198. — Stzctina hirsuta Nyl. in Annal. Scienc. Nat., Bot., ser. 4, yol. XV, 1861, p. 41; Sizrser, in Flora, vol. LXXXI, 18095; p: 13% eee cyphellaria hirsuta Wain. in Hedwigia, vol. XXXVIII, 1899, p. (187); Marme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXV, afd. III, no. 6, 1890, De 10. Masatierra: Auf Baumrinden (BERTERO). SKOTTSBERG hat diese Art nicht gesammelt; viclleicht bezieht sich die Angabe MONTAGNE’s auf die nachste Art. S. Guillemini Mont. Mont: in Annal. Science?” Nat} Bot, Ser: 2, vol. 1V,.<7835; pp; ,S7jte nee hirsuta var. Guillemint Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 107. —- Stictina Guillemini Ny apud Hue in Nouv. Archiv. du Muséum, ser. 3, vol. Il, 1890, p. 207; SizBeR. in, Flora, vol. LXXXI. 1895, p.. 13m Receptaculum apotheciorum extus pulverulentum, non hirsutum; hypothe- cium ochraceo-lutescens, ex hyphis intricatis formatum; hymenium gO—10O »p. altum, purum, I e coeruleo aeruginascens; paraphyses filiformes, gelatinoso- conglutinatae, simplices, eseptatae, ad apicem clavato-incrassatae; asci ellip- soideo-clavati, ad apicem rotundati et membrana mediocri cincti, 8-spori; sporae in ascis bi- vel subtriseriales, obliquae, flavescentes, ellipsoideae vel elongato- ellipsoideae, ad apices angustato-rotundatae, rectae, triseptatae, 20—25 y, longae Cis Ouueratae: Von Stcta hirsuta durch die Behaarung der Lageroberseite, der heileren Thallusunterseite und die etwas kiirzeren Sporen verschieden. STIZENBERGER (a. o O,) irrt, wenn er fiir Stzcta Guzllemini die langeren Sporen angibt, in Wirklichkeit ist das Gegenteil der Fall; die Sporen der letzten Art werden bis 36 p. lang und sind beiderseits mehr zugespitzt. Masatierra: Auf Baumrinden, ohne nahere Standortsangabe (BERTERO); Cord6én Chifladores, auf dem Stamme einer Rodbinsonia thurifera, fruchtend (C. und I. SKOTTSBERG). *var. stictica A. Zahlbr. nov. var. Thallus superne marginem versus tenuiter pubescens, caeterum subnudus vel passim omnino nudus, punctis parvis, cyphelliformibus, emergentibus, luteis DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 349 obsitus; receptaculum apotheciorum extus tenuiter pubescens et punctis luteis munitum. Masatierra: Quebrada Juanango, 250 m, auf Baumrinde, fruchtend (C. und I. SKOTTSBERG). Das gelegentliche Vorkonimen von pseudozyphellenartigen, gelben Punkten auf der Thallusoberseite bei Sécfa Azrsuta erwahnt NYLANDER. S. carpoloma Del. Det., Hist. Lich. Sticta, 1822, p. 159; Mont. in Annal. Scienc. Nat., Bot., Seoa,) volo LV, 1895, op: Ss et apud Gay, Hist. Fisic. y Polit. Chile, Bot.,, vol: VII, 1852, p. 111; Jonow, Estud. Flora Juan Fernand., 1896, p. 198. — Stctina carpoloma Nyl., Synops. Lich., vol. I, 1860, p. 339; StzBer. in Flora, vol. LXXXI, 1895, p. 130. — Pseudocyphellaria carpoloma Wain. in Hedwigia, vol. XXXVII, 1898, p. 34; Matme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXV, afd. nie no. 6,°1899, Pp: 33. Die Pseudozyphellen sind an den meisten Stiicken hellgelb, mitunter fast weiss, aber mit unverkennbarem Stich ins Gelbe. Masatierra: Auf Baumrinden (BERTERO); an Steilhangen bei Portezuelo, 400—625 m, auf Baumzweigen fruchtend (C. und I. SKOTTSBERG). Masafuera: Las Torres, bei 1200 m, steril (C. und I. SKOTTSBERG). *f ornata A. Zahlbr. nov. f. Thalli lobi in margine in lobulis linearibus, simplicibus vel ramosis dis- secti. Pseudocyphellae luteo-albidae; medulla alba. Masatierra: Portezuelo, bei 600 m, auf Baumrinden, steril und Cord6én Salsipuedes, auf Straéuchern, fruchtend (C. und I. SKOTTSBERG). S. Mougeotiana Del. Wer, Est. Lich, Sticta, 1822, ip-'62, tabs MN, figs 13 Mon. in. Annaly Serene Wat.. (Bot.,, ser.) 2, vol: IV,)1835, p- 88! et apud) Gay,.- Hist. Fisic. y, Polit. Chile, Bot., vol VIII, 1852, p. 110; JoHow, Estud, Flora Juan Fernand., 1896, p. 198; Hve in Nouv. Archiv. du Muséum, ser. 4, vol. II], 1901, p. 84. — Sticta erythro- scypha ‘Tayl. apud Hoox in London Journ. of Bot., vol. VI, 1847, p. 181. — Sézct- ina Mougeotiana Nyl., Synops. Lich., vol. I, 1860, p. 340. — Pseudocyphellaria Mougeotiana Wain. in Hedwigia, vol. XXXVI, 1898, p. (36). Masatierra: in der Bergregion auf dem Erdboden und Baumrinden (BER- THRO! no, 1657, 16050, 1661); Cordon Centinela, 530 mi, auf der Erde; Pico Central, 360 m, auf Blocken, steril; Tres Puntas, 2—300 m, steril (C. und I. SKOTTSBERG). Masafuera: Cerro Correspondencia, bei 1350 (C. und I. SKOTTSBERG). *var. xantholoma Del. Dev. Hist. Lich. Sticta, 1822, p. 63, tab. V, fig. 14; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. III, 1890, p. 85. — Stictina Mougeotiana var. xantholoma Nyl., Synops. Lich., vol. I, 1860, p. 340; StzBer. in Flora, vol. LXXX,, 1895, pe i3t- 350 A, ZAHLBRUCKNER Masatierra: Cordén Salsipuedes, 625 m, auf dem Erdboden, in einer mehr kontrakten Form (C. und I. SKOTTSBERG). Masafuera: haiufig im unteren Teil der Heide, oberhalb der Quebrada del Mono, bei 800 m (C. und I. SKOTTSBERG). var. aurigera Nyl. Nyx. in Annal. Scienc. Nat., Bot., ser. 4, vol. XI, 1859, Pp. 254; HvE in Nouv. Archiv. du Muséum, ser. 4, vol. III, 1901, p. 85. — Sticta aurigera VDel., Hist. Lich. Sticta, 1822, p. 54, tab. III, fig. 8; Monr. in Annal. Scienc. Nat., Bot., ser. 2, vole IV, 1825,0p. So. Masatierra: Auf Felsen in den Waldern (BERTERO). *S. Weigelii Isert. IsErRt apud Acu., Lichenogr. Univers, 1810, p. 446; Wain., Etud. Lich. Brésil, vol. I, 1890, p. 189; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. III, 1901, p., 94. — Stetina quercizans' Ny)l., Synops. Lich., vol. 1,:'1860, ps) 344ei—— Stictina Wergelii Stzbgr. in Flora, vol. LXXXI, 1895, p. 133. Masatierra: Tres Puntas, 2—300 m, steril (C. und I. SKOTTSBERG). In dem obigen Sttick sind die Zyphellen sehr klein, gleichen auf den ersten Blick Pseudozyphellen, aber bei naherer Untersuchung findet man an einigen Stellen deutliche echte Zyphellen. Sect. Eusticta Hue. *S. nitida Tayl. Tayl. in Hoox., London. Journ. of Bot., vol. VI, 1847, p. 178; NyL., Synops. Lich., vol. I, 1860, p. 359; StzBer. in Flora, vol. LXXXI, 1895, p.116; A Zane. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 19. — Pseudo- cyphellaria nitida Malme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXViafd. Il. no. 6, 1890; \p. 26: Masafuera: El Barril, bei 1200 m, auf Drémys-Stammen, fruchtend (C. und I. SKOTTSBERG). S. aurata Ach. Acu., Method. Lich., 1803, p. 279; Det., Hist. Lich. Sticta, 1822, p. 49, tab. II, fig. 5; Monr. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 88; STZBGR. in Flora, vol. LXXXI, 1805, p. 118; A. Zauipr. in Sitzungsber. Akad. Wiss. Wien, math.-naturw. Cl., vol. CXI, Abt. 1, p. 495. Masatierra: Auf Baumrinden in den Bergwaldern (BERTERO no. 1658); Puerto Frances, auf Baumzweigen; Cordén Chifladores, 350 m, auf Baumrin- den; Cordén Centinela, 530 m, auf Dendroseris micrantha; Quebrada Dama- juana, auf losen Steinen im Walde; Plazoleta del Yunque; Portezuelo, 600 m, auf dem Stamme einer Rodinsonia thurifera; Quebrada seca, 455 m, auf DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 351 Dendroserts micrantha; Quebrada Juanango, 250 m, auf Baumstammen; alle Stiicke nur steril (C. und I. SKOTTSBERG). S. orygmaea Ach. Acu., Method. Lich., 1803, p. 276. — Pseudocyphellaria orygmaea Malme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXV, afd. III, no. 6, 1899, p: 28. Der Formenkreis der Sé#cfa orygmaea ist schwierig und durchaus nicht geklart. In der neueren Zeit haben HUE und MALME versucht, die Arten, bzw. Varietaten besser zu unterscheiden, HuE (Nouv. Arch. du Muséum, ser. 4, vol. III, 1901, p. 48) unterscheidet zwei Arten: Sticta endochrysea und Sticta orygmaca und findet, dass das wich- tigste Unterscheidungsmerkmal der anatomische Bau der Rinde sei. Bei Sécta orygmaca soll die Rinde der Lageroberseite und das Rezeptakulums aus per- pendicular verlaufenden, mit Seitenasten, welche anastomisieren und ein Netz- werk bilden, versehenen eingeschnirten und septierten Hyphen gebildet wer- den; bei Stécta endochrysea soll diese Rinde paraplektenchymatisch sein und mehr wenig eckige diinnwandige Zellen aufweisen, doch wird eine eingehendere Beschreibung der Rinde nur fiir die zu Stzcta endochrysea gestellten Varietaten: Urvillet, flavicans, rubescens und orygmaeoides gebracht, fir den Typus selbst als welchen die von DELISE (Hist. Lich. Sticta) auf Tafel I unter Figur 1 ge- brachte Abbildung zitiert, wird die Kinde nicht beschrieben. Ich habe an den in Juan Fernandez gesammelten Arten stets nur eine paraplektenchymatische Rinde mit eckigen und diinnwandigen Zellen gesehen, dadurch ergibt sich ein Widerspruch, den ich derzeit noch nicht aufzuklaren vermag. MALME (a. o. O.) fithrt Sz¢cta orygmaca mit den Varietaten Urvz//ez und flavicans an; zum Typus wird als Synonym Sécta endochrysea var. orygmacotdes Nyl. gestellt, aber Svcta endochrysea Del. wird nicht erwahnt. Ich bin daher dariiber nicht ganz sicher, ob MALME zur Ganze Sécfa endochrysea mit Stecta orygmaea vereinigt, halte dies aber fiir wahrscheinlich, und dann muss der Spe- ziesname des ACHARIUS als 4lterer zur Bezeichnung der Art Verwendung fin- den. Ich halte mich vorlaufig daran, betone jedoch, dass itber diesen Punkt die Akten noch nicht geschlossen sind und dass erst eine eingehende, auf dem Studium der Urstiicke beruhende Behandlung der Frage Klarung zu bringen vermag. Ich gliedere provisorisch die Art folgendermassen. Sticta orygmaea. Lobi thalli majores, superne laevigati var. endochrysea. Lobi thalli minores, superne scrobiculato-inaequales vel reticulati typus specter. Lobi magis divisi, in margine nudi var. Durvillet. Lobi in margine minute dissecta | var. flavicans. 352 A. ZAHLBRUCKNER var. endochrysea A. Zahlbr. nov. comb. Sticta endochrysea Del., Hist. Lich. Sticta, 1822, p. 43, tab. I, fig. 1; Mont. in Annal. Scienc Nat. Bot., ser. 2, vol. IV, 1835, p. 87 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 104; JoHow, Estud. Flora Juan Fernand., 1896, p. 198; Nyt. Synops. Lich., vol. I, 1860, p. 358; Srzper. in Flora, vol. LXXXI, 1895, p. 116; Hue in Nouv. Archiv. du Muséum, ser. 4, vol. IID ergom p. 48; A. ZauLBr. in Kgl. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 19. Masatierra: Auf Baumrinden (BERTERO no. 1631); El Yunque, 500 m, in den Waldern des Siidhanges, tiber Moosen fruchtend (C. und I. SKOTTS- BERG). Masafuera: im Hochlande, bei 1200 m und Heide bei Campo Correspon- dencia, 1,150 m, steril (C. und I. SKOTTSBERG). var. Durvillei A. Zahlbr. nov. comb. Sticta Durvillei Del., Hist. Lich. Sticta, 1822, p. 170. — Stcta Urorller Baw: apud Hoox., Flor. Nov. Zeland., vol. II, 1855, p. 275; Nyz., Synops. Lich., vol. I, 1860, p. 360; Stzpcr. in Flora, vol. LXXXI, 1895, p. 117. — Sticta imbricatula Tayl. in Hoox., London. Journ. of Bot., vol. VI, 1847, p. 180. — Pseudocyphellaria orygmaca var. Urvillec Malme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXV, afd. III, no. 6, 1899, p. 30. Masatierra: In den hodher gelegenen Waldern, auf Baumrinden (BER- TERO no. 1662). Cordén Salsipuedes, 615—625 m, auf Drzmys Stammen, steril; Puerto Ingles, auf Rinden, fruchtend; Villagra, 250 m, fruchtend (C. und I. SKOTTSBERG). var. flavicans A. Zahlbr. nov comb. Sticta flavicans Hook. f. et Tayl. in London. Journ. of Bot., vol. III, 1844, p. 648. — Sticta endochrysea var. flavicans Mill. Arg. in Flora, vol. LXXXI, 1888, p. 136. — Pseudocyphellaria orvgmaca var. flavicans Malme in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXV, afd. III, no. 6, 1899, p. 29. Masatierra: Quebrada Salsipuedes, bei 500 m, auf Dyrzmzys-Rinde, steril (C. und I. SKOTTSBERG). S. lineariloba Nyl. Nyt., Synops. Lich., vol. I, 1860, p. 355; StzBer. in Flora, vol..LXXXI, 1895, p. 124. — Sticta filicina Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 89 (non alior.); Croms. in Journ. Linn. Soc. London, Bot., vol. XVI, 1877, p. 223. — Sticta filicina var. lineariloba Mont. apud Gay, Hist, Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 122; JoHow, Estud. Flora Juan Fernand., 1896, p. 199. Masatierra: ohne nahere Standortsangabe, auf Baumstammen (BERTERO no. 1660; MOSELEY); Puerto Frances, bei 500 m, fruchtend; Bergriicken zwi- schen Quebrada Piedra agujereada und Quebrada Laura, 625 m, fruchtend; Portezuelo, 590 m, auf Baumrinden fruchtend; Quebrada seca, 435 m, auf Den- droseris micrantha, fruchtend; Cordén Salsipuedes, 625 m, auf Rinden fruchtend (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 353 var. hypopsila Ny]. Nyt., Synops. Lich., vol. I, 1860, p. 355; Cromp. in Journ. Linn. Soc. Lon- don, Bot., vol. XVI, 1877, p. 223; StzBeR. in Flora, vol. LXXXI, 1895, p. 124; Jonow, Estud. Flora Juan Fernand., 1896, p. 199. — Stcta filicina var. lineartloba f. Aypopsila Mont. in Annal. Scienc. Nat., Bot., ser. 3, vol. XVII rég2pp2308.-— Sticta filicina var. lineariloba Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 125. Masatierra: an nicht naher bezeichnetem Standorte (BERTERO; MOSELEY); auf dem Bergriicken zwischen Quebrada Laura und Quebrada Piedra agujereada, 625 m, auf Rinden fruchtend; Cordon Central, 390 m, in den Waldern auf Baumen, fruchtend; Abhange bei Portezuelo, 500 m, fruchtend; Cord6n Salsi- puedes, 500 m, fruchtend; am Fusse des Berges El Yunque, auf Baumrinden fruchtend (C. und I. SKOTTSBERG). Der Zuschnitt der Thalluslappen ist recht wechselnd. Ausser Individuen mit nur linearen, demmach der Beschreibung entsprechenden Lagerabschnitten, finden sich Stiicke, bei denen sich neben den normalen auch verkiirzte und ge- dringte Lappen vorfinden. Dadurch andert sich das habituelle Bild dieser Stiicke besonders dann, wenn die normalen Lappen nur vereinzelt auftreten, doch ist der Zusammenhang klar. Leicht ist nach dem Ausseren allein die Varietat mit Sticta laciniata var. denudata Ny\. zu verwechseln, vornehmlich wenn an den eingesammelten Exemplaren der Thallusstiel abgebrochen ist, indes ist es nicht schwer, beide aus einander zu halten, da bei der letzteren Art die Markschicht durch Kalilauge blutrot gefarbt wird. *S. latifrons A. Rich. A. Ricu., Voyage de Découv. de 1’Astrolabe, Bot., vol. I, 1832, p. 27, tab. VIII, fig. 2; Nvyv., Lich. Nov. Zeland., 1888, p. 33; REINKE in Prinesu., Jahrbiich. fiir wiss., Bot., vol. XXVIII, 1895, p. 441, fig. 157, Il et 158; StzBer. in Flora, vol. LXXXI, 1895, p. 125; Hue in Nouv. Archiv. du Muséum, ser. 4, vol. III, Ig01, p. 8o. Masatierra: auf dem Bergriicken zwischen Quebrada Laura und Que- brada Piedra agujereada, bei 625 m, fruchtend (C. und I. SKOTTSBERG). War bisher nur aus Neuseeland bekannt. *S. laciniata (Huds.) A. Zahlbr. var. denudata Nyl. Synops. Lich., vol. I, 1860, p. 354; Krmph. in Linnaea, vol. XLI, 1877, p. 138; Stzber. in Flora, vol. LXXXI, 1895, p. 122. Masatierra: auf dem Bergriicken zwischen Quebrada Laura und Que- brada Piedra agujereada, 625 m, auf Baumrinden fruchtend; Cordén Centinela, 530 m, auf Baumstammen steril; auf dem Bergriicken tiber Pangal, 775- 795 m, auf Rinden fruchtend; Quebrada seca, 435 m, auf Coprosma triflorum, fruchtend (C. und I, SKOTTSBERG),. 23—2391t. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. a A. ZAHLBRUCKNER 354 S. Freycinetii Del. Det., | Hist.. Lich.) Sticta:) 1822; p. 124, tab. XIV, ne. 513 Mont. apudmitGaas Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 120; Nyx., Synops. Lich., vol. I, 1860, p. 365; Sirzper. in Flora, vol. LXXXI, 1895, p. 115; Jonow, Estud. Flora Juan Fernand., 1896, p, 199. Masatierra; ohne Angabe des Standortes (JOHOW). Masafuera: auf der Hochebene, 9—1000 m, steril (C. und I. SKOTTs- BERG), var. fimbriata Mont. Mont. apud “Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIN, 1Sij2\ipesmenene Sticta variabilis Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 89. Masatierra (BERTERO): Eine zweifelhafte Form, die erst geklart wer- den muss. var. lactucaefolia Mill. Arg. Mitr. Arc. in Mission Scient. Cap Horn, vol. V, 1889, p. 157; A. ZAHLBR. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1907, p. 18. — Farmela lactucaefolia Pers. apud Gaupicu., Voyage Uranie, Bot., 18.26, p. 199. Masatierra: ohne nahere Standortsangabe (C. SKOTTSBERG). S. Richardi Mont. Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 89 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 110; Jonow,) Estuds\Mlora juan Fernand., 1896, p. 198; Hue in Nouv. Archiv. du Muséum, ser. 4, vol. III, 1901, Pp. 55. — Sticta fossulata var. Richardi Nyl., Synops. Lich., vol. I, 1860, p. 364. Masatierra: in den Bergwaldern auf Baumrinden (BERTERO); Portezuelo, 590 m, auf Rinden fruchtend; Cordon Salsipuedes, 500-625 m, auf Drzmys (C. und I. SKOTTSBERG). var. divulsa Hue. Hur in Nouv. Archiv. du. Muséum, ser. 4, vol. III, 1007, p. 56) tab. ieee 1—2..— sucla dwulsa’ Vayl. in London Journ: of Bot., vol. Vij 1847, qpaetocees Sticta fossulata f. divulsa Stzbgr. in Flora, vol. LXXXI, 1895, p. 114. Masatierra: Corddn Salsipuedes, 615 m, auf Gestrauchen (C. und I. SKOTTSBERG). Pelimoaenraceae: Nephroma Ach. A. Gonidia palmellacea vel cystococcoidea; thallus plus minus stramineus. a. Thallus major, superne foveolatus et lacunosus, inferne bullatus N. antarcticum. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 355 b. Thallus minor, utrinque laevis N. australe B. Gonidia nostocacea; thallus obscurus. a. Thallus plumbeus, subtus fuscescens et plus minus tomentellus N. plumbeum. b. Thallus castaneo-fuscus, subtus albidus et glaber J. cellulosum. Sect. Nephromium (Nyl.) Stzbgr. N. plumbeum Mont. Mont. apud. Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 100; Jonow, Estud. Flora Juan Fernand., 1896, p. 197. — Peltigera plumbea Mont. in Annal. Scienc. Nat., Bot, ser. 2, vol. IV, 1835, p. 87. — Wephromium plumbeum Nyl. in Mémoir. Soc. Scienc. Nat. Cherbourg, vol. V, 1857, p. ror et Synops. Lich., vol. I, 1860, p. 321. — Opisterta plumbea Wain. in Arkiv f6ér Bot., vol. VIII, no. 4, 1909, P- 93: Masatierra: auf Baumzweigen in den Bergwaldern (BERTERO no. 1656). N. cellulosum Ach. Acu., Lichenogr. Univers., 1810, p. 523; Bas. et Mitr. apud Hoox., Flora Tasman., vol. Il, 1860, p. 345, tab. CXCIX, fig. A. — Lzchen cellulosus Sm. apud Acu., Method. Lich., 1803, p. 289. — Aephromium cellulosum Nyl. in Mémoir. Soc. Sewne. Nat. Cherbourg, vol. V, 1857, p. 10ot et, Synops: Lich., vol. I, 1860; p. 321. — Opfisteria cellulosa Wain. in Arkiv f6r Bot., vol. VIII, no. 4, 1909, p. 93. Ohne nahere Standortsangabe (nach NYL. a. o. O). Masatierra: siidliche Abhange des Portezuelo, 600 m; Quebrada Juanango, 250 m, auf Baumstammen und auf Zweigen, fruchtend (C. und I. SKOTTSBERG). Sect. Eunephroma Stzbgr. N. antarcticum Nyl. Nye Synops, wich,- vol 1) 1860, pr 317, tab. VIE fig: 273 ue im iNeuy. Archiv. du Muséum, ser. 4, vol. II, 1900, p. 106. — Lichen antarcticus Wulf. in JAco., Miscell. Bot. Austr., vol. II, 1781, p. 370, tab. X;)fig. 1. —= Opesterza ant arctica Wain. in Arkiv fdr Bot., vol., VIII, no. 4, 1909, p. 93. — Lephroma arc- fon Mont. “apud: Gaye Hist.’ Misic: y Polit. Chile) Bor, vol Vill, 1652, "pe 111 (non alior.). Masatierra: auf Baumzweigen (BERTERO). Masafuera: Cord6én del Barril, 1,200, auf Drzmys und auf dem Siidost- abhange des Berges Los Inocentes, 900 m, auf Drimys im Decksonia-Wald (C. und I. SKOTTSBERG) Es unterliegt keinem Zweifel, dass sich die Angabe MONTAGNE's nicht auf Nephroma arcticum bezieht. Beide Arten wurden vielfach verwechselt, sie sind habituell sehr ahnlich, HUE und NYLANDER (an den oben angetihrten 356 A. ZAHLBRUCKNER Stellen) haben die Unterschiede auseinandergesetzt und legen dabei ein grosses Gewicht auf die Farbe der Lagerunterseite. Indes ist dieses Merkmal nur unter einer gewissen Einschrankung und mit Vorsicht zu benutzen. Bei Nephroma arcticum ist, von einem schmalen, hellen Rande abgesehen die La- gerunterseite dunkel; bei Mephroma antarcticum hingegen ist die Lagerunter- seite zum grdssten Teil hell, im Zentrum jedoch ebenfalls mehr oder weniger dunkel, nur fehlen an diesen dunklen Stellen die Rhizinen. Ich betrachte beide als geographische Rassen. N. australe A. Rich. A. Ricu, Woyage de Découv. de l’Astrolabe, Bot., vol. 1, 1832 i sieetaie IX, fig. 2; Nyt... Synops. Lich., vol. I, 1860, p. 318; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. Il, 1900, p. 106. — Feltigera ausiralis Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 86. — Wephromium antarcticum var. tenue Nyl., Synops. Lich., vol. I, 1860, p. 317. — Opzsteria australis Wain. in Arkiv fér Bot., vol. VIM; no: “4, 1909, (p: 93. Masatierra: In den Bergwaldern auf Baumrinden (BERTERO no. 1653, 1655). Peltigera Pers. *P. rufescens (Neck.) Humb. Masatierra: Valle de Anson am Fusse des Cordén Damajuana, auf dem Erdboden, steril; niedriger Riicken bei der Kolonie, 147 m, auf der Erde, steril (C. und I. SKOTTSBERG). P. polydactyla (Neck.) Hoffm. Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 87 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 96; Jonow, Estud. Flora Juan Fernand., 1896, p. 108. Masatierra: auf der Erde, auf Baumwurzeln und am Fusse der Baum- stamme (BERTERO no. 1650). f. pellucida Dietr. Dietr., Lichenogr. German., 1832—37, p. 27, tab. 128. — Luchen caninus, pellucidus Web., Spicil. Flor. Gé6tting., 1778, p. 270. — Leltidea polydactyla var. pellucida Ach., Method. Lich., 1803, p. 287. Masafuera: im Hochland, bei 1,200 m, auf der Erde fruchtend (C. und I. SKOTTSBERG). *var. dolichorrhiza Ny]. Ny. Synops. Lich., vol. I, 1860, p. 327; Hur in Nouv. Archiv. du Muséum, ser. 4, vol. II, 1900, p. 99. — Leltigera dolichorrhiza Nyl|., Lich. Nov. Zeland., 1888, p. 43. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 357 Masatierra: Bahia Cumberland bei 200 m, auf moosigen Steinen; Puerto Ingles, an felsigen Abstiirzen, 575 m, fruchtend (C. und I. SKOTTSBERG). Lecideaceae. Lecidea (Ach.) Th. Fr. A. Thallus crustaceus, uniformis. a. Apothecia biatorina; species corticola L. mutabilis. b. Apothecia lecideina; species saxicolae. I. Hypothecium incolor vel pallidum. 1. Apothecia in margine primum distincte albo-cincta L. leucozonata. 2. Apothecia semper nigro-marginata. 7. Paraphyses facile liberae. *_ Epithecium coerulescens L. enteroleuca. Epithecium fuscescens. §. Thallus cinerascens, Ca Cl,O, —, L. latypea. §§. Thallus flavo-viridescens, Ca Cl,O, rubens L. viridans. 6. Paraphyses conglutinatae L. leucophaca. II. Hypothecium obscurum, plus minus nigrum et carbonaceum. 1. Thallus flavidus, discus nudus. ok a. Thallus Ca Cl,O, rubens, pulvinatus L. cyanosarca. 6. Thallus Ca Cl,O, —, crustaceus L. tnactiva. 2. Thallus cinereus vel albidus; discus apotheciorum prui- nosus L. avium. B. Thallus squamulosus, hepatico-luteus Le cbterieg. Sect. Eulecidea Th. Fr. L. avium A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, tenuis. 0,z—0,3 mm _ Crassus, cinerascenti-, plumbeo- vel fulvo-cinerascens, KHO e luteo demum sordide aurantiacus vel sordide rubens, Ca Cl,0, —, opacus, effusus, areolatus, areolis fissuris tenuissimis limitatis, parvis 0,3—0,6 (0,8) mm latis, angulosis, planis, superne laevigatis, sorediis et isidiis nullis, in margine thalli protothallo nigri- cante cinctus; medulla alba, 1 —, KHO lutescens, CaCl,O, —, ex hyphis dense inspersis formata. Apothecia lecideina, adpresso-sessilia, dispersa, usque I mm lata, sed nor- maliter minora, rotunda, demum in margine leviter sinuoso-flexuosa; discus livido-, subsulphurascenti- vel caesio-pruinosus, opacus, planus vel convexiu- sculus; margo proprius niger, valde tenuis, persistens, parum prominulus, disco madefacto bene distinctus; excipulum sat angustum, extus nigrum, intus 358 A. ZAHLBRUCKNER cinereus (hyphis dense inspersis), infra hymenium leviter inflexum; hypothecium crassiusculum, decolor, ex hyphis intricatis formatum, molle; hymenium su- perne sordide fuscescens et pulverulentum, NO, non tinctum, KHO particulae pulveraceae dissolvuntur, caeterum decolor, purum, I intense violaceo-coeruleum, demum aeruginoso-sordidescens, 70—85 wp. altum; paraphyses filiformes, strictae, simplices, eseptatae, ad apicem vix latiores, conglutinatae; asci oblongo-clavati, hymenio subaequilong!, recti, ad apicem rotundati et membrana primum bene incrassata cincti, 8-spori; sporae in ascis biseriales, decolores, simplices, elli- psoideae vel ovali-ellipsoideae, rectae, parvae, 8,5—9 wu. longae et plus minus 4 », latae, membrana tenui cinctae. Conceptacula pycnoconidiorum minuta, vertice convexo, nigro, nitidulo prominula; perifulcrium dimidiatum, fusco-nigrescens; fulcra exobasidialia,; pycno- conidia filiformi-bacillaria, arcuata, rarius hamata, utrinque retusata, 14—I7 wp. longa et ad 1 wy lata. Masatierra: niedriger Felsriicken am Siidabhang von Tres Puntas und auf dem unweit davon gelegenen »Vogelgipfel», 380 m, an Felswanden (C. und I. SKOTTSBERG). Der chilenischen Lecizdea aeruginosa Nyl. ziemlich nahe stehend unter- scheidet sie sich von dieser schon durch die Farbe des Lagers. *L. inactiva A, Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, sat late expansus, versus am- bitum attenuatus, in centro crassior et ibidem usque I mm crassus; stramineo- vel isabellino-pallescens vel cremicolor, opacus, KHO e flavo subaurantiacus, Ca Cl,O, non tinctus, KHO+CaCl,O, sordide subaurantiacus, minute verru- culoso-areolatus et praeterea irrugulariter rimosus, areolis 0,2—0,s5 (0,7) mm latis, in centro thalli continuis et convexis, in ambitu thalli paulum dispersis et hypo- thallo tenuissimo, albido insidentibus, ad ipsum ambitum linea nigra tenuique cinctus, sorediis et isidiis destitutus; medulla alba, tartarea, I —. Apothecia lecideina, primum thallo arcte adpressa, demum sessilia, dispersa vel hinc inde approximata et dein subirregularia, parva, 0,3—0,6 (0,7) mm lata, planiuscula vel demum convexiuscula, nigra, opaca; discus niger, epruinosus; margo proprius primum valde tenuis, haud prominulus, integer, demum de- pressus et indistinctus; excipulum integrum, fusconigrum, infra hymenium crasse productum; hymenium superne anguste aeruginoso-nigricans vel nigricans, non inspersum, KHO —, NO, obscure coeruleum, caeterum pallide coerulescens et purum, 90-—100 ». altum, I coeruleum; paraphyses filiformes, strictae, contextae, simplices et eseptatae, ad apicem clavato-capitatae et obscuratae; asci hymenio subaequilongi, oblongo-clavati, ad apicem rotundati et membrana incrassata cincti, 8-spori; sporae in ascis biseriales, decolores, simplices, oblongo-ellipsoideae vel ellipsoideo-ovales, rectae, membrana tenul cinctae, minutae, 8,5—I0,5 u longae et 4,5—5,5 vy. latae. Pycnoconidia non visa. Masatierra: an Felswanden, Pico Central, 363 m; beim SELKIRK-Denk- mal, 590 m; Cordén Salsipuedes, 465 m und Puerto Ingles (C. und I. SKOTTS- BERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 359 *L. cyanosarca A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, crassiusculus, usque 2,5 mm altus, pulvinulos leviter convexos vel plus minus deplanatos, demum confluentes formans, isabellinus vel pallide ochraceo-stramineus, opacus, KHO_ flavens, Ca Cl,O, lateritio-ruber, sat irregulariter vel subareolatim rimosus, fissuris altis et angustis vel paulum hiantibus, in centro plus minus verruculosus, ad am- bitum passim subradians, breviter longitudinaliter torulosus, bene determinatus, sed linea obscuriore non cinctus, sorediis et isidiis destitutus; stratum corticale superficiem thalli et latera fissurarum obducens, angustum, 30— 40 yp, crassum, ex hyphis intricatis et inspersis formatum; medulla alba, crassa, KHO sordide flavens, Ca Cl,O, lateritio-rubens, ex hyphis non amylaceis, inspersis et intri- catis formata; stratum gonidiale infra corticem situm, sat crassum, continuum, gonidiis laete viridibus, globosis vel subglobosis, usque 15 y, latis. Apothecia lecideina, immersa vel adpressa, sed superficiem thalli non superantia, dispersa vel approximata, e rotundo subangulosa, nigra, opaca, parva, usque I mm lata, e plano convexiuscula; margo niger, valde tenuis et parum conspicuus, non prominulus; excipulum integrum, sat crassum, fusco- nigrum; hypothecium tenue, pallidum; hymenium superne late aeruginoso- coerulescens, NO, magis coeruleum, caeterum decolor, non inspersum, 120— 140 » altum, I intense coeruleum; paraphyses conglutinatae, filiformns, eseptatae, ad apicem clavatae; asci hymenio subaequilongi, oblongo-clavati, ad apicem rotundati et membrana paulum incrassata cincti, 8-spori; sporae in ascis bi- vel triseriales, decolores, simplices, ovales vel ovali-ellipsoideae, rectae, membrana tenui cinctae, 8,5—10 y, longae et 3,5—5 p. latae. Conceptacula pycnoconidiorum minuta, vertice convexo, nigro, nitidulo thallum paulum superantia; perifulcrium dimidiatum, nigrescens; fulcra exobasi- dialia; basidia subfiliformia, densa, pycnoconidiis vix longiora; pycnoconidia filiformia, bacillaria, recta vel subrecta, ad apices retusa, 8—9 p longa et I—I,5 » lata. Masatierra: steinige, trockene Heide bei Tres Puntas (C. und I. SKorTs- BERG). Masafuera: Cordon del Barril, bei 100 m (C. und I. SKOTTSBERG), =f, superfusa A. Zahlbr. nov. f. Apothecia tenuiter glauco- vel subsulphureo-suffusa. Masatierra: Portezuelo del Villagra, beim SELKIRK-Denkmal, 590 m, auf Felsen (C_ und I. SKOTTSBERG). *L. leucoplaca Mill. Arg. Mttiri. Arc. in Hedwigia, vol. XXXI, 1892, p. 281. Thallus crustaceus, uniformis, tartareus, tenuis, lacteus vel cinerascenti- albidus, opacus, KHO —, Ca Cl,0O, —, continuus, laevigatus, in margine linea obscuriore non cinctus, sorediis et isidiis destitutus; hyphae medullares non amylaceae. 360 A. ZAHLBRUCKNER Apothecia lecideina, sessilia, dispersa vel approximata, rotunda, atra, opaca, usque 1,5 mm _ lata, e concaviusculo demum convexula vel inaequalia, margo proprius crassus, integer, prominulus, demum depressus; excipulum integrum, inferne (sub hymenio) fere decolor, ex hyphbis tangentialibus for- matum, in parte marginali crassum, extus anguste nigricans, intus decolor, ex hyphis radiantibus, dense contextis, non septatis formatum; hypothecium an- gustum, decolor; hymenium superne obscure olivaceum, KHO —, NOs; —, caeterum decolor et purum; paraphyses filiformes. strictae, simplices, eseptatae, ad apicem vix latiores, dense contextae, in KHO autem facile liberae; asci hymenio paulum breviores, oblongo-clavati, ad apicem rotundati et membrana primum bene incrassata cincti, 8-spori; sporae in ascis subbiseriales, decolores, simplices, ellipsoideae vel ovali-ellipsoideae, membrana tenui cinctae, rectae, I2—14,5 ». longae et 5—5,5 uw latae. Masatierra: an Felsen im obersten Teile des Villagratales (C. und I. SKOTTSBERG). * L. leucozonata A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, tenuis, tartareus, cinereus, opacus, KHO —, Ca Cl,0, —, areolatus, areolis confertis, versus ambitum thalli tantum plus minus dispersis, angulosis, 0,;—1 mm latis, fissuris tenuibus limitatis, planis vel planiusculis, sorediis et isidiis nullis, hypothallus distinctus non evolutus; medulla alba, KHO —, CaCl,O, —, hyphis non amyloideis. Apothecia lecideina, sessilia, rotunda, usque 1,5 mm lata, dispersa vel approximata; discus niger, fere opacus, epruinosus, e concaviusculo planus, demum convexiusculus vel leviter gibbosus; margo primum leviter prominulus, integer, tenuis, albus, demum obscuratus et depressus; excipulum dimidiatum, crassiusculum, extus fusconigrum, intus pallidum, cinereum vel cinereo-fuscescens; hypothecium angustum, pallide lutescens, molle, ex hyphis intricatis, non inspersis formatum; hymenium superne umbrino- nigricans, KHO —, NO; —, non inspersum, caeterum fere decolor et in parte inferiore pallide aeruginoso- coerulescens, 60—80 y, altum, I intense coeruleum; paraphyses filiformes, strictae, simplices, eseptatae, ad apicem clavatae et obscuratae, conglutinatae; asci oblongo-clavati, hymenio subaequilongi, 8-spori; sporae in ascis biseriales, decolores, simplices, late ellipsoideae vel ovales, rectae, membrana tenui cinctae, minutae, 8.5—11 p, longae et 4—6 y, latae. Pycnoconidia non visa. Masafuera: Cerro Correspondencia, bei 1,400 m, auf vulkanischem Ge- stein (C. und I. SKOTTSBERG). * L. enteroleuca Ny]. Nyt. in Flora, vol. LXIV, 1881, p. 187; Arn. in Flora, vol. LXVII, 1884, p. 558 et Zur Lich., Flora Miinchen in Bericht Bayr. Bot. Gesellsch., vol. I, Anhang, 169m p. $0. — Hepp, Flecht. Burop., no: 129: Masatierra: Pangal, auf Strandfelsen, und Steilwand unweit des Porte- zuelopasses, 600 m (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 365 Masafuera: im untersten Teil der Quebrada de las Casas, auf vulka- nischem Gestein (C. und I. SKOTTSBERG). *L. latypea Ach. Acu., Method. Lich., 1803, suppl., p. 10; ARN. in Flora, vol. LXXVII, 1884, p- 562; Warn. in Arkiv f6r Bot., vol. VIII, no. 4, 1909, p. 133. Masatierra: auf Blocken zwischen der Kolonie und Pangal (C. und I. SKOTTSBERG). An den gesammelten Stiicken sind die Apothezien etwas kleiner und etwas mehr gewolbt als in den europdischen Exemplaren, iibrigens aber ganz ibereinstimmend. *L. viridans Lamy. Lamy in Bull. Soc. Bot. France, vol. XXV, 1878, p. 446; ARN. in Flora, vol. LXXVII, 1884, p 559; A. L. Smit, Monogr. Brit. Lich., vol. II, 1911, p. 55. — Lectdella viridans K6rb., Syst. Lich. German., 1855, p. 242 et Parerg. Lich., 1861, p. 213. — Sratore viridans Hepp, Flecht. Europ. no. 726. Masatierra: Portezuelo de Villagra, 590 m, auf vulkanischem Gestein (C. und I. SKOTTSBERG). Sect. Biatora (Ach.) Br. et Rostr. L. mutabilis Fée. Fér, Suppl. Essai Crypt. Ecorc. Officin., 1837, p. 105; tab. XLII, fig. 16; Nyt. in Acta Soc. Scient. Fennic., vol. VII, 1863, p. 495; Fixx in Contrib U. S. Nation. Herbarium, vol. XIV, 1910, p. 71; A. L. SmirxH, Monogr. Brit. Lich., vol. II, 1911, p. 43. — Bvatora vernalis var. varians Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 93. — Statora mutabilis Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 171; JoHow, Estud. Flora Juan Fernand., 1896, p. 201. Masatierra: auf Baumrinden (BERTERO no. 1615). Sect. Psora (Hoffm.) Schaer. “Letetericaslayl: Tay. in Hoox., London Journ. of Bot., vol. VI, 1847, p. 150; Nyt in Annal. Scienc. Nat., Bot., ser. 4, vol. XV, 1861, p. 360. — Siatora icterica Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. II, 1834, p. 373 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 170, tab. XII, fig. 4. — Lecedea endochlora Tayl. in. Hoox., London Journ. of Bot., vol. VI, 1847, p. 151. — Lecanora Wrightze Tuck. im Americ. Journ. Arts and Scienc., ser. 2, vol. XXV, 1858, p. 425. — Biatora Wrighttt Tuck., Lich. Californ., 1866, p. 31, not. — sora icterica Mill. Arg. in Flora, vol. LXXI, 1888, p. 45; Finx in Contrib. U. N. Nation. Her- bar., vol. XIV, 1910, p. 103: 262 A. ZAHLBRUCKNER Masafuera: Playa Ancha, am Meeresstrand auf humoser Erde (C. und I. SKOTTSBERG). Catillaria (Ach.) Th. Fr. A. Apothecia pallide marginata, habitu lecanarino. a. Apothecia nigra; sporae oblongae vel subfusiformes, 14—30 p. longae et 5—7 wu. latae C. endochroma. b. Apothecia pallida, ceracea, livida vel livido-nigricantia; sporae ovales vel ovali-ellipsoideae, 17—20 ». longae et 8—g y, latae — C. leucochlora. B. Apothecia nigro-marginata, typice lecideina. a. Hypothecium obscuratum; apothecia mediocria. I. Hypothecium rubrofuscum; sporae 15— 19 ». longae et 6—8,5 u. latae C. intermixta. Il. Hypothecium coerulescenti-nigricans, in parte marginali strato albo praeditum; sporae 25—30 uw. longae et I12—16y, crassae C. melastegia f. mesoleucodes. b. Hypothecium decolor; apothecia minuta, convexa et immarginata C. theobromina. *=C. intermixta Arn. Arn. apud Gtowacki in Verhandl. zool.-bot. Gesellsch., Wien, vol. XX, 1870, p- 455. — Lecidea intermixta Nyl. in Annal. Scienc. Nat., Bot., ser. 4, vol. III, 1855, p. 161 et Lich. Scandin, 1861, p. 194; Harm. in Bull. Soc. Scienc. Nancy, ser. 2, vol. XXXIII, (1898) 1899, p. 59, tab. XXII, fig..54. — Cetilaria Laureri Hepp apud Arvn., Lich. exsicc. no. 353 (1867); TH. Fr., Lichenogr. Scand., vol. I, 1874, p. 582. — Buatorina imtermixta Kieffer in Bull. Soc. Hist. Nat. Metz, 1895, p-184; 7A. 1s. Smira, “Monogr. Brit: Michs>vol- IL* 19o1r,-p. £25. Masatierra: Cordon Chifladores, 350 m, auf der Rinde einer Rodznsonia thurifera; Cordén Centinela, 530, auf Stammen der Dendroseris micrantha; Valle Colonial, Quebrada seca, 435 m, auf Dendroseris micrantha und Fagara (C. und I. SKOTTSBERG). *C. melastegia (Nyl.) A. Zahlbr. f. mesoleucodes A. Zahlbr. nov. comb. Lecidea melastegia f. mesoleucodes Ny\., Lich. Fueg. et Patagon., 1888, p. 15. — Catillaria grossa var. mesoleucodes A. Zahlbr. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917. p. 22. Masatierra: auf dem Bergriicken iiber Pangal, 400 m, auf Coprosma py- rifolium (C. und I. SKOTTSBERG). Masafuera: Quebrada del Blindado, 446 m, auf der Rinde der Dendroseris giganiea und Quebrada de la Loberia, 300 m, auf dlterer Lumarinde (C. und I. SKOTTSBERG). Excipulum crassiusculum, integrum, coeruleo-nigrum, in parte marginali intus anguste decolor; hymenium superne anguste coeruleo-nigrum, caeterum decolor, aquoso-pellucidum, purum, 180—200 yp, altum, I e coerulescente mox DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 363 cupreo-rufidulum; paraphyses capillares, strictae, simplices, eseptatae, conglu- tinatae; asci hymenio subaequilongi, oblongo~clavati, superne rotundati et mem- brana incrassata cincti, 8-spori; sporae in ascis biseriales, decolores, ovales vel ovali-ellipsoideae, utrinque rotundatae, rectae, uniseptatae, septo et membrana tenui, ad septum non constrictae, 26—30 p. longae et 12—15 yp. latae. *C, endochroma A. Zahlbr. A. ZAucpr. in ENGLERR-PRANTL, Natiirl. Pflanzenfam., r. ‘Teil, Abt. 1*, 1995, p. 134. — Lecanora endochroma Feée, Essai . Crypt. Ecorc. Officin., 1824, p. 114, tab. XXIX, fig. 1. — Lecidea endochroma Nyl. in Flora, vol. XLI, 1858, p. 380 et in Bull. Soc. Linn. Normand., ser. 2, vol. III, 1869, p. 271. — Psorothecium endo- chromum Mass. in Atti I. R. Istit. Veneto, ser. 3, vol. V, 1860, p. 261; MULL. ARc. in Revue Mycol., vol. IX, 1887, p. 89. — Heterothectum endochromum Fw. in Bot. Zeitung, vol. VIII, 1850, p. 555; TucKx., Synops. North Americ. Lich., vol. I, 1888, p. 55. Masatierra: Quebrada de la Choza, 250 m, auf Baumrinden im Walde (C. und I. SKOTTSBERG). C. leucochlora A. Zahlbr. nov. comb. Parmelia leucochlora Mont.! in Annal. Scienc. Nat., Bot. ser. 3, vol. XVIII, 1852, p. 310 et apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIL, 1852, p. 152. — Parmelia varia ¥r. var. leucochlora Mont. in Annal. Scienc. Nat., Bot., ser. 2. vol. IV, 1835, p. 91. —- Bérengeria leucochlora ‘Vrevis., Spighe e Paglie, 1853, p. 5. — Lecidea leucochlora Nyl. in Annal. Scienc. Nat., Bot., ser. 4, vol. IJ], 1855, p. 184; Hue in Nouv. Archiv. du Muséum, ser. 3, vol. II], 1891, p. 107. — Bayrhofferia leucochlora Trevis. in Rivist. Period. Lavori Accad. Padova, vol. V, 1857, p. 69. Thallus epiphloeodes, crustaceus, uniformis, sat expansus, tenuis, albidus vel dilute alutaceus, fere opacus, KHO flavens, Ca Cl,O, —, continuus vel sub- strato rupto passim irregulariter fissus, utplurimum sat laevigatus vel subpul- veraceo-inaequalis, rarius minute granulosus, sorediis et isidiis non praeditus, in margine hinc inde linea tenui nigricante cinctus. Apothecia pseudolecanorina, sessilia, ad basin leviter constricta, dispersa vel plus minus approximata, rotunda vel rotundata, usque 2 mm lata, e con- cavo convexiuscula vel convexa, cerina vel livida, demum nigricantia, caesio- pruinosa vel subnuda; margo proprius primum bene prominens, integer, disco pallidior, dein flexuosus et demum depressus; excipulum crassum, chondroideum, dimidiatum, in sectione late obtriangulare, decolor vel in margine tenuiter in- fuscatum, ex hyphis radiantibus, strictis, tenuibus, arcte conglutinatis, eseptatis formatum, I vix lutescens, gonidia nulla includens; hypothecium sat crassum, fere decolor, ex hyphis dense intricatis formatum, I e coeruleo aeruginoso- obscuratum; hymenium superne anguste sordide fuscidulum et pulverulento- inspersum, caeterum decolor et purum, 120-130 y altum, I primum intense coeruleum, demum in sordide cupreum vergens; paraphyses filiformes. con- glutinatae, simplices, eseptatae, ad apicem non latiores; asci hymenio subae- quilongi, ovali-clavati, superne rotundati et membrana sat bene incrassata cincti, 8-spori; sporae in ascis bi- vel triseriales, decolores, ovali-ellipsoideae, curvulae, 364 A. ZAHLBRUCKNER uniseptatae, septo tenui, ad septum non constrictae, membrana tenui cinctae, 17—20 ». longae et 8—g yp. latae. Pycnoconidia non visa. f. typica A. Zahlbr. nov. f. Thallus granulosus, alutaceo-cinerascens. Masatierra; auf Rinde (BERTERO no. 1616). Masafuera: Hochplateau zwischen Quebrada Sanchez und Ensenada Toltén, 515 m, auf Baumrinden (C. und I. SKOTTSBERG). Diese Form entspricht dem MONTAGNE'schen Urstiick. f. laevigata A. Zahlbr. nov. f. Thallus laevigatus, granulis nullis, albidus. Masatierra: auf dem Riicken des Pangal, 400 m, auf Coprosma-Rinde; Portezuelo, beim SELKIRK-Denkmal, 590 m, auf Aodzusonza-Rinde; Quebrada Salsipuedes, 625 m, auf Rinden (C. und I. SKOTTSBERG). *C. theobromina A. Zahlbr. nov. spec. Thallus epilithicus, crustaceus, uniformis, substratum tenuissime obducens, late expansus, theobrominus vel passim theobromino-cinerascens, opacus, KHO —, Ca Cl,0, —, continuus, leviter inaequalis, sorediis et isidiis nullis, in margine linea obscuriore non cinctus. Apothecia sat copiosa, plus minus dispersa, nigra, opaca, lecideina, ses- silia, rotunda, mox convexa et demum fere semiglobosa, parva, 0,3—0.4 (— 0,5) mm lata; margo proprius in juventute apotheciorum valde tenuis, integer et vix prominulus, mox depressus; excipulum dimidiatum, fusconigrum, ad basin ver- sus hymenium productum, 18—28 uw, crassum; hypothecium fere decolor, molle, ex hyphis intricatis formatum; hymenium superne coceruleo-nigricans, NOs in- tense coeruleum, non inspersum, in centro et in parte marginali dilute coeru- lescens, caeterum decolor, purum, 34—4o vu. altum; paraphyses filiformi capillares, strictae, conglutinatae, simplices, eseptatae, ad apicem capitatae et obscuratae; asci hymenio subaequilongi, oblongo-clavati, ad apicem rotundati et membrana primum bene incrassata cincti, 8-spori; sporae in ascis biseriales, decolores, oblongae vel ellipsoideo-oblongae, utrinque rotundatae, rectae, uniseptatae, septo et membrana tenui, parvae, 7—8 ». longae et ad 2 », latae. Pycnoconidia non visa. Masatierra: Portezuelo de Villagra, 590 m, auf vulkanischem Gestein (C. und I. SKOTYTSBERG). Megalospora Mey. et Fw. *M. versicolor A. Zahlbr. A. Zantspr. in ENGLER PRantt, Natiirl. Pflanzenfamil., 1 Teil, Abt. 1*, 1905, p. 134; RippLe in Bull. Torrey Bot. Club, vol. XLIV, 1917, p. 327, tab. XXI, fig. 9. — DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 365 Lecanora versicolor Fée, Essai Crypt. Ecorc. Officin., 1824, p. 115, tab. XXVIII, fig. 4. — Lecidea versicolor Nyl. in Acta Soc. Scient. Fennic., vol. VII, 1863, p. 461. — Feterothecium versicolor Fw. in Bot. Zeitung. vol. VIII, 1850, p. 555; Tuck., Synops. North Americ Lich., vol. II, 1888, p. 54. — VPatellaria versicolor Miill. Arg. in Flora, vol. LXV, 1882, p. 330. — Bvatorina versicolor Hellb. in Bihang till Kgl. Svensk. Vetensk.-Akad. Handl., vol. XXI, afd. II, no. 13, 1896, p. 109. var. microcarpa A. Zahlbr. nov. var. Thallus tenuis, albidus, continuus vel imprimis versus ambitum discon- tinuus et hypothallo cinerascente vel umbrino insidens. Apothecia nigra, nuda, minuta, 0,2—0,3 mm lata, mox convexa, margine haud conspicuo; sporae 27—38 wp. longae et 17—20 y. latae, in ascis binae. Masatierra: Quebrada Salsipuedes, 615 m, auf Baumzweigen (C. und I. SKOTTSBERG). Bacidia (De Not.) A. Zahlbr. A. Apothecia saltem primum distincte pallide marginata. I. Hypothallus distinctus deest B. delapsans. II. Hypothallus distinctus, niger B. subluteota. B. Margo apotheciorum disco concolor. I. Apothecia nigra B. endoleuca. Il. Apothecia plus minus fusca. a. Sporae crassiores; hymenium superne olivaceo-coerulescens vel nigro-violaceum B. endoleuca var. laurocerasz. b. Sporae graciles, subfiliformes; hymenium superne fuscescens vel subdecolor B. arceutina var. hyposcotina =B. endoleuca Kickx. Kicxz, Flore Crypt. Flandr., vol. I, 1857, p. 261; Tu. Fr., Lichenogr. Scand., vol. I, 1874, p. 347; Watn. in Acta Soc. Scient. Fennic., vol. LIII, no. 1, 1922, p. 148 et 177- — Biatora luteola f. endoleuca Nyl\. in Bot. Notiser, 1853, p. 98. Masatierra: Plazoleta del Yunque, 235 m, auf Avzstote/ia-Rinde (C. und I, SKOTTSBERG). Masafuera: Quebrada Blindado, 440 m, auf der glatten Rinde von Den- droserts gigantea (C. und I. SKOTTSBERG). *var. laurocerasi Arn. Arn. in Flora, vol. LIII, 1870, p. 472; Oxiv., Expos. Lich. Ouest France, vol. II, 1900, p. 19. — VPatellaria laurocerasi Del. apud Duby, Botanic. Gallic., vol. II, 1830, p. 652. — Bacidia laurocerasi Wain. in Acta Soc. Scient. Fennic., Moll, no. 1; 1922, Ps 147 eb 17 5- Masatierra: Plazoleta del Yunque, 235 m, auf A7zstofelza-Rinde (C. und I. SKOTTSBERG). 366 A. ZAHLBRUCKNER Masafuera: Hochebene zwischen Quebrada Sanchez und Ensenada Toltén, 515 m, auf Lumarinde (C. und I. SKOTTSBERG). b *B. arceutina (Ach.) Arn. var. hyposcotina A. Zahlbr. nov. var. Apothecia usque I mm lata, primum rubricoso-fusca, demum obscurata, nigricantia; hypothecium crassiusculum, rufofuscum; hymenium superne dilute fuscescens vel fere decolor, non inspersum, caeterum decolor, 50—70 . altum, I praecedente coerulescentia mox cupreo-fuscum; paraphyses filiformes, sim- plices, eseptatae, ad apicem paulum latiores, conglutinatae; asci hymenio subae- quilongi, anguste clavati, ad apicem rotundati et membrana modice incrassata cincti, 8-spori; sporae decolores, filiformi-capillares, subrectae, in uno apice le- viter latiores, subindistincte pluriseptatae, 42— 50 u. longae et ad 1 yw. latae. Masatierra: Cordén Chifladores, 350 m, auf dem Stamme einer Kodzn- sonia thurifera (C. und I. SKOTTSBERG). Habituell gleicht diese neue Varietat der Bacidia fuscorubclla (Hoffm.), wie sie bei HEpP unter no. 520 ausgegeben ist; die grazilen Sporen zeigen indes, dass sie mit ihr nichts zu tun hat. Das Hypothezium ist schon in den jungsten Friichten dunkel gefarbt und kraftig entwickelt. *B. delapsans A. Zahlbr. nov. spec. Thallus epiphloeodes, tenuissimus, substratum arcte obducens, maculas formans irregulares, sat extenuatas, margine linea nigra cinctas et demum plus minus confluentes, lutoso-cinerascens, opacus, KHO lutescens, Ca C]l,O, —, pri- mum pulverulento~leprosulus, demum leprosus, continuus, sorediis et isidiis non praeditus. Apothecia lecideina, sessilia, mox elabentia et foveolas rotundas relin- quentia, normaliter dispersa, rotunda, minuta, 0.5—O.8 mm lata, e plano mox convexa, umbrino nigricantia, opaca, madefacta magis umbrina; margo proprius primum valde tenuis, integer, cinerascens, mox depressus; excipulum ad am- bitum nigricans, intus pallidum, ex hyphis dense intricatis, elementa substrati includentibus formatum, I —; hypothecium crassum, fusconigrum, versus am- bitum rufescens, KHO vix mutatum; hymenium angustum, 45—55 uw. altum, superne anguste umbrino~fuscum, KHO violaceum, caeterum maculatim dilute fuscescens vel decolor, purum, I e coeruleo mox aeruginoso-obscuratum; pa- raphyses filiformes, simplices, eseptatae, ad apicem non latiores; asci crebri, oblongo-clavati, ad apicem rotundati et membrana bene incrassata cincti, 8- spori; sporae in ascis subtriseriales, verticales, decolores, aciculari-fusiformes, subrectae vel curvulae, in uno apice rotundatae vel acutatae, in altero apice normaliter fusiformi-acutatae, subindistincte pluriseptatae, 28—45 u longae et I,5—I.8 v. latae. Conceptacula pycnoconidiorum punctiformia, nigra, concavo-emersa; peri- fulcrium superne obscuratum, caeterum pallidum; fulcra exobasidialia; pycno- conidia oblonga, minuta, I.s—2 wv, longa. Masatierra: Quebrada Gutierrez, und Quebrada seca, auf Rinde von Fuanta australis, 435 m (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 367 Das Excipulum ist eigentlich dem anatomischen Bau nach lekanorinisch, indem es aus verwebten Hyphen, welche jenen der Markschicht ahnlich sind, gebildet wird. Es scheint sich auch unter das Hypothezium zu erstrecken, doch ist die Begrenzung schwer wahrzunehmen, da die Hyphen beider Schichten wenig verschieden sind. Der Fruchtrand ist demnach hier ahnlich gebaut, wie bei manchen Arten der Gattung Lecedea sect. Lulecidea, z. B. bei Lecidea po- lycarpa F\ik., wo das Verhaltnis nur insoferne deutlicher hervortritt, als alle Markhyphen durch Kalilauge blutrot gefarbt werden und diese Farbung sich auch auf die Hyphen des Excipulums erstreckt. *B. subluteola A. Zahlbr. A. ZaAHLBR. in Sitzungsber. Akad. Wiss. Wien, math.-naturw, Cl., vol. CXI, Abt. 1, 1902, p. 395 et in Denkschrift. math.-naturw Cl. Akad. Wiss. Wien, vol. LXXXIII, 1909, p. 129. — Lecidea subluteola Ny\. in Flora, vol. LU, 1869, p. 122. Latellaria subluteola Mill. Arg. in Flora, vol. LXIV, 1881, p. 522. Masafuera: Quebrada de la Loberia, 300 m, auf Lumarinde (C. und I. SKOTTSBERG). Toninia Th. Fr. *T. (sect. Eutoninia) bullata A. Zahlbr. A. Zauipr. in Beihefte zum Bot. Centralbl., vol. XIX, Abt. II, 1905, p. jo. — Lecidea bullata Mey. et Fw. in Nova Acta. Acad. Caes. Leopold.-Carolin., vol. XIX, suppl. 1, 1843, p. 227. Apothecia primum subpedicellata, cupularia, demum applanata, usque 3 mm lata, rotunda; discus niger, nudus; margo primum crassiusculus, obtusus, inflexus, niger, demum angustatus, minus prominulus et leviter crenulatus; receptaculum subtus primum nigricans et verruculosum, evolutum pallide ru- fescens, verruculosum, sed verruculae minus prominentes, extus et inferne cras- siuscule corticatum, cortice chondroideo, lutescente, ex hyphis ramosis et dense contextis formato; hypothecium angustum, obscure fuscum vel fere nigrum, ex hyphis perpendicularibus formatum; hymenium 70- gO u. crassum, superne anguste nigricans, non inspersum, KHO —, caeterum decolor et purum, I pri- mum coerulescens, tandem sordide aeruginosum; paraphyses filiformes, ad 2 vu. crassae, strictae, simplices, primum facile liberae, ad apicem clavatae et obscu- ratae, clavo 3,5—4 y. lato; asci clavati, ad apicem rotundati, hymenio subae- quilongi, recti, membrana in parte apicali primum valde incrassata cincti, 8- spori; sporae in ascis 3—4 seriales, decolores, verticales, subcylindricae, in uno apice paulum latiores, utrinque rotundatae, subrectae vel curvulae, 8—10-locu- lares, septis pertenuibus, 38—52 v. longae et 3,5—4 y. latae. Masafuera: iiber abgestorbenen Moosen beim Correspondencialager, 1,150 m (C. und I. SKOTTSBERG). Ich hatte bisher die Flechte bevor ich die Apothezien kennen lernte wegen der Gestalt des Lagers bei der sect. Yhalloidima untergebracht; nunmehr 368 A. ZAHLBRUCKNER unterliegt es keinem Zweifel, dass sie zur sect. Eudoninza gehort. Der Thallus wurde von MEYER und FLOTOW zutreffend beschrieben, ich fitige erganzungs- weise nur hinzu, dass die Lagerwarzen bis 3 mm breit werden, bald zerstreut stehen, bald aneinander geriickt sind. Lopadium Korb. *L. leucoxanthum A. Zahlbr. A. ZAHLBR. in Sitzungsber. Akad. Wiss. Wien, math.-naturw. Cl., vol. CX], Abt. 1, 1902, p. 398. — Lecedea leucoxantha Sprgl. in Kgl. Vetensk.-Akad. Handl., 1820, p. 46. var. albidius A. Zahlbr. nov. var. Thallus lacteus vel albus. Apothecia usque 2 mm lata, ad basin constricta, basi angusta elato-sessilia; discus croceus vel vitellino-subcroceus; margo proprius crassiusculus, integer, albus vel lutescenti-albidus. Masatierra: Cordén Chifladores, 350 m, auf Robinsonien; Quebrada seca, 435 m, auf der Rinde einer Dendroseris micrantha; Quebrada Salsipuedes, auf Rinden; Quebrada de la Choza, 250 m, auf Baumrinden (e und I. SKOTTS- BERG). Rhizocarpon DC. A. Thallus flavus R. geographicum. B. Thallus cinereus vel albidus. Sporae 13—16 ». longae et 8,5--9 p. latae R. microspernum. Sporae 22—30u. longae et 10-14. latae A. obscuratum var. demimutum. *R. geographicum (L.) DC. f. contiguum (Schaer.) Mass. Masafuera: auf einem vulkanischen Block im oberen Teil der Quebrada de las Vacas und Strandfelsen in der Quebrada del Mono (C. und I. SKOTTs- BERG). *R. (sect. Eurhizocarpon) microspermum A. Zahlbr. nov. spec. Thallus epilithicus, tenuis, crustaceus, uniformis, tartareus, plumbeo- vel cinereo-albidus, opacus, KHO —, Ca Cl,0, —, rimoso-areolatus, areolis parvis, O.2z—1 mm latis, subangulosis, contiguis, planis, fissuris angustis et plus minus flexuosis separatis, superne laevigatus, sorediis et isidiis destitutus, in margine linea obscuriore distincta non cinctus; medulla alba, ex hyphis non amyloideis formata. Apothecia lecideina, primum inter areolas thalli sedentia, mox elavata et adpresse sessilia, nigra, opaca, rotunda, usque I mm lata, dispersa vel approxi- mata, e plano leviter convexa, non umbonata; margo primum leviter promi- nulus, integer angustusque, demum depressus; excipulum nigrum, integrum, olivaceo-fuscum, infra hymenium sat crassum, ad latera hymenii nigrum; hyme- DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 369 nium superne obscure fuscum, KHO haud mutatum, non pulverulentum, cae- terum decolor, purum, 130—150 y. altum, I intense coeruleum; paraphyses strictiusculae, subsimplices, gelatinose conglutinatae, eseptatae; asci hymenio subaequilongi, oblongo-clavati, ad apicem rotundati, 8-spori; sporae in ascis 2—3 seriales, decolores, ellipsoideae vel ovales, depauperato-murales, septis hori- zontalibus 3—4, septo verticali utplurimum unico, halone non circumdatae, parvae, 13—16 yw. longae et 8,5—9 y. latae. Pycnoconidia non visa. Masafuera: Heide des Hochplateaus, 1000—1200 m, auf vulkanischem Gestein (C. und I. SKOTTSBERG). Gehort in den Formenkreis des Rhzzocarpon obscuratum (Ach.) Mass.; die kleinen Sporen und die flachen Areolen des hellen Thallus kennzeichnen die neue Art. *Rh. obscuratum (Ach.) Mass. var. deminutum A. Zahlbr. nov. var. Thallus cinereo-caesius, minute areolatus, areolis ad 0,15 mm latis. Apothecia minuta, O2—o,3 mm lata, mox convexiuscula; sporae ut in typo. Masafuera: Cordon del Barril, auf vulkanischem Gestein (C. und I. SKOTTSBERG). Phyllopsoraceae. Phyllopsora Mill. Arg. Ph. parvifolia Mill. Arg. Mii. Arc. in Bull. Herbier Boissier, vol. II, 1894, p. 90. — Lecadea parvi- folta Pers. apud Gaupicu., Voyage Uranic, Bot., 1826, p. 192 — Farmelia parvi- folia Mont. apud Sacra, Hist. de d’Ile Cuba, Botan., 1838—42, p. 214, tab. X, meaeer apud Gay, Hist-< Fisic)..y Polit:, Chile, Bot-,, vol) Vil n852)'p 145; Jonow, Estud. Flora Juan Fernand., 1896, p. 209. — Svatora parvifolia Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835, p. 92. — Lecanora bibula ‘Yayl. in Hoox., London Journ. of Bot., vol. VI, 1847, p. 160. Masatierra; auf Baumrinden in den hoher gelegenen Bergwaldern (BER- TERO no. 1648). Cladoniaceae. Baeomyces Pers. B. chilensis Cromb. Croms. in Journ. Linn. Soc London, Bot., vol. XVI, 1876, p. 223; Jonow, Estud. Flora Juan Fernand., 1896, p. 201. — Jvatora byssoides var. chilensis Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p..173. — Bacomyces rufus var. chilensis Nyl., Synops. Lich., vol. I, 1860, p. 177. Masatierra: ohne nahere Standortsangabe (MOSELEY). 24—2391. The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 3709 A. ZAHLBRUCKNER Masafuera: an Steinen und Felsen an einem Bache unweit des Correspon- dencialagers bei 1100 m (C. und I. SKOTTSBERG). Cladonia (Hill.) Wain. C. alpestris (L.) Rabh. Wain., Monogr. Cladon., vol. I, 1887, p. 41. — Cladonia rangiferina var. al- pestris Schaer.: Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p- 166; Jonow, Estud. Flora Juan Fernand., 1896, p. 196. Masatierra: zwischen Moosen auf den hoheren Bergen (BERTERO no. 1605). *C. pycenoclada (Gaud.) Nyl. var. flavida Wain. Wain., Monogr. Cladon., vol. I, 1887, p. 38. Masatierra: Cordon Salsipuedes, 625, an sonnigen und steinigen Stellen, steril (C. und I. SKOTTSBERG). Masafuera: unter dem Gipfel von Los Inocentes, steril (C. und I. SKOTTSBERG). C. bacillaris (S. Gray) Nyl. Wain., Monogr. Cladon., vol. I, 1887, p. 88; A. ZauiBr. in Kgl. Svensk. Ve- tensk.-Akad> Hand; «vol; Vil, no; 6; 1917, 1p. 24: Masatierra: Bahia Cumberland, auf morschen Baumstriinken (C. SKOTTS- BERG). *C. didyma Wain. var. vulcanica (Zoll.) Wain. Wain., Monogr. Cladon., vol. I, 1887, p. 145. Masatierra: auf dem Bergriicken tiber Pangal, 795 m, auf dem Erdboden, fruchtend (C. und I. SKOTTSBERG). C. coccifera (L.) Willd. var. stemmatina Ach. Wain., Monogr. Cladon., vol. I, 1887, p. 158; A. ZaHtpr. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 24. — Cladonia cornucopioides Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 163; JoHow, Estud. Flora Juan Fernand., 1896, p. 197. Ohne nahere Standortsangabe (CUMING). Masatierra: Cordén Salsipuedes, 500—600 m, auf der Erde, fruchtend (C. und I. SKOTTSBERG). Masafuera: Heide oberhalb Quebrada del Mono, 850 m, schén fruchtend, darunter einige Podezien mit Bechern, welche in ihrer Mitte proliferieren (= f. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 371 asotea Ach.); auf dem Hochplateau, bei 1200 m; Heide unweit des Campo Correspondencia, 1130 m, auf der Erde, fruchtend (C. und I. SKOTTSBERG),. C. aggregata (Sw.) Ach. Wain., Monogr. Cladon., vol. I, 1887, p. 224; Mont. apud Gay, Hist. Fisic. y Polit. Chile, Bot., vol. VIII, 1852, p. 1673; Jonow, Estud. Flora Juan Fernand., 1896, p. 196; A. ZAuLBR. in Kgl. Svensk.-Vetensk.-Akad. Handl., vol. LVII, no. 6, 1917, p. 25. — Cladia ageregata Nyl.; Croms. in Journ. Linn. Soc. London, Bot., Wolo el, 1877; ps 223. Masatierra: ohne nahere Standortsangabe (BERTERO, MOSELEY); Puerto Frances, Loma del Incienso; Portezuelo del Villagra, auf losem Tuff und auf der Erde, steril (C. und I. SKOTTSBERG). Masafuera: oberhalb Quebrada del Mono, 850 m, auf dem Hochplateau, 1000—1350 m und beim Campo Correspondencia, 1130 m, steril (C. und I. SKOTTSBERG). var. straminea Mill. Arg. Wain., Monogr. Cladon., vol. I, 1887, p. 230. Masatierra: Am Weg nach Portezuelo, 200 m, auf dem Erdboden, steril (C. und I. SKOTTSBERG). *Cl. furcata (Huds.) Schrad. f. foliosa Del. Watn., Monogr. Cladon., vol. I, 1887, p. 333. Masatierra: Cordén Salsipuedes, 615 m, in Buschwerk zwischen Moosen auf der Erde (C. und I. SKOTTSBERG). Cl. gracilis (L.) Willd. var. chordalis (Flk.) Schaer. Wain., Monogr. Cladon., vol. II, 1894, p. 97. Masatierra: Portezuelo, bei 600 m, zwischen Moosen auf der Erde (C. und I. SKOTTSBERG). Masafuera: auf dem Hochplateau, 750—1200 m, auf dem Erdboden, fruchtend (C. und I. SKOTTSBERG). var. tenera Nyl. Wain., Monogr. Cladon., vol. II, 1894, p. 109. Ohne nahere Standortsangabe (BERTERO). var. aspera Flk. Watn., Monogr. Cladon., vol. II, 1894, p. 110. 2g 7 A. ZAHLBRUCKNER Masafuera: Heide unweit des Campo Correspondencia, 1100—1350 m, auf dem Erdboden (C. und I. SKOTTSBERG). var. elongata (Jacqu.) Flk. Wain., Monogr. Cladon., vol. II, 1894, p. 116; Mont. in Annal. Scienc, Nat., Both ser= 25vol. eV 2 1os5, up. o2: Ohne Standortsangabe (CUMING). Masatierra: Portezuelo, auf den hoheren Bergen auf der Erde (BERTERO no. 1606). Cl. pyxidata (L.) Fr. var chlorophaea Fk. Wain., Monogr. Cladon., vol. II, 1894, p. 232. Ohne Standortsangabe (CUMING). Masafuera: Heide unweit Correspondencia, bei 1350 m, auf der Erde, fruchtend (C. und I. SKOTTSBERG). C. fimbriata (L) Fr. var. prolifera (Retz.) Mass. Wain , Monogr. Cladon., vol. II, 1894, p. 271. — Cladonia fimbriata var. ra- diata. Mont. in> Annal, sScienc.. Nat., “Bot., ser. 2, vol: LV, S35, p; 92K): Masatierra: Auf steinigen Htigeln (BERTERO no. 1607). var. Balfourii (Crb.) Wain. Wain., Monogr.. Cladon., vol. II, 1894, p. 339. Masafuera: Quebrada del Mono, an der Basis eines Decksonta-Stammes (C. und I. SKOTTSBERG). Cl. pityrea (Flk.) Fr. I. Zwackhii f. scyphifera (Del.) Wain. Wain., Monogr. Cladon., vol. II, 1894, p. 354. Masatierra; Valle Colonial, auf der Erde, fruchtend (C. und I. SKOTTS- BERG). Masafuera: Heide oberhalb des Monotals, 850 m (C. und I. SKOTTSBERG)- f. subacuta Wain. Wain., Monogr. Cladon., vol. II, 1894, p. 355. Masatierra: am Wege zwischen Portezuelo und Villagra, bei 400 m, auf der Erde (C. und I. SKOTTSBERG). DIE FLECHTEN DER JUAN FERNANDEZ-INSELN >) —I os) f. squamulifera Wain. Wain., Monogr. Cladon., vol. II, 1894, p. 355. Masatierra: Cordon Centinela, auf Baumstriinken; Quebrada Monte Ma- derugo, auf abgestorbenen Baumstémmen (C. und I. SKOTTSBERG). f. sorediosa Wain. Wain. apud A. Zanvpr. in Kgl. Svensk. Vetensk.-Akad. Handl., vol. LVII, Pred LG07, Pp: 27. Ohne Standortsangabe (C. SKOTTSBERG). Stereocaulon Schreb. A. Phyllocladia linearia, simplicia vel divisa; cephalodia plus minus pe- dicellata. a. Cephalodia gonidiis nostocaceis S. vamulosum. b. Cephalodia gonidiis scytonemaceis S. proximum. B. Phyllocladia granuliformia; cephalodia sessilia S. paschale var. magellanicum. S. paschale (L.) Ach. var. magellanicum Nyl. Nyt. in Mémoir. Soc. Scienc. Nat. Cherbourg, vol. V, 1857, p. 96 et Synops. Lich., vol. I, 1860, p. 243. — Stereocaulon tomentossum var. magellanicum "Yh. Fr., De Stereoc. et Piloph. Comment., 1857, p. 31. — Stereocaulon magellanicum Th. Fr. imeNova Acta Soc, Scient.. Upsal., ser. 3, vol. Il, pars 1, 16587 p. 350: Masatierra: auf Felsen in den Bergen (BERTERO no. 20). Masafuera: auf dem Hochplateau. 1200 m und beim Campo Correspon- dencia, 1130 m, fruchtend (C. und I. SKOTTSBERG). *S. proximum Nyl. Nvyr. in Annal. Scienc. Nat., Bot., ser. 4, vol. XI, 1859, p. 210 et Synops. Lich., vol. I, 1860, p. 237, tab. VII, fig. 14; Hue in Nouv. Archiv. du Muséum, ser. 2, vol. X, 1898, p 154. — Stereocaulon ramulosum £. proximum Tuck. in Pro- ceed. Americ. Acad. Arts and Scienc., vol. VII, 1868, p. 228; Watn. in Annal. Acad. Scient. Fennic., ser. A, vol. VI, no. 7, 1915, p- 25. Masatierra: Puerto Frances, Loma del Incienso; Cordén Escarpado, 365 m; Portezuelo, auf Felsklippen, 590—600; am Wege zwischen Portezuelo und Villagra, 420 m, auf dem Erdboden; Quebrada Villagra 380 m, auf Fels- spalten und Cordén Salsipuedes, 625 m; iiberall fruchtend (C. und I. SKOTTS- BERG). var. compactius A. Zahlbr. nov. var. Thallus substrato arcte adhaerens, fastigiato-conglomeratus. Podetia usque 3,5 cm alta, I—1I,5 mm lata, stricta, rigida, in parte inferiore phyllocladis fili- 374 A, ZAHLBRUCKNER formibus obsita, in parte superiore phyllocladiis verruculoso-inaequalibus, ver- rucis thallum fere omnino obducentibus et quasi corticem formantibus. Masafuera: Las Torres, c. 1350 m, auf Gestein, fruchtend und auf der Heide unweit Correspondencia, 1200 m, fruchtend (C. und I. SKOTTSBERG). Durch das kompakte Wachstum und durch das Fehlen typhischer Phyllo- kladien im oberen Teile der Podezien von den tbrigen Varietaten leicht zu unterscheiden. S. ramulosum Raenschel. RaAENSCHEL., Nomencl. Bot., edit. 3, 1797, p. 328; Monr. in Annal. Scienc. Nat., Bot., ser. 2, vol. IV, 1835; p. o2 ‘et apud -Gay, "Hist. Fisie/y Polit) @iamle; Bot., vol. VII, 1852, p. 154; Nyz., Synops. Lich:, vol. I, 1860, p. 235-s@rome: in Journ. Linn. Soc. London, Bot., vol. XVI, 1877, p. 223; Hur in Nouv. Archiv. du Muséum, ser. 3, vol. X, 1898, p. 243; REINKE in Pringsh., Jahrbtich. fiir wiss. Bot., vol. XXVIII, 1895, p. 115, fig. 48, II—IV. — Lzchen ramulosus Sw., Nova Gener set, spec. Plant.) 31738, 40> 1271. Masatierra: auf der Erde und auf Felsen in den Bergen (BERTERO no. 1652; MOSELEY); Klippen bei Portezuelo, 590 m; Cordon Salsipuedes, 615 m; zwischen Villagra und Tres Puntas, 2—300 m (C. und I. SKOTTSBERG). Masafuera: Las Torres, 1350 m, auf der Erde, fruchtend (C. und I. SKOTTSBERG). var. implexum Nyl. Nvt., Synops. Lich., vol. I, 1860, p. 236, tab. VII, fig. 11; MULL. ARG. in Mission. Scient. Cap. Horn, vol. V, 1889, p. 150. — Stereocaulon implexum ‘Th, Fr., De Stereoc. et Piloph. Comment., 1857, p. 23; Warn., Etud. Lich. Brésil, vol. I, 1890, p. 68. — Stereocaulon corallinum Mont. in Annal. Scienc. Nat., Bot., ser. 2, vol... IV,. 1835, .p:. 02, et apud Gay, Hist.. Fisic. y Polit. Chile), Bot vole vaaie 1852, p. 155; JoHow, Estud. Flora Juan Fernand., 1896, p. 197. Masatierra: Bahia Cumberland, in Waldern (C. und I. SKOTTSBERG). Masafuera: Heide oberhalb des Monotales, bei 850 m, fruchtend (C. und I. SKOTTSBERG). Acarosporaceae. Acarospora Mass. A. Thallus fuscus A. smaragdula. B. Thallus flavus A. citrina. *A. smaragdula Mass. ; Mass., Ricerch. Auton. Lich., 1852, p. 29, fig. 47; A. L. Smiru, Monogr. Brit. Lich., vol. I, 1898, p. 336. — Zndocarpon smaragdulum Wahlbg. apud Acu., Method. Lich., 1803, p. 29. — Lecanora smaragdula Nyl. in Flora, vol. LV, 1872, p. 429. DIE FLECHTEN DER JUAN FERNANDEZ-INSELN 375 Masafuera: im untersten Teil der Quebrada de las Casas, auf vulka- nischem Gestein haufig (C. und I, SKOTTSBERG),. *A. citrina A. Zahlbr. A. Zantpr. apud Rechinger in Denkschrft. Akad. Wiss. Wien, math.-naturw. Cl., vol. LXXXVIII, 1911, p 26. — Wreeolaria citrina 'Tayl. in Hoox., London Journ. of Bot., vol. VI, 1847, p. 158. — Lecanora xanthophana Nyl. in Annal. Scienc. Nat., Bot, ser. 4. vol. XV, 1861, p. 379. — etue sete MenTOSUl a. akle Poe cgs MSS a ais. = x Se oh Bviny RO. o. 2 a eee eS > SURO ys 2 a RGM tats ts ES as, cn ue ee » PCRS hs i SAS AMICECHSIS: =! a, \\ miata. > la Gee (a tet oes. ge lel ORE ce, 2 i ee » pabteGnicdst. SOLeGiGsHiay .<..s° ska coaex) <0» Ya U6) oh ae aoagee deft Uw anreaeee = gene cAI Rs nl so eke ow Suma fae, os, Sabbatini : : 357 MEEUOCISALCE epee 6 kate, «acoder caiacnis «| 2" a), xr ecaneme a apes 359 > > fo: SHPGYRISA, cue «in Wer Seana erie ne MMI We a ahd ie an ay ane RESAGLEG A: cine oe eo valtw. We herby ol aS, mn ele te EMRE CIE ENEL ALI Bee val och Mot | an noe bie ny Copies ao ond So, be ee 360 MMMERET Se BOLYCINGIOICES: 6's cs ce ss eo Js (Sar wk ee ee «+ i ws 8 ie 955 Pevanun Jeucoxanthum var..albidius: 2 2... Ss 2 Se ee ecmiapora. versicolor vat. Mucrocarpa =/s <6 26+ jh > se mo = She lm A yw hes 365 upEnIPeEMESTSTCEHAUGC AANA oar. 3 3: ca var (xs Vet whe act's omy, a RE ew ae 318 feaaetctyen lopadioides:. (95 - 2. «1 sea 5 Ge ee 8 ow ees Weculwia subdenticulata, 2. 5°. -g ee es ee Se ce ee Pannaria Hilatis .. . . » WEIPMMOS2 Vat VUlCANICA > “awa nai ails, 3-0 ua! ©. Jos la 5 angst nea a aamenCt Conspcrsa yar-/Obtecta 2. . 9 oS es se 5 we js le em » EMEXEISHICCA, £.. dIVISA. . s.a_-o0s0 SNR aie. x <<) oie © oust arent onde ee ee Peanenella symptychia . . . . 1... « Pertusaria hadrocarpa . » Helcarps. Var TIONOSPOra:, Ss meses |S. =<) 5h Ya eee, bee ae a= eee rd > Skottsbergii . RPOAECMEACMEESSENA DS | 5S! s.ce er ta) ot 3 ee eee Se! tae) \s nme Nal ee tae eee MBEERSRAMECTIANA <5 af = 5. iw “gale oe erie Ow te a) See ee OS i re ee eo ae Sea Seasrotet: SNSUSHSECEUM: fs 2 ej 0s) 2 dae een ce eo oS Ye a SS) Tat tah lee es eee > Eepltledingnl + yo o.s= 6. is)=)cakleee she 2% 2 0 )S, ». pat) en ee » sis LEC (ee ore Pre Me nC meee Ace 2G oS ee > MELCANNICHIN 5 «consti “ey /e).5 ist, lehiey sca. ©) oy 2 8e oa ieee SUSIE CHE AUNEL A 58 so Salas oh ays ce oe Fe sy ocee) im pale de ek ina (ade are ye . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . WW WW UW WW WWW WWwWwWw vw “] a | iizeearnon'microspermum 00°40 02' 2. 6S. i 5 ee ee > obscuratum: var. deminuium® ...°. 2 22. 2 2 ee ee SEhisniatomm AECeGenSie easy cary al cars ot (2) <2) oie’ sd) eg a bpath ces) De meted ot Sy) Steccocaulon proximum! var, ComIpactius 2... << io. «is ya; amie ~ ee te ee et 29 1a Serta carpolonia. f.@riiatae Paine Bt cere (ook 0 yp a Sse eee ey as i ots at © fesoillima “war. Wietiss: Sete re cies = 2S x lake, See ee Si Soy mcs pe 340 So enallepiind ware SHEMEN = oe one) Sc. oe = ce koe ees Se es 948 wy eames, var. Durvilleiy Se Go: 2, 4)... - «56. 2) =. ste eee ee ms ees > SS » » 3 “enldochigset Sree sn oo = Oy a cee ete = = oe > » oo flavicans = ot ) Th. proboscideum Broth. n. sp. Gracilescens, sordide viridissimum, opacum. Caules secundarii usque ad 7 cm alti, stipitati, stipite usque ad 3 cm longo, foliis squamaeformibus dense vestito, dein pinnatim ramosi, ramis usque ad 3 cm longis, simplicibus vel in- fimis subpinnatim ramulosis. Folia erecto-patentia, concaviuscula, Janceolata, apice plerumque excisa ibidemque minute serrulata, nervo longe excedente terminata, c. 2 mm longa; nervo crassissimo, ubique aequilato, longe excedente; cellulis haud incrassatis, acuminis irregularibus, ovalibus vel triangularibus, cae- teris hexagono-oblongis, basin versus sensim longioribus, infimis linearibus. Caetera ignota. Masatierra: Pangal; ad rupem praeruptam in cataracta; 205 m. s. m. (256). Species praecedenti affinis, sed foliis acumine exciso, nervo crassiore, longe excedente dignoscenda. Th. assimile Broth. n. sp. Praecedenti gracilius, atroviride, opacum. Caules secundarii usque ad 8 cm alti, stipitati, stipite 2—5 cm longo, foliis squamaeformibus dense vestito, dein pinnatim ramosi, densiuscule foliosi, complanati, ramis usque ad 6 cm longis, infimis pinnatim ramulosis. Folia erecto-patentia, concaviuscula, lanceo- lata, raptim in acumen breve, argute serratum contracta, c. 2 mm longa; nervo crassissimo, ubique aequilato, infra summum apicem folii evanido; cellulis haud incrassatis, superioribus ovalibus, caeteris hexagono-oblongis, basilaribus linearibus. Caetera ignota. Masatierra: Quebrada Damajuana; ad marginem cataractae (350); 250 m. s. m. Species praecedenti valde affinis, sed foliis in acumen breve, argute ser- ratum contractis dignoscenda. Th. confertum (Mitt. sub Porotricho in Hemsley Challenger Rep. Bot. I.). Masatierra: Sine loco designato (MOSELEY). Area: Endemic. Pinnatella (C. Miill.) Fleisch. P. (Eupinnatella) macrosticta Broth. n. sp. — Plate 27, fig. 14—15. Tenella, laxe caespitosa, fuscescenti-viridis. Caulis primarius elongatus, repens; caules secundarii vix ultra 5 mm alti, + distincte pinnatim ramosi, ramis patentibus brevibus, laxiuscule et complanate foliosis, simplicibus, obtusis. Folia patula, breviter oblonga, acumine brevi, acutiusculo terminata, c. 0,75 mm longa et c. 0,2 mm lata, apice minute denticulata; nervo infra apicem folii evanido; cellulis rotundato-hexagonis, papilla medig distinctissima instructis, inferioribus sensim longioribus. Caetera ignota. 25 2391- The Nat. Hist. of Juan Fernandez and Easter Isl. Vol. II. 434 Vv. F. BROTHERUS Masafuera: Quebrada de las Casas; ad corticem (271, 273). Species distinctissima, foliis cellulis papilla media elevata instructis a con- generibus raptim dignoscenda. Hookeriaceae. Distichophyllum Doz. et Molk. D. (Discophyllum) subelimbatum Broth. n. sp. — Plate 27, fig. 9—10. Dioicum; gracile caespitosum, caespitibus densiusculis, mollibus, depressis, glauco-viridibus, aetate lutescenti-viridibus, opacis. Caulis fragilis, repens, hic illic rhizoideis fuscis instructus, densiuscule et complanate foliosus, cum foltis 2—3 mm latus, ramosus. Folia sicca vix mutata, planiuscula, lateralia patentia, e basi anguste et breviter spathulata obovata, obtusa, mutira vel apiculata, marginibus minutissime crenulatis; nervo tenuissimo, ad vel paulum ultra me- dium folii evanido; cellulis laminalibus superioribus angulato-rotundatis, O,or5— 0,020 mm, dein sensim majoribus, ovali-hexagonis, basilaribus oblongo-hexagonis, marginalibus inferioribus elongatis, angustis, limbum uniseriatum, hyalinum, vix ad medium folii productum efformantibus. Folia intermedia subconformia, media multo minora, ovalia, nervo breviore. Bracteae perichaetii erectae, minutae, e basi oblonga breviter ligulatae, acutae, apice crenulatae, enerves. Seta c. 8 mm alta, tenuis, rubra, laevissima. Calyptra superne scabra, basi pilis longis instructa. Caetera ignota. Masatierra: In jugo inter valles Piedra agujereada et Laura; ad ligna putrida; 650 m. s. m. (306, 309). Valle Colonial; ad parietes cavernae plus minusve humidas (304, 305, 308, 310, 312, 313, 316, 317). In jugo Centinela; ad truncos Filicum (311). Salsipuedes; ad ligna putrida in Dicksonieto; 660 MreSa me SU4e 275 ). Species distinctissima, ob folia subelimbata cum nulla alia commutanda. D. (Discophyllum) assimile Broth. n. sp. Dioicum: gracilescens, caespitosum, caespitibus densiusculis, mollibus, glauco-viridibus, opacis. Caulis procumbens, inferne fusco-radiculosus, densiuscule et complanate foliosus, cum foliis c. 3 mm latus, ramosus. Folia sicca vix mutata, planiuscula, e basi anguste et breviter spathulata obovata, obtusa, api- culata, limbata, superne minutissime serrulata; nervo tenuissimo, ultra medium folii evanido; cellulis laminalibus superioribus angulato-rotundatis, 0,o15—O,o20 mm, dein sensim majoribus, ovali-hexagonis, basilaribus oblongo-hexagonis, marginalibus elongatis, angustis, limbum biseriatum, lutescentem, usque ad apicem productum efformantibus. Folia intermedia subconformia, media multo minora, ovalia. Caetera ignota. Masatierra: Portezuelo; ad ligna putrida in Dicksonieto;.c. 550 m. s. m. (307). THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 435 Species praecedenti habitu persimilis, sed foliis limbo usque ad apicem producto dignoscenda. D. (Discophyllum) fernandezianum Broth. n. sp. — Plate 27, fig. 7—8. Gracile, caespitosum, caespitibus densis, mollibus, viridissimis, opacis. Caulis erectus, usque ad 2,5 cm altus, per totam longitudinem parce fusco- radiculosus, dense et vix complanate foliosus, superne dense fasciculatim ra- mosus. Folia sicca vix mutata, inferiora limbo excepto plerumque destructa, planiuscula, lateralia erecto-patentia, ovalia, raptim in apiculum contracta, c. I,x mm longa et c. 0,57 mm lata, limbata, e medio ad apicem serrulata; nervo tenuiusculo, lutescente, ultra medium folii evanido; cellulis laminalibus angulato- rotundatis, O,oo7—0,o1o mm, basilaribus raptim multo majoribus, oblongo-hexa- gonis, marginalibus elongatis, angustis, limbum lutescentem, 3 seriatum effor- mantibus. Folia ventralia et dorsalia minora. Caetera ignota. Masafuera: In alpinis prope campos »Correspondencia» dictos; ad saxa humida rivuli; 1100 m. s. m. (303). Species D). rotundifolio (Hook. fil. et Wils.) Broth. forsan affinis, sed foliis densius areolatis, latius limbatis jam dignoscenda. Pterygophyllum Brid. P. anomalum (Schwaegr.) Mitt. Masatierra: Pangal; ad marginem cataractae; 205 m. s. m. (290). Area: Fuegia. P. obscurum (Mont.) Mitt. Masafuera: Quebrada Loberia; in rivulo una cum Philonotide vagante (291). Area: Chile, W. Patagonia, Falkland Isl., Tasmania. P. tenuinerve Broth. n. sp. — Plate 27, fig. 5—6. Dioicum; gracilescens, caespitosum, caespitibus laxiusculis, mollissimis, sordide glauco-viridibus, opacis. Caulis arcuato-procumbens, vix ultra I cm longus, basi fusco-radiculosus, laxe et complanate foliosus, simplex. Folia facillime emollita, sicca vix mutata, planiuscula, lateralia patula, e basi bre- vissime spathulata oblonga, obtusa, usque ad 4 mm longa et usque ad 1,7 mm lata, e medio ad apicem grosse serrata; nervo tenui, brevi vel brevissimo; cellulis laminalibus valde pellucidis, rotundatis, superioribus 0,o5—-0,o6 mm, basin versus sensim majoribus. Folia ventralia et dorsalia erectiora, minora. Bracteae perichaetii erectae, e basi oblonga anguste lanceolatae, integrae. Seta c. 12 mm alta, tenuis, nigrescens, laevissima. Theca nutans, minuta, breviter oblonga, sicca deoperculata sub ore constricta, atrofusca. Caetera ignota. 436 V. F. BROTHERUS Masatierra: Cumberland, Bay, Pico Central; ad ligna putrida; 390 m. s. m. (386 fo. major). Salsipuedes; ad vegetabilia putrescentia in Dicksonieto parce; 660 m. s. m. (330). In jugo inter valles Piedra agujereada et Laura; ad ligna putrida; 650 m. s. m. (287, 288, 289 fo. major). Species praecedenti valde affinis, sed foliis glauco-viridibus, facillime emol- litis, grosse serratis, nervi tenui, brevissimo instructis dignoscenda. P. denticulatum (Hook. fil. et Wils.) Mitt. Masatierra: Sine loco designato (E. C. REED). Area: Fuegia, Falkland Isl., East Australia, Tasmania, New Zealand, Auckland und Campbell Isl. Eriopus (Brid.) C. Mull. E. leptoloma Broth. n. sp. — Plate 27, fig. 1—2. Autoicus; robustiusculus, caespitosus, caespitibus laxiusculis, laete viri- dibus, aetate lutescentibus, vernicoso-nitidis. Caulis erectus vel adscendens, I cm vel paulum ultra altus, cum foliis usque ad 5 mm latus, basi fusco-radicu- losus, laxiuscule foliosus, simplex. Folia sicca vix mutata, planiuscula, lateralia patentia, paulum asymmetrica, e basi brevissime spathulata ovalia, acuta, lim- bata, e medio ad apicem serrata; nervo brevissimo, furcato; cellulis ovali-hexa- gonis, basin versus sensim longioribus, marginalibus elongatis, angustis, lim- bum 2—3 seriatum efformantibus. Folia ventralia et dorsalia minora. Seta c. 4 mm alta, lutea, mamillis humilibus, latis, hyalinis instructa. Caetera ignota. Masatierra: V. Colonial, Quebrada Gutierrez; ad ligna putrida silvae; c. 300 m. s. m. (302). Masafuera: Quebrada del Blindado; ad ligna putrida silvae; c. 440 m. s. m. (301). Species £. apzculato (Hook. fil. et Wils.) Mitt. affinis, sed inflorescentia autoica, foliis acutis nec cuspidatis, argutius serratis, angustius limbatis nec non seta breviore dignoscenda. E. grandiretis Broth. n. sp. — Plate 27, fig. 3—4. Robustus, mollis, glaucoviridis, opacus. Caulis usque ad 6 cm altus, cum foliis usque ad 8 mm latus, inferne fusco-radiculosus, densiuscule foliosus, atte- nuatus, simplex vel parce ramosus. Folia inferiora remota, dein magis con- ferta, sensim majora, lateralia patentia, oblonga, raptim in acumen breve, acu- tum attenuata, usque ad 4 mm longa et I,9 mm lata, marginibus e medio ad apicem argute serratis; nervo e basi fere bifurco, cruribus inaequalibus, longiore ad medium folii evanido; cellulis laxis, valde pellucidis, oblongo- vel ovali- hexagonis, marginalibus angustis, limbum uniseriatum, hyalinum efformantibus. Caetera ignota. THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 437 Masafuera: In parte intima faucis Casas; ad truncum (265). Species pulcherrima, praecedenti affinis, sed statura robustiore foliisque laxe areolatis, limbo uniseriato longe diversa. Lamprophyllum Schimp. L. splendidissimum (Mont.) Schimp. Masatierra: Sine loco designato (BERTERO, MOSELEY). Quebrada Gu- tierrez; ad ramulos arborum in silva; c. 300 m. s. m. (300). Ad rupes humidas in silva secus viam ad Portezuelo (298). Portezuelo; ad ligna putrescentia Dicksonieti; c. 550 m. s. m. (297 p. p., 299). Salsipuedes; ad truncos Drimydis in Dicksonieto; 660 m. s. m. (292, 293). Villagra, Quebrada Choza; ad truncos iesiiva C. 250.m, S. m:) (295, 208 p. p.): Masafuera: Quebrada de las Casas; ad rupem praeruptam (294). Que- brada del Blindado; ad truncum; 440 m. s. m. (296). Area: Chile, W. Patagonia. Hypopterygiaceae. Lopidium Hook. fil. et Wils. L. concinnum (Hook.) Fleisch. Masatierra: In declivi montis Yunque; ad truncos; c. 450 m. s. m. (331, 333). In jugo inter Quebrada Piedra Agujereada et Quebrada Laura; 650 m. s s. m. (332). Area: Chile, W. Patagonia, Australia, Tasmania, New Zealand, Auck- land, Isl. Hypopterygium Brid. H. Thouini (Schwaegr.) Mont. Masafuera: Quebrada de las Casas raro; c. 200 m. s. m. (334). Area: Chile, W. Patagonia, Fuegia. Rhacopilaceae. Rhacopilum Palis. Rh. fernandezianum Card. in Soc. Havr. p. 7 (1921) (22. fernandeztanum Broth. in sched.). Dioicum; gracilescens, caespitosum, caespitibus densis, depressis, late extensis, viridibus vel lutescentibus, caulibus intertextis. Caulis elongatus, repens, 438 V. F. BROTHERUS per totam longitudinem fusco-tomentosus, dense foliosus, pinnatim ramosus, ramis patulis, simplicibus, vix ultra 5 mm longis, cum foliis c. 2 mm vel paulum ultra latis, singulis longioribus, pinnatim ramulosis. Folia postica siccitate in series duas reflexas, aristatas disposita, madore applanata, ovato-oblonga, ob- tusiuscula, aristata, apice tantum minute serrulata; nervo in aristam longiusculam, rigidam, subintegram excedente; cellulis laminalibus minutis, angulato-ovalibus vel subrotundis, papilla media minutissima instructis, basilaribus oblongis. Folia antica minora, late cordato-lanceolata, apice minutissime serrulata; nervo in aristam longam, rigidam, subintegram excedente. Seta c. I,5 cm alta, tenuis, rubra. Theca suberecta vel inclinata, anguste cylindrica, arcuata, c. 5 mm longa, pallida, aetate fusca, sicca plicatula. Operculum e basi convexa longe rostratum. Calyptra parce pilosa. Masatierra: Sine loco designato (BERTERO). In declivi septemtrionali montis Yunque; ad truncos Filicum; 4—500 m. s. m. (282). Valle de Anson, prope Plazoleta; ad truncos putridos, c. 260 m. s. m. (277, 281, 284). Valle Colonial, Pico Central, ad terram 390 m. s. m. (261); Quebrada Gutierrez, in silva c. 300 m. s. m. (318, 324); Quebrada del Monte Maderugo, ad truncos silvae siccioris (260); Quebrada seca, 435 m. s. m. ad truncos J/yrceugeniae (278, 321) et ad terram (320); in jugo ad Portezuelo, in Dicksonieto, c. 550 m. s. m. (297 p. p.) et in silva ad arbores et ad terram (262). Quebrada Juanango, ad truncos in rivulo sicco (257, 262). Inter Villagra et Tres Puntas locis humidioribus (323). Quebrada de la Choza, in silva 250 m. s. m. ad terram (263). Masafuera: inter Sanchez et Toltén; ad truncos putridos in silva, 515 m. s. m. (259, 325). Quebrada del Mono; ad truncos in silva, 475 m. s. m. (280). Quebrada de las Casas; ad rupes praeruptas humidas (274, 326). Species Rk. tomentoso (Sw.) Brid. affinis, sed foliis apice minute serrulatis vel subintegris, cellulis laminalibus minoribus nec non theca anguste cylindrica dignoscenda. Some expressions in CARDOT’s diagnose led me to the conclusion that his species was different from that collected by SKOTTSBERG, but the examina- tion of type material removed all doubts as to their identity. The briefness of CARDOT’s description justifies the more detailed one given above. Area: Endemic. Thuidiaceae. Thuidium Bryol. eur. Th. (Thuidiopsis) Masafuerae Broth. n. Sp. Dioicum. Robustiusculum, caespitosum, caespitibus laxis, lutescentibus. Caulis elongatus, procumbens, flexuosus, parce radiculosus, laxe bipinnatim ra- mosus, ramis patulis, ad 1,5 cm longis, arcuatulis, attenuatis, ramulis subremotis, brevibus. Folia caulina arcuato-patula, plicata, e basi late cordata raptim lanceolato-subulata, c. 1,1 mm longa etc. 0,75 mm lata, marginibus basi recurvis, THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 439 apice minutissime serrulatis; nervo valido, infra apicem folii evanido; cellulis subrotundis O,o1o—O,or2 mm, papilla media instructis, subpellucidis. Folia ramea brevius acuminata, ramulina ovato-lanceolata, obscura. Caetera ignota. Masafuera: Cordon del Barril, in filiceto (513). Species 7h. fulvastro (Mitt.) Jaeg. affinis, sed caule laxe bipinnatim ra- moso oculo nudo jam dignoscenda. Th. Valdiviae Broth. in Herb. Dioicum. Gracilescens, caespitosum, caespitibus densis, extensis, viridibus, aetate lutescentibus. Caulis elongatus, arcuato-procumbens, interrupte fusco- radiculosus, dense bipinnatim ramosus, ramis suberectis, vix ultra I cm longis, strictis vel arcuatulis, ramulis brevissimis. Folia caulina arcuato-patula, plicata, e basi late cordata raptim lanceolato-subulata, c. 0,3 mm longa et c. 0,65 mm lata, marginibus basi recurvis, subintegris; nervo valido; infra apicem folii eva- nido; cellulis angulato-rotundatis, 0,o07—O,oro mm, papilla media instructis, subpellucidis. Folia ramea angustiora, brevius acuminata, ramulina ovata, acuta, cellulis obscuris. Bracteae perichaetii erectae, piliformiter attenuatae, basi ciliis elongatis instructae. Caetera ignota. Masatierra: Quebrada Damajuana, in declivi silvatico, c. 200 m. s. m. (514, 515). Masafuera: Quebrada de las Casas. Species a praecedente statura graciliore, caule dense bipinnatim ramoso, foliis caulinis subintegris, densius areolatis dignoscenda. Area: Chile, Valdivia (A. HOFMANN in herb. BROTH.). Amblystegiaceae. Sciaromium Mitt. S. pachyloma (Mont.) Par. Masatierra: Pangal; in cataracta; 205 m. s. m. (239). Quebrada Juanango; in aqua fluitante (238). Cerro Chumacera; ad rupes praeruptas irroratas (237). Masafuera: Quebrada de las Casas; in aqua fluitante. Quebrada de la Loberia, in cataracta (235). Area: Chile. Hypnaceae. Hypnum Dill. emend. H. Lechleri C. Miill. Masatierra; Sine loco designato (BERTERO, MOSELEY); Puerto Frances, c. 500 m. s. m. (399). In jugo inter Quebrada Piedra agujereada et Quebrada 440 V. F. BROTHERUS Laura; 650 m. s. m. (392, 398). Cordén Centinela; 530 m. s. m. (453, 457). Quebrada Damajuana (459). Cerro Damajuana (458). El Yunque, in declivi septemtrionali; 4—500 m. s. m. (391). Portezuelo de Villagra; c. 500 m.s. m. (394, 455) et c. 550 m.s. m. (456). Quebrada Gutierrez (490). Quebrada seca; 435 m. s. m. (393, 396, 397). Cordon Salsipuedes; c. 600 m. s. m. (395). Puerto Ingles; c. 575 m. s. m. (474). Quebrada de la Choza; c. 250 m. s. m. (452). — Ubique ad truncos et ad ligna putrida. Area: Chile, Patagonia. Isopterygium Mitt. I. fernandezianum Broth. n. sp. Autoicum; gracilescens, caespitosum, caespitibus densiusculis, depressis, laete viridibus, sericeo-nitidis. Caulis repens, per totam longitudinem fusco- radiculosus, dense et complanate foliosus, vage ramosus, ramis brevibus, com- planatis, cum foliis c. 1,5 mm_ latis, simplicibus, obtusis. Folia falcatula, con- caviuscula, caulina ovato-lanceolata, in acumen + elongatum, subpiliforme sen- sim attenuata, marginibus erectis, integris; enervia; cellulis anguste linearibus, basilaribus infimis laxis, oblongis, hyalinis vel aureis, alaribus haud diversis. Folia ramea brevius acuminata, superne + distincte serrulata. Bracteae pe- richaetii internae erectae, subsensim in acumen elongatum, lanceolato-subulatum, serrulatum attenuatae. Seta c. 5 mm alta vel paulum ultra, tenuis, rubra. Theca horizontalis vel subnutans, ovalis, regularis, sicca sub ore paulum con- stricta, fusca. Masatierra: In jugo Centinela; ad ligna putrida; 530 m. s. m. (479). var. longisetum Broth. n. var. Seta I—I,s cm alta. Masatierra: Cordén Centinela; 530 m, s. m. (485). Valle Anson prope Plazoleta; 260 m. s. m. (491). Cumberland Bay, Pico Central; 390 m. s. m. (492). Valle Colonial, Quebrada seca; 435 m. s. m. (481, 484, 487). Cerro Piramide; c. 600 m. s. m. (488). Cordén Salsipuedes; in Dzcksonzeto; 660 m. s. m. (480, 483). El Yunque; 4—500 m. s. m. (482). Villagra, Quebrada de la Choza; c. 250 m. s. m. (486). Ubique ad ligna putrida. Species ob folia cellulis infimis laxis instructa cum /. ¢exero (Sw.) Mitt. comparanda, sed statura robustiore jam dignoscenda. Sematophyllaceae. Rhapidostegium (Bryol. eur.) De Not. Rh. (Cupressinopsis) Masafuerae Broth. n. sp. Autoicum; robustiusculum, caespitosum, caespitibus densis, mollibus, viri- dibus vel lutescenti-viridibus, sericeo-nitidis. Caulis clongatus, repens, dense THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 441 ramosus, ramis usque ad 6 cm longis, adscendentibus, dense foliosis, dense ramulosis, ramulis vix ultra 1 cm longis, sursum spectantibus. Folia falcata, concava, oblongo-lanceolata, sensim in subulam + elongatam, serrulatam atte- nuata; enervia; cellulis angustissime linearibus, basilaribus infimis brevioribus, aureis, alaribus magnis, oblongis, vesiculosis, aureis. Bracteae perichaetii in- ternae vaginantes, in acumen breve lanceolatum, serrulatum attenuatae. Seta I cm vel paulum ultra alta, tenuis, rubra, laevis. Theca suberecta, turgide ovalis, fusca. Caetera ignota. Masafuera: Correspondencia; 1150 m. s. m. (499). Species pulchra, habitu 4. cyparzssozdi (Hornsch.) Besch. similis, sed foliis superne serrulatis nec non seta brevi jam dignoscenda. Rh. (Cupressinopsis) aberrans Broth. n. sp. Autoicum; gracile, caespitosum, caespitibus densiusculis, depressis, viri- dibus vel lutescenti-viridibus, nitidis. Caulis repens, per totam longitudinem radiculosus, dense et complanate foliosus, pinnatim ramosus, ramis brevibus, patentibus, complanatis, simplicibus, obtusis. Folia patentia, + distincte sub- secunda, concava, lanceolata vel oblongo-lanceolata, sensim + longe subulata, marginibus erectis, subintegris vel superne minutissime serrulatis; enervia; cellulis angustissime linearibus, infimis aureis, alaribus oblongis, vesiculosis, aureis, omnibus laevissimis. Bracteae perichaetii erectae, internae raptim in acumen breve, serrulatum attenuatae. Seta c. I cm alta, tenuis, rubra, laevis- sima. Theca horizontalis vel subnutans, minuta, ovalis, sicca sub ore paulum constricta, fuscidula. Operculum longe rostratum. Masatierra: In jugo Centinela; 530 m. s. m. (385, 389). Plazoleta in valle Anson; c. 260 m. s. m. (379). Quebrada del Monte Maderugo (377). Salsi- puedes, in Dicksonieto; 660 m. s. m. (386 p. p.). Quebrada Gutierrez; in silva; c. 300 m. s. m. (387). Quebrada Juanango; c. 250 m.s. m. (383, 384). Ubique ad ligna putrida. Masafuera: Quebrada del Mono; ad truncos arborum (388). In valle Mono; ad truncum Dzcksoniae (378). Quebrada del Blindado; 400—440 m. s. m. (489, 500). In silva inter Sanchez et Toltén; ad ligna putrida; 515 m. s. m. (390). Cordon Barril (498). Species Rk. callido (Mont.) Jaeg. foliorum forma affinis, sed caule com- planato foliisque + distincte subsecundis nec falcatis oculo nudo jam digno- scenda. Rh. caespitosum (Sw.) Jaeg. Masatierra: Cordén Escarpado, El Pico; ad moles humidiusculos; 365 m. s. m. (497). Quebrada Juanango; ad arbores (370). Masafuera: Quebrada de la Loberia; ad arbores; 280 m. s. m. (496). Area: Antilles, Trinidad, tropical and subtropical parts of South America. 442 V. F. BROTHERUS Rh. (Aptychus) caespitosoides Broth. n. sp. Autoicum; gracilescens, caespitosum, caespitibus densiusculis, late extensis, lutescenti-viridibus, nitidis. Caulis elongatus, repens, per totam longitudinem hic illic fusco-radiculosus, dense foliosus, dense ramosus, ramis brevibus. Folia sicca laxe imbricata, humida erecto-patentia, concava, oblongo-elliptica, sensim lanceolato-acuminata, c. 1,7 mm longa et c. 0,5 mm lata, marginibus erectis, integerrimis; enervia; cellulis angustissimis, haud incrassatis, infimis aureis, alaribus oblongis, versiculosis, aureis vel hyalinis. Bracteae perichaetii erectae, acumine brevi, serrulato. Seta usque ad 1,5 cm alta, tenuis, rubra. Theca horizontalis, e collo distincto ovalis, sicca deoperculata sub ore parum con- stricta, pallida. Operculum longe subulatum. Masafuera: Quebrada de las Casas; ad truncos in rivulo; c. 200 m. s. m. (493). Ibidem ad saxa (371). Ibidem ad saxa rivuli (494, 495). Species praecedenti similis, sed bracteis perichaetii acumine brevi in- structis dignoscenda. Rh. (Aptychus) brachycladulum Broth. n. sp. Autoicum; gracilescens, caespitosum, caespitibus densiusculis, laete viri- dibus, aetate fuscescenti-viridibus, sericeo-nitidis. Caulis elongatus, repens, per totam longitudinem fusco-radiculosus, dense foliosus, densissime ramosus, ramis brevissimis, erectis, simplicibus, obtusis. Folia erecto-patentia, concava, oblongo- elliptica, lanceolato-acuminata c. 1,5 mm longa et usque ad 0,6 mm lata, mar- ginibus erectis, integerrimis; enervia; cellulis angustissimis, haud incrassatis, infimis aureis, alaribus’ oblongis, vesiculosis, plerumque hyalinis. Bracteae perichaetii erectae, sensim lanceolato-subulatae, integrae. Seta vix ultra 5 mm alta, tenuissima, rubra, laevissima. Theca suberecta vel inclinata, regularis, oblonga, sicca deoperculata sub ore constricta, fuscidula. Operculum longe rostratum. Masatierra: Valle Colonial, Quebrada seca; ad truncum Cofrosmatis triflort; 435 m. s. m. (369). Yunque ad truncum (368). Rigodium Kunz.; Schwaegr. R. toxarion (Schwaegr.) Schimp. Masatierra: Sine loco designato (BERTERO, DOUGLAS, E. C. REED); Puerto Frances; ad basin truncorum; c. 500 m. s. m. (412, 416, 419). In jugo inter Quebrada Laura et Quebrada Piedra Agujereada; c. 600 m. s. m. (415) et 650 m. s. m. (406). Quebrada de la Damajuana; in silva (374, 413, 428). El Yunque; 4—500 m. s. m. (431, 436, 438). Cumberland Bay, Pico Central; 390 m. s. m. (421). Valle Colonial, Quebrada seca; ad basin truncorum; 435 m. s. m. (400, 408, 437, 441). Cerro Pirdmide; c. 600 m. s. m. (407, 411). Porte- zuelo; c. 600 m. s. m. (430) et c. 550 m. (417). Infra Portezuelo de Villagra; ad ligna putrida; c. 500 m. s. m. (460). Quebrada Gutierrez; in silva; c. 300 THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 443 m. s. m. (410). Cordon Salsipuedes: 615 m. s. m. (401, 405); in Dicksonieto; 660 m. s. m. (420); in silva lapidosa; 625 m. s. m. (427); in fruticetis; 615 m. s. m. (462). Quebrada Juanango; ad cataractam (403) et ad terram; c. 250 m: s. m. (422). In trajectu Villagra; ad rupem praeruptam; c. 600 mi. s. m. (414). Quebrada de la Choza; ad truncos; c. 250 m. s. m. (298, 402, 461, 482). Masafuera: Quebrada del Mono; in silva (439, 443). Quebrada del Blin- dado; ad terram et lapides; 440 m. s. m. (442). In jugo supra Varadero; ad truncos in Dicksonieto; 8—go00 m. s. m. (409). Quebrada de la Loberia; ad terram et lapides silvae; 280 m. s. m. (440). Area: Tropical and subtropical parts of South America, Patagonia. R. arborescens (C. Mill.) Broth. Masatierra: Ad rupem praeruptam infra Portezuelo; c. 500 m. s. m. (424, I I 5 4 426). Salsipuedes; in silva jugi (425). Masafuera: Quebrada del Mono; in silva (463). Quebrada de las Casas; ad moles sat frequenter (464, 465). Ad terram inter Sanchez et Toltén; 515 m; Ss: m. (423). Area: Chile. R. hylocomioides Card. et Broth. Masafuera: Correspondencia; c. 1100 m. s. m. (466). Area: Patagonia. R. robustum Broth. n. sp Dioicum; robustum, caespitosum, caespitibus laxis, rigidis, fuscescentibus, vix nitidiusculis. Caulis elongatus, repens, nudus, divisionibus usque ad 5 cm longis, inferne simplicibus, dein bipinnatim ramosis, ramis vix ultra 2 cm longis, arcuatulis, attenuatis vel obtusis. Folia caulina subsquarrosa, late cordato- ovata, raptim in acumen lanceolato-subulatum attenuata, c. 1,5 mm longa et c. 0,8 mm lata, marginibus basi recurvis, integris vel acumine minutissime serru- latis; nervo in acumine evanido; cellulis incrassatis, lumine angusto. Folia ramea erectiora, minora, ovato-lanceolata, superne minutissime serrulata. Caetera ignota. Masatierra: Salsipuedes; in silva lapidosa jugi; 625 m. s. m. (429). Species a congeneribus statura multo robustiore oculo nudo jam digno- scenda. Rhynchostegium Bryol. eur. Rh. complanum (Mitt.) Jaeg. Masatierra: Quebrada de la Damajuana; in silva (432, 435). Ad rupem praeruptam infra Portezuelo; c. 500 m. s. m. (433). In jugo Portezuelo; in silva; c. 550 m. s. m. (434). 444 V. F. BROTHERUS Masafuera: Quebrada del Blindado; ad truncos; 440 m. s. m. (506). Area: Chile. Rh. tenuifolium (Hedw.) Jaeg. Masatierra: Sine loco designato (E. C. REED). Area: Uruguay, Chile, East Australia, Tasmania, New Zealand. Catagoniopsis Broth. C. Berteroana (Mont.) Broth. Masatierra; Sine loco designato (MOSELEY). Peace nile: Hy pnodendraceae. Hypnodendron Lindb. H. microstictum Mitt. Masafuera: Quebrada del Mono; in silva (267 p. p.). Cerro Correspon- dencia; c. 1100 m. (340). Area: Chile. Polytrichaceae. Oligotrichum Lam. et DC. QO. canaliculatum (Hook.) Mitt. var. minus Broth. n. var. A typo foliis brevioribus, seta vix ultra 1 cm alta nec non theca multo minore differt. Masafuera: Ad rupes terra obtectas inter Casas et Chozas; c. 450 m. s. m. (467, 468). Area of O. canaliculatum typicum: Chile. Psilopilum Brid. P. antarcticum C. Miill. Masafuera: Las Torres; ad rupes; 1370 m. s. m. (469). Area: Fuegia, Kerguelen, South Georgia. THE MUSCI OF THE JUAN FERNANDEZ ISLANDS 445 Polytrichadelphus (C. Mil.) Mitt. P. magellanicus (L.) Mitt. Masafuera: Correspondencia; in campo; 1I—1300 m. s. m. (470, 471) et c. 1400 m. s. m. (472). Area: Patagonia, Fuegia, Falkland Isl., Tasmania, New Zealand, Auck- land and Campbell Isl. Dendroligotrichum (C. Mill.) Broth. D. dendroides (Brid.) Broth. Masafuera: Correspondencia (473); c. 1100 m. s. m. (474) et 1150 m. s. m. (475). Quebrada de las Casas; c. 200 m. s. m. (476). Area: Peru, Chile, Patagonia, New Zealand. As JOHOW’s book on the Botany of Juan Fernandez has long been the principal source of information regarding these islands and will, it is assumed, continue to be consulted, I have found it useful to copy his list of mosses and add the name for each species as used in my list. In several cases an identification, for want of original specimens from the islands, has not been possible. Racopilum tomentosum Brid. Probably the same as Rh. fernandezianum Card. : Stereodon Lechleri (C. Mill.) Mitt. = Hypnum Lechleri C. Mill. Hypnum circinale Hook. =? Hypnum tenuifolium Hedw. = Rhynchostegium tenuifolium (Hedw.) Jaeg. Hypnum neckeroides Hook. =? Hypnum Berteroanum Mont. = Catagoniopsis Berteroana (Mont.) Broth. Hypnum serrulatum Hedw. =? Hypnum aciculare Brid. Probably the same as Ptychomnium subaciculare Besch. Hypnum toxarion Schwaegr. = Rigodium toxarion (Schwaegr.) Schimp. Hypnum crassiusculum Schwagr. =? Leskia mollis Hedw. = Weymouthia mollis (Hedw.) Broth. Leptodon Smithii Mohr. = L. Smithii (Dicks.) Mohr. Lepidopilum splendidissimum Mitt. = Lamprophyllum_ splendidissimum (Mont.) Schimp. : Pterygophyllum denticulatum Mitt. = P. denticulatum (Hk. f. et Wils.) Mitt. Porotrichum confertum Mitt. = Thamnium confertum (Mitt.). Porotrichum latinerveS Mitt. — Thamnium latinerve (Mitt.). Porotrichum rigidum Mitt. = Thamnium rigidum (Mitt.) Broth. Lepyrodon parvulus Mitt. = L. parvulus Mitt. 446 Vv. F. BROTHERUS Fissidens asplenioides Hedw. = F. asplenioides (Sw.) Hedw. Mielichhoferia longiseta C. Mill. = M. longiseta C. Miill. Bartramia stricta Brid. =? Funaria hygrometrica Hedw. = F. hygrometrica (L.) Hedw. Macromitrium urceolatum Brid. =? Macromitrium fimbriatum Schwaeegr. =? Macromitrium saxatile Mitt. = M. saxatile Mitt. Glyphomitrium fernandezianum Mitt. = Ptychomitrium fernandezianum (Mitt.). Notarisia crispata Mont. =; Oncophorus cyathicarpus Mitt. = Amphidium cyathicarpum (Mont.) Broth. Oncophorus striatus Mitt. =? Zygodon obovalis Mitt. = Z. obovalis Mitt. Tortula flagellaris Mont. = T. flagellaris (Schimp.) Mont. Trichostomum longifolium Brid. =? Campylopus introflexus Brid. = C. introflexus (Hedw.) Brid. Campylopus leptodus Mitt. = Thysanomitrium leptodus (Mitt.). Dicranum Billardieri Schwagr. = Dicranoloma Billardieri (Schwaegr.) Par. Phascum nervosum Hook. = Pleuridium Robinsonii (Mont.) Mitt. Amphidum cyathicarpum....... AnacOlia SUDSESSIS & % u\ sce. sayeque Bariranbaparistatay. "5. o.¢-2, +. 5\2s™ 5 dense = wrertiandeziana (Gard)... 2. => \ [MERE Soke ECM Cni enmerianaee sue Brentelia Masaiuerae=, 2.5... . 2: Bryum fernandezianum-........ == ILGa nlc Sis re ene oe eee Canipylopusyaberrans 3). sire yw ie S—MIEEOGICIVON oe wots Soe) = HUE GOT CS Ft re rae Sa mtrOHexus =.) +! cee BPOUVETICHOTCES i fs eo ss SS SUDALCOCICtYON teyeh hf cyy) 23) . 03 SeMMRELULMICALUIS Geese lge Co P-L EP sp sy 3. * Catagoniopsis Berteroana....... Cyptodon crassinervis . Dendrocryphaea cuspidata ...... Dendroligotrichum dendroides Dicranella costata Cliewy ?: wi lat neshe fiefs = — capillifolioides ....... ee —saeaputolim. 83S res Se) — fernandezianum win eier am) =, a)... Index. AII Dicranoloma Menziesii |. 2/-92) saan 413 25 —( nigricauley’. 2)°5 eee 435 42 —* subelimbatum (1.1. 22 eeee 434 422 Ditrichumvaffine; 2ti-))- "205. see 410 414 — Jongisetiim.< 42.)-\s-:5) =) eee 410 415 Eriopus Srandiretis — = . -)-)). eee 436 416 — leptoloma . . =.5 >. eee 436 414 Eucamptodon perichaetialis ...... 416 414 Eustichia Poeppigii. 26). esa 435 415 Fissidens ‘asplenioides .1(2);-)- pean 418 414 —» CEassicUSPiS.:4.)-.© <5 pees 417 444 — fernandezianus . < ~,. -joss ame 417 428 —Teptochzte -.-*... ">. =) eaeaea ": aa 428 — maschalanthus. |-42°2"2ieeeee 417 445 — | pycnotylus > <2) cose 418 412 — |Tigidulus;«i42. sj hee 417 413 Funaria ,hygrometrica 7/5. ==) 423 413 Grimmia phyllorhizans . . ~ > ee). 419 413 Gymnostomum calcareum. ...... 4II 412 Hymenostomum Kunzeanum .... . 4II THE MUSCI Hypnodendron microstictum. . , yoni Leclerc gs) 6 ss Hypopterygium Thouini ... . Isopterygium fernandezianum Lamprophyllum splendidissimum Eeptodon Smithii....... 3 > . Leptodontium fernandezianum . Leptostomum Menziesii. . . Eepyrodon implexus ...... Bem PALVULUS) 5) “a Go stem sce aw BE — tomentosus .... Lopidium concinnum : Macromitrium fernandezianum . . — hymenostomum — Masafuerae. . Semmes ACE “sie Vat e)- Mielichhoferia longiseta . Neckera rotundata . . : Oligotrichum canaliculatum . . Oncophorus fuegianus ..... Phmonots clabrata .-.<.. . . meCralscandris) a ss. tbo wits = MeSGADMTOM Ase, 3.3 lw se SEAMS STL SUS) (07. iat \2,) ae Pinnatella macrosticta ..... Pleuridium Robinsonii ..... Polytrichadelphus magellanicus . Porothamnium arbusculans .. . ——MASCICHIAtHIM 4. . . « - - SMMITINEEVE) schist eye -2)15) 2 Psilopilum antarcticum .... Pterygophyllum anomalum .. . S—wedenticulatam . «4. -) MOUSCUEMIM srs <4 ee = fein A soa Ptychomnium falcatulum .... Seaptychocarplumn 2°. >...) EEnISubacrculares . 2. =\io) «6 Ptychomitrium fernandezianum . Rhacocarpus Humboldti. . .. . Or THE 444 439 437 440 437 430 418 424 429 428 JUAN FERNANDEZ ISLANDS Rhacomitrium convolutum . — lanuginosum.... ==). 1OMLORMIEY gener cee ke — striatipilum... 2... — subnigrium ..... — symphyodontum Rhacopilum fernandezianum . Rhaphidostegium aberrans . brachycladulum — caespitosoides — caespitosum .. . — Masafuerae . Rhizogonium mnioides — Novae Hollandiae . Rhynchostegium complanum — tenuifolium ... Rigodium arborescens . — hylocomioides . . — robustum ...-.. =——'* TOXaTION, 4) ws eA ee Sciaromium pachyloma .. . Stenomitrium pentastichum . Thamnium assimile. ... . =" Garoliee eaves =~ confertum (:'6). = — crassinervium..... =n Hin acre het as Maer — proboscideum .... ==) sLICiGUmy Fleet ay es Thuidium Masafuerae. .. . a AVAGINIAC SS .01s ey Thysanomitrium leptodus . . — Richardt <2... . Mortulastlace arise <5 ie == Scabrinmervis, -)-) sac Trichostomum brachydontium lope ruilagy ice ek er oe Weymouthia mollis... - . Zygodon Menziesii ..... a ODO NALS! oy cine elton 2 of . 448 Vv. F. BROTHERUS Explanation of Plates. Plate 26. 1—4. Fissidens fernandezianus Broth.: 7—2 sterile plant (*/1, '°/1); 3 leaf (75/1); 7 apex or dear (22/4): s—8. Fussidens crassicuspes Broth.: 5—6 sterile plant (*/, 10/1). 7 leaf (25/2); 6) apex of leann(229/3)) g—12. Fissidens pycnotylus Broth.: 9—zo sterile plant (1/1, 1/1); 77 leaf (78/4); 72 apex of leaf (29/1). 13—15. Leptodontium fernandezianum Broth.: 73 sterile plant (1/1); 74 leaf (1/1); 75 apex Ormleata(st/s): 16—20. Didymodon calymperidictyon Broth.: 76—r7 sterile plant (1/1, 1%/1); 78 leaves (75/1); zg apex of leaf (1/1); 20 base of leaf (7°91). 21—24. Grimmia phyllorhizans Broth.: 27 fertile plant (1/1); 22 leaf (?/1); 23 apex of leaf (59/1); 2¢ base of leaf (5/1). 25—28. Macromitrium fernandezianum Broth.: 25 fertile plant (*/1); 26 leaves (9/1); 27 apex of leaf (7/1); 28 base of leaf ('7/1). 29—32. Macromitrium Masafuerae Broth.: 29 sterile plant (*/1); 30 leaves (?°/1) 37 apex Of deafyG'/1); 92) base of leaf (279/s). 3335. Ltvchomnium falcatulum Broth.: 33 fertile plant (#/1); 34 leaf (/1); 35 apex of keane (Gaui) Plate 27. 1—2. Lriopus leptoloma Broth.; 7 leaves (?°/1); 2 apex of leaf (8/1). 3—4. Eriopus grandiretis Broth.: 3 leaves (#5/1); ¢ apex of leaf (°/1). s—6. Pterygophyllum tenuinerve Broth.: 5 leaf (?°/1); 6 apex of leaf (°/1). 7—8. Distichophyllum fernandezianum Broth.: 7 leaves (75/1); 8 apex of leaf (7/1). g—10. Distichophyllum subelimbatum Broth.: 9 leaves (25/1); zo apex of leaf (1°/1). 11-13. Weckera rotundata Broth,: 77 sterile plant ('/1); 72 leaf (75/1); 73 apex of leaf (*%/1). 14—15. Pinnatella macrosticta Broth.: 7g leaves (75/1); 75 apex of leaf (#°9/1). Printed 4/3 1024. Nat. Hist. Fuan Fernandez and Easter Isl. Vol. I. \ 4 Mey © ie \ AEIg 0 70b> SS \ Jeo9\\ Poo ‘ | HOO PLS: | KOO 2% } = —————— az rr a a j@ay TS ee. —— SS SE NE Wey eases aves Tererta (Lu iat} tl == ee ~~ } ¢ ——_————— ——/ —— a area ——— | OC Re sh ages OS Sa aime Ss \ 5 ——} 3 = ——— — ——— _———— AN a sa pee cee if LSJ \W7) ‘wer Hl ned NM | i) ‘M8 LU | 7 CES. fi SPOS Aly | ie | NN \|| ; Ove sere ttze 4)|| 100698259 ee LPO ESET LA ie y CEPR SA pace Bane de lit PLATE 26. o,f 00% PCOCy UBIO YN oY Kn NOONE OO 0 OG Ao0%C 005 ain lets al [ ace (Vod0ds n N06 N() 0 OCAU (4000 noc He 09.000 NS ()\ 0 10000000 Ng qqedoan0o4 Nooasa i nelod old YOMOOMIPMSaaeh qn 0690901 Ado HER eay090 Meseotetees k sagen O54 O0n900R090 009 On" ON ‘ LY IQO 1 [ () 00250 YUAN RSD Osi Moa secsze NeDegs WIS VOUT AHO 9O° OG 2007 Gndna YSs8i0% ORUNN, { 6000 Deval M Letnany i 008 Coal sang YOO, | Nd 80 YC \) i" i i jC OU oy ). ay ) \ | Aor py (h nM 28 wor ar wou aH vl) : noon oP or Sun tt | UDA YyovoqynoQyeyoconr DOgneutr oS gn0ndoy 0 WW Do0ue% of 1G i lil AULD) e525 Coe Se Bes 20Ooc SS OO r> > => ‘gy Ss 5 o S =o 050 0097 Ok RIVA) 0 of? is WOK 0h oP raya | 0! Me Hale ie I RNEGONUT EH LE IN cee 10007 EE BE OOO och a od RRO Ra HH ldoeeeon cane WY \\ Oya enccat™ We | “inn GeCgo noon tp AON 92 EOC hap T. Ekblom del. Vol. LT, Nat. Hist. Fuan Fernandes and Easter Isl. AF ie Wa gee A> yy « ZL a ss if} Ia PLATE: 27; OVO 00 DO OOK xOx 000550, Crome) ang eiexe OK OOKGOWOSO LA? SOMO QS ORI Oo OW. I vies \ NN I ANY 0920) () al, 2. 802050) N sy ae WRI 7 £49 (Bama T. Ekblom del. ies a { Peis " . 2 ¢ ‘ a “ ‘ Ya . ' Mio) i } : i 4 7 New Y 85 00235 5731 3 51 < an Se Aut inpenamtnipconibccen ak m ms a i, Whe ss ibipnkinpalnad ee aha eek * aces rs : Oo Oth an stnameyegiee eee oa je 4 ee ae . 9 oom embehe ® - ae