ARNOLD ARBORETUM HARVA RD UNIVERSITY ALFRED REHDER EDITOR [OSEPH H. FAULL anp CLARENCE E. KOBUSKI ASSOCIATE EDITORS eee = TABLE OF CONTENTS CONTRIBUTIONS TO THE FLORA OF THE NEW Hepripes; PLants Cot- LECTED BY 8S. F. KAsEwsxki 1n 1928 anp 1929. With one text figure. A. Guillaumin DR OR Bele Scie ee Sree wok ok ee a eee ee FN Se es Katmiopsis, A New Genus or ErtcaAcear FROM NORTHWEST AMERICA. With plate ay ey Alived Mehiler.s 2530s 8 ee Ree, Diaxy.ary LatTIcIrFEROUS ones oF BEAUMONTIA GRANDIFLORA. With pete Ai; Be Br ee. Woodworth oe ce is oe So A ESET CHROMOSOME NUMBERS jee, THE ANATOMY OF THE SECONDARY XYLE IN be OLEACEAE. With two text figures. By Karl Sax and Ernst C. Js eH Owe 650. C Wo ee 6 6 we ho wee oO 6 One Hes Ee eb 0a lee mw eee be & O08) * 8 ee 6 een ON THE PRECIPITIN REACTION IN Piants. I. THe Speciricrry OF nN NorM MAL PrecipitiIn Reaction. With plate 42. By Kenneth Proc! oe eb RE ee ee Cem Gn a eee ea ee ee THE OccURRENCE IN THE UNITED States or Cryptococcus Faer (Barr) DovG.., THe Insect Factor In A Menacine Disease oF Brscu. By J. ohn RACH, FSC ee ee en, Bek a Resta aie se CONTRIBUTIONS TO THE FLorA oF THE NEw Hesrives; Piants Cot- LECTED BY S., F. Kasewski In 1928 anp 1929. With plate 43 and 2 text figures) By A. Gralla te 6 o58 56 ake PO ote is CONTRIBUTIONS TO THE FLORA OF THE New HEBRIDES AND Santa Cruz IsLANDs; Orcnips CoLiectrepD By 8. F. Kasewski1 1n 1928 anp 1929. A NEW yaneg OF THE ORCHIDACEZ FROM THE NEW HEBRIDES. ‘By Oakes MONE ee OE ee oi cls 1 ee g's Ba come FER ey eee THE rates 0K eae pave feisetlag ae ebie leyborde Pamermms By Mat Sarl. o.oo ees CONTRIBUTIONS TO THE TRADESCANTIA PROBLEM. With _ 45 and 3 text figures. By cag Ey vesiesy gg te oo) | Be Se eee eee A Migs — TIVE STUDY & PHytopaTHora Diseases or Linac F THEIR conceit =. With plates 46 and 47 rye diagram. By. ouch pe ESPON ee ee SS ea eey Cae cee mee ey eee r sM oF MyxomyceTre PLASMODIA ON THE SPOROPHORES z HYMENOMYCETES. With pla plates 48 and 49 and two text figures. -Frank L. Howard and Mary E. Currie... .......2.-+.2--+ 20: ea Nores ON soOME CHINESE PLANTS. pda Franklin P. Metcalf.......... A PREVIOUSLY UNDESCRIBED VARIETY OF I RANTHUS VITELLINUs F. v. Muve.tit. From Norra NSLAND. By W. F. Blakeley.......... Nores on THe LigNeous nts Descripep By H. LEVEILLE FROM Eastern Asta. “By Alfred Rehder........ preaie PLE ER ISLE G) ‘1 New reabahieg, VARIETIES ia CoMBINATIONS FROM M THE HERBARIUM 52 75 AND THE ‘Cot.ections OF THE ARNOLD “ARBORETUM. By Alfred © ee 1 e Oe Clee te we Ss wre ee & & So ele ee ee ae ose 9 le © See 2 & Se ee oF © © Tan Crararcus Prostem. By Ernest J. Palmer.............-.---- | RELATIONSHIPS IN THE POMOIDEAE. With plate 50. ‘iy ME Pairing IN Sas Graces. With ~ 51 and one text ae Heue By Hala Toke Se ee 8 ess aed 2a: TABLE OF CONTENTS Meiosis AND CHIASMA FoRMATION IN PAEONIA SUFFRUTICOSA. plate 52 ae two text figures. By Karl Sar.....-.-.-++-+-+--+--005 7 or THE LiGNEous PLants Couiectep By J. F. Rock on = AnsouD ArporetuM ExpepiTion TO NORTHWE : ORTHEASTERN TIBET (Additions and Continuation). By — Alfred ehder and Clarence E. 385 CyroLocicaL StupiEs oF CoRNUS. With plate 53. By Haig Dermen... 410 LEAVES FROM A CoLLector’s Note Book. With one text goto By 4 Aly Ernest J. Palmer... 2. 02+. 206266 fo ba wer be eens ois ee ners ParasitisM OF Myxomy plate 54. By Frank c Howard aud Mary E. Currie. ......-.-++-++ 42 Nores. The Arnold Arboretum during the Fiscal year en ended June 30, 1932; The Arboretum; The Pathological Laboratory; The Cytological Lab The Herbarium; Library; Bibicenphy of Published — ry; The Writings of the Staff and Students i in 1931-1932. The Staff of the — Arnold Arboretum 1932-1933...........---2. eee eee cece eee cers JOURNAL OF THE ARNOLD ARBORETUM Votume XIII JANUARY, 1932 NuMBER 1 CONTRIBUTION TO THE FLORA OF THE NEW HEBRIDES PLANTS COLLECTED BY S. F. KAJEWSKI IN 1928 AND 1929* A. GUILLAUMIN With one text figure RUBIACEAE Nauclea sp. Aneityum: Anelgauhat Bay, rain-forest, alt. 60-450 m., no. 954 (coll. J. P. Wilson), Sept. 1929 (tree 12 m. high; diam. 0.5 m.; flowers yellow; fruit light brown, non edible). —Vernacular name “‘ Nepec.” Uncaria orientalis Guillaumin, sp. nov.—Fig. 4. Scandens, ramuli tetragoni angulis obtusatis, glabri vel sparse hirtelli, stipulae ovatae (1-1.3 cm. X 0.5—-0.8 cm.), lamina sparsis- sime hirtella vel glaberrima ad marginem puberula, folia ovato- apice subito lineari apice rotundato acuminata, membranacea, rigida vel coriacea, glabra vel infra in costa nervisque et praecipue ad nervorum angulos plus minusve hirtella, nervis 6-7-jugis valde prominentibus, nervis parallelis in venis perpendicularibus non prominentibus sed conspicuis, petiolo 1 cm. longo glabro vel sparse irtello, uncis magnis compressis ad 3 cm. longis, basi 5 mm. latis glabris vel sparse hirtellis, apice bracteis 3 ovatis 3-4 mm. longis, una aliis duplo (2 mm.) latiore ad marginem puberulis, pedunculo medio articulato bracteolatoque dense piloso vel plus minusve glabro, primum horizontali, postea decurvato, deinde (parte superiore et capitulo delapsis) uncato, capitulo fere 3 cm. diam. ebracteolato, calyce 2 mm. longo, ovario aequilongo dense rufo- velutino, lobis rotundatis tubo 3-plo brevioribus, corolla infundi- buliformi fere 1 cm. longa, tubo gracile cylindrico extra dense griseo rufo-velutino intus glaberrimo, lobiis rotundatis 2-3 mm. longis extra rufo-velutinis, intus basin versus linea media longi- tudinali pilosa excepta, glabris, stigmate apice cylindrice dilatato vix exserto. 1 Continued from Vol. XII, 264 (1931). 2 JOURNAL OF THE ARNOLD ARBORETUM [VoL. xm neityum: Anelgauhat Bay, rain-forest at sea level, no. 910, March 11, 1929 (common vine with sweetly scented, creamy flowers, growing on rain-forest trees). Eromanga: Dillon Bay, common in rain-forest, alt. 400 m., no. 348, June 1, 1928 (vine growing on top of trees; used by natives in case of necessity for making fire by rubbing thick part of vine against a hard wood).— Vernacular name “ Tevi-cow.’’ Fic. 4. Uncarta orrentaus Guillaumin (X 34). 4 Quite different from the species of New Guinea and the Bismarck 3 Islands which so far have been the most eastern species. It is easy — to recognize in these specimens the transformation of the inferior — Dolicholobium aneityense Guillaumin, sp. nov. 3 Arbor parva, 7 m. alta, ramis primum setulosis deinde glabres- : centibus, foliis ovato-lanceolatis (usque ad 14 em. X 6 cm.) atro- viridibus membranaceis apice acutis basin versus longe attenuatis, — Supra primum in costa setulosis deinde glabrescentibus, subtus — ‘Guillaumin in Compt. Rend. Acad. Sei. Paris, 192, p. 1264(1931). a 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 3 primum costa, nervis marginibusque setulosis, deinde costa excepta glabrescentibus, nervis 14—-16-jugis tenuibus subtus prominentibus, venis tenuissimis, petiolo gracili circa 2 cm. longo setuloso stipulis ellipticis apice rotundatis fere 2 cm. longis setulosis cito caducis; inflorescentia axillaris, cymosa, 4-5 cm. longa, 3—5-flora, floribus albis, pedunculo dense rufo-setuloso circa 1.5 cm. longo, calyx membranaceus, obconicus, 0.5 cm. longus, margine undulatus setulosusque, extra glaber, intus reticulose nervosus et basin versus appresse pilosus, corolla 3.5 cm. longa, tubo gracili cylindrico extra appresse setuloso intus glaberrimo, lobis 5 ovato-falcatis apice obtusis circa 6-7 mm. latis aequilongis, extra marginem versus sparse setulosis, antherae 5, sessiles, ad tubi apicem insertae, lineares, 2.5 mm. longae, apice rotundatae, basi parum sagittatae, tubi orem non attingentes, discus tubulosus, basin styli cingens, stylus gracillimus ad apicem valde dilatatus compressusque, antherarum apicem attingens, lobis 2 apice triangularibus, ovarium pedicello indistinctum, 2-loculare, ovulis in quoque loculo o. Aneityum: Anelgauhat Bay, common in rain-forest, alt. 450 m., no. 775, Feb. 18, 1929 (small tree up to 6 m. high, with white flowers and dark green leaves). Related particularly to D. graciliflorum Val. of New Guinea and to D. oblongifolium A. Gray of the Fiji Islands. Badusa occidentalis Guillaumin, sp. nov Arbor parva, 10 m. alta, ramis eardibas: foliis ovatis vel ovato- lanceolatis tusaie ad 15 em. X 5 em.) membranaceis apice breviter obtuseque acuminatis basi: cuneatis, nervis 6—8-jugis tenuibus subtus prominulis, venis inconspicuis, petiolo usque ad 2.5 cm. longo, stipulis interpetiolaribus tubulosis breviter mucronatis; inflorescentia corymbosa vel paniculatim corymbosa, foliis fere aequilonga, pedunculo 5-8 cm. longo compresso, bracteis foliaceis ad 2 cm. longis petiolatis, petiolo supra piloso, pedicello 1-1.5 cm. longo gracili apice decurvo, bracteolis triangularibus margine niinute erosis supra _pilosis, floribus niveis formosissimis, calyx clavatus, apice tubuloso-cupulatus, dentibus 5 minimis margine suberosis, intus velutinus, corolla tubulosa, tubo 5-6 mm. longo, extra glaber, intus medio leviter puberulo lobis 5 recurvis elliptice linearibus apice rotundatis, medio supra longitudinaliter costatis tubo aequilongis, stamina 5, 1.5 cm. longa, filamentis ima basi corollae insertis et tubum brevem formantibus, parte libera dimidio inferiore hispidis, antheris linearibus basi sagittatis, apice obtusis, discus dentibus 5 minimis obtusis circa styli basin, stylus staminibus aequilongus, apice valde clavatus, integerrimus, ovarium 2-loculare, malorinare _ Fructus clavatus, 11.2 cm. longus, tubo calycino s ¥. 4 4 JOURNAL OF THE ARNOLD ARBORETUM [vou. xn coronatus, septicide 2-valvis, seminibus valde compressis breviter alatis. 4 Eromanga: Dillon Bay, common in rain-forest at sea leve no. 393, June 8, 1928 (small tree up to 10 meters, with pretty snow- white flowers).—Vernacular name “ Ney in watu.” Only one more species of this genus is known which is restricted to Tonga, while another, B. philippica Vidal, non Blanco, remains doubtful. 4 Oldenlandia Crataeogonum Guillaumin in Lecomte, Not. Syst.mL_ 160 (1915). gy Eromanga: Dillon Bay, common in poor, red soil country, alt. 400 m., no. 352, June 1, 1928. Also in New Caledonia- Vernacular name “Noo lay yelong.” Mussaenda cylindrocarpa Burck in Ann. Jard. Bot. Buitenzorg, mit. 118, t. 17 (1883). Eromanga: Dillon Bay, common in rain-forest, alt. 460 m., no. 349, June 1, 1928 (small tree about 7 m. high with yellow flowers).—Also New Guinea and Bismarck Islands.—Vernacular name “Tear vess.”’ Mussaenda frondosa Seemann, Fl. Vit., 123 (1886); non Li naeus. — e Aneityum: Anelgauhat Bay, common on seashore, no. 922, March 17, 1929 (small tree up to 9 m. high with brown fruit wh ripe).—Already found in Aneityum. Identical with Polynesian specimens (Fiji, Tonga, Wallis, Sam: and Solomon Islands), but different from those of New Caledor and the Bismarck Islands and also from those of the Caroli Islands which Valeton considers as two distinct species. Me has suggested that this may be the same plant as M. philippiea Rich. of the Philippines. ; Chomelia (Tarenna) banksiana Guillaumin, sp. nov. : Arbor parva, 6 m. alta, ramis cylindraceis, glabris, raro foliatis, folia glaberrima, liptico-lanceolata, usque ad 19 em. X 8 z papyracea, basi apiceque acute cuneata, costa infra valde ned ig = : a ing ' longis) glabris apice acutis; inflorescentiae latiores, terminales, _ foliis multo breviores, ramis minutissi | | M™unito, calyx turbinatus, haud costatus, lobis bene distinctis _ Fotundatis dense ciliolatis, intus glaber, extra basin versus mi 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 5 sime adpresseque setulosus, 2 mm. longus, corolla tubo cylindrico lobis longiore extra minute adpresseque puberulo intus medio apicemque versus sparse barbato 5 mm. longo, lobis 5 patulis ellipticis apice rotundatis (3.5 mm. X 2 mm.) intus glabris extra ut tubus minutissime adpresseque puberulis, stamina 5, fauce inserta, filamentis brevissimis, antheris lineari-lanceolatis basi bre- viter sagittatis lobis leviter brevioribus, ovarium 2-loculare, quoque loculo 6-ovulato, stylo subulato 1 cm. longo tubi medio sparse barbato. Fructus globosi, 4 mm. diam. anks Group: Vanua Lava, common in rain-forest at sea level, no. 470, July 12, 1928 (small tree up to 6 m. high with white flowers). Remarkable for the distinct rounded lobes of its calyx and the small acute stipules. Chomelia sp. ‘Aneityum: Anelgauhat Bay, common in rain-forest alt. 60 m., no. 744, Feb. 12, 1929 (large tree up to 12 m. with black fruit). Eromanga: Dillon Bay, common in rain-forest, alt. 300 m., no. 290, May 23, 1928 (small tree up to 8 m.).—Vernacular name “ Deuv-yourit.” Randia sezitat Guillaumin in Arch. Bot. Mém. m1. no. 5, 9 (1930). Aneityum: Aname, common on seashore along w. coast, alt. 30 m., no. 991 (coll. J. P. Wilson) Sept. 1929 (small tree 15 cm. diam. with white flowers and brown fruit).—Also New Caledonia and Loyalty Islands.—Vernacular name “‘ Daramdaram.” Gardenia tannaensis Guillaumin, sp. nov. Arbor parva, 6 m. alta, foliis lanceolatis (usque ad 28 em. X 8.5 em.) apice basique acutis rarius apice rotundatis basi cuneatis, membranaceis, nervis 7—15-jugis, subtus prominentibus, venis numerosis, stipulis lanceolatis vel deltoideis 1 cm. longis basi tubum 0.5 cm. longum formantibus, petiolo 1.5-2 em. longo, gemmis resinosis, flores solitarii, breviter pedicellati, calyce obconico extra glabro, lobis 5, 1.5 cm. longis falciformibus acutis brevissime puberulis, corollae tubo sepalis aequilongo extra brevissime puber- ulo, lobis 5 tubo aequilongis. T anna: Lenakel, very rare in rain-forest soil, alt. 100 m. no. 33, Feb. 21, 1928 (small tree about 6 m.). Gardenia sp. Eromanga: Dillon Bay, common in rain-forest at sea level, no. 362, June 4, 1928 (shrub about 4 m. high with white flowers and with leaves which are used for healing by being placed on wound after being heated and rubbed).—Vernacular name “ Neace ya vot.” 6 JOURNAL OF THE ARNOLD ARBORETUM [VoL, XHI Gardenia sp. Aneityum: Anelgauhat Bay, common in rain-forest, alt. F 180 m., no. 913, March 17, 1929 (small tree up to 6 m. high, with yellow fruit). a This and the preceding species resemble each other in their falciform calyx-lobes. Gardenia sp. Banks Group: Vanua Lava, common in rain-forest at sea — level, no. 489, July 17, 1928 (small tree up to 8 m. high). Gardenia sp., praecedenti affinis. Eromanga: Dillon Bay, common in rain-forest, alt. 300 m., no. 369, June 5, 1928 (large tree up to 20 m. high).—Vernacular — name “ Nor-rooka.”’ These different species belong to the group with well developed calyx-lobes, not abundant in New Guinea, but well represented in A New Caledonia. Guettarda speciosa Linneaus, Sp. Pl. 991 (1753). 8 Aneityum: Anelgauhat Bay, common on seashore, no. 799, 4 Feb. 21, 1929 (tree 9 m. high; flowers white, sweetly scented). — Efate: Undine Bay, common in rain-forest along sea shore, no. — 210, April 25, 1928 (tree up to 20 m.; flowers white to cream). _E ro- manga : Dillon Bay, common in rain forest along sea shore, no. 363, June 4, 1928 (large tree up to 20 m. high).—Already found in Efate; also in New Caledonia, Australia (Queensland, North — Australia), Fiji, Samoa, Cook, Society, Union, Ellice, Gilbert, — Marshall, Caroline, Mariana, Solomon, Bismarck and Admiralty | Ms Gs ps) be =a) Avur ys Islands, New Guinea and Malaysia —Vernacular name ‘‘Ou-ven-— ou-ven.”” Guettarda Kajewskii Guillaumin, n. sp. Arbor medioeris, 5 m. alta, foliis obovatis (usque ad 11 cm. X 6 em.) atro-viridibus membranaceis basi cuneatis apice breviter ob- . tusissimeque acuminatis, nervis 5-8-jugis valde tenuibus parum _ conspicuis, venis inconspicuis, petiolo ad 1.5 cm. longo, stipulis . lanceolatis acutis 2-3 mm. longis extra adpresse argenteo-puberulis, citissime caducis, inflorescentiae densius cymosae, 1.5-8 cm, — edentatus, corolla 4—5 mm. longa, tubo linearibus 3 mm. | iformis, _— cylindrico extra praecipue — basin versus argenteo-puberulo, lobis 5 rotundatis 2 minoribus (an _ semper?), stamina 5, fauce inserta, sessilia, leviter exserta, antheris _ Aneityum: Anelgauhat Bay, common in rain-forest, alt. a | 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 7 30 m., no. 724, Feb. 11, 1929 (medium sized tree up to 12 m.; leaves dark green; corolla cream-colored). Guettarda sp. Aneityum: Anelgauhat Bay, common in rain-forest, alt. 210 m., no. 826, Feb. 28, 1929 (medium sized tree up to 9 m. high; wood hard; fruit purple when ripe). Plectronia odorata (Forst.) Hillebrand, Fl. Hawaii Isl. 175 (1888). Aneityum: Anelgauhat Bay, common in rain-forest, alt. 30 m., no. 726, Feb. 11, 1928 (tree up to 12 m. high; fruit yellow when ripe). Ef ate: Undine Bay, common in rain-forest, alt. 30 m., no. 191, April 23, 1928 (tree up to 20 m. high; fruit greenish purple when ripe).—Already found on Tanna; also New Caledonia, Loyalty, Fiji, Tonga, Society and Marquesas Islands.—Vernac- ular name “'Toolepau”’ (under no. 191). Plectronia sp. affinis P. barbata K. Schum. Tanna: Mt Tokosh Meru, common in tainicresk: alt. 300 m., no. 170, March 15, 1928 (tree about 20 m. high). Ixora aneityensis Guillaumin, sp. nov. Arbor parva, 7 m. alta, ramis sat magnis ut gemmis roseis, foliorum delapsorum cicatricibus valde prominentibus, _foliis sessilibus lanceolatis (usque ad 30 em. X 8.5 cm.) apice attenuatis basi auriculatis rigide membranaceis, costa subtus valde promi- nente, nervis 13-jugis subtus prominentibus, venis reticulatis _ subtus prominulis, stipulis fere 7 cm. longis triangularibus carinatis apice 3 mm. longe caudato-acuminatis, inflorescentia terminalis, 7 em. longa, paniculata, ramis brevissime puberulis, bracteis tri- angularibus longe caudato-acuminatis vel deficientibus, floribus albo-roseis pedicello 3-8 mm. longo vulgo apice subulate bracteolato brevissime puberulo suffultis, calyx ovatus, extra minute puberulus, 3 mm: longus, lobis 4 acutis brevissimis, corolla ad 2.5 cm. longa, lobis 4 lanceolatis extra subglabris tubo cylindrico extra minute puberulo leviter brevioribus, stamina 4, fauce inserta, filamentis brevibus, antheris acutissime lanceolatis basi sagittatis 6 mm. longis, stylus gracilis, apice fusiformis, antherarum apicem at- tingens. Aneity um: Anelgauhat hay, common in rain-forest, alt. 300 m., no. 851, March 2, 1929 (small tree up to 6 m. high; flowers pink-white; with pink buds and stems). This approaches especially I. montana Schlechter of New Cale- donia, but the leaves and the calyx are very different. iieekpess opulina D De Candolle, ‘Prodr. rv. 492 (1830). Vel Fi 8 JOURNAL OF THE ARNOLD ARBORETUM [vou. xa Aneityum: Anelgauhat, common in lower ranges, alt. 300 m., no. 969 (coll. J. P. Wilson), Sept. 1929 (small tree; flowers small, white; fruit a cluster of brown berries). Efate: Fila : Island, Vila, common in rain-forest along sea coast, no. 187, April — 14, 1928 (small tree 5 m., flowers white).—Already recorded from i Eromanga; also in New Caledonia and Loyalty Islands.—Vernac- i ular name “‘Natge” (under no. 969). a Pavetta sambucina De Candolle, Prodr. 1v. 492 (1830); non : Tarenna sambucina K. Schum. A Tanna: Lenakel, common in rain-forest soil, alt. 100 m., no. — 30, Feb. 21, 1928 (shrub up to 5 m.; flowers white).—Also in Fiji — and Society Islands. : These two species, extremely close, seem to me hardly distinct — from the var. tomentosa Hook. f. of New Guinea and Malaysia. Coffea arabica Linnaeus, Sp. Pl. 172 (1753). i Eromanga: Dillon Bay, common, growing wild in rain- — forest at sea level, no. 368, June 5, 1928 (common shrub gathered — Dy Cie we = to see if any difference takes place when it grows wild). Morinda citrifolia Linnaeus, Sp. Pl. 176 (1758). y Aneityum: Anelgauhat Bay, common in rain-forest at sea és level, no. 709, Feb. 9, 1928 (small tree up to 10 m. high; petals — white; fruit cream-colored when ripe, 5.5 cm. long, 4 cm. in diam.). _ Eromanga: Dillon Bay, common in rain-forest at sea level, — no. 260, May 15, 1928 (small tree up to 10 m. high; fruit cream- _ colored; sap of root used by natives as a red dye for grass skirts) .— Already found in Efate; also New Caledonia, Loyalty Islands, — Australia (Queensland, North Australia), Fiji, Tonga, Samoal ena) Society, Marquesas, Union, Ellice, Marshall, Caroline, — i : res lariana, Hawaii, Solomon, Bismarck and Admiralty Islands, New _ Guinea and Malaysia.—Vernacular name “Noah-i-rat.” f Morinda Forsteri Seemann, Fl. Vit. 129 (1866). z Aneityum: Anelgauhat Bay, common in rain-forest at 30 z m., no. 725, Feb. 11, 1929 (vine climbing up the rain-forest trees; _ fruit dark purple when ripe).—Also New Caledonia, Loyalty, Fiji, — _ Tonga, Samoa, Pitcairn, Gambier and Mariana Islands. . Calycosia Milnei A. Gray in Proc. Am. Acad. tv. 307 (1860). : Aneityum : Anelgauhat Bay, common in rain-forest at 180 — m., no. 817, Feb. 28, 1929 (small tree up to 6 m. high; flowers — white; fruit orange-colored when ripe, 2 em. long, 1.5 em. in diam., _ slightly flattened at each end). Tanna: Mt. Tokosh Meru, — : common in rain-forest at 800 m., no. 162, March 15, 1928 (small tree up to 8 m. high; flowers white). Eroman ga: Dillon Bay, 4 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 9 common in rain-forest at 400 m., nos. 315 and 315a, May 28, 1928 (small tree up to 7 m. high; flowers white; fruit. red).—Already found on Aneityum; also Fiji Islands.—Vernacular name “‘ Ney-in-war-tu.” Calycosia sp. Eromanga: Dillon Bay, alt. 300 m., rain-forest, common, no. 402, June 8, 1928 (small tree, 10 m. high; fruit red when ripe).— Vernacular name ““Nem-pel-ted.”’ Psychotria aneityensis Guillaumin, sp. nov. Frutex glaberrimus, 4 m. altus, ramis gracilibus viridibus, foliis glabris lanceolatis (usque ad 10 cm. X 3 em.) apice acutis basi anguste attenuatis membranaceis, nervis 8—12-jugis tenuissimis, petiolo 1-2 cm. longo, stipulis brevissimis deltoideis apice mucro- nulatis, inflorescentia axillaris vel pseudo-terminalis, cymose umbellata, 2 cm. longa, pedunculo gracili, floribus subflavis, 3—5- nis, pedicello capillari cirea 5 mm. longo, bracteis minutissimis, calyx hypocrateriformis, 1 mm. longus, lobis indistinctis, corolla 2 mm. longa, campanulata, lobis ovatis tubo fere 2-plo brevioribus, stamina inclusa, sub fauce inserta, filamentis brevibus, antheris 0.5 mm. longis linearibus apice truncatis basi leviter sagittatis, discus epigynus, tubulosus, stylus antheras leviter superans, basi gracili, apice dilatato, ramis 2 apice truncatis. Fructus pedicello circa 1 cm. longo suffultus, 6-7 mm. longus, pyrenis longitudinaliter suleatis. ~ Aneityum:Anelgauhat Bay, common in rain-forest at 30 m., no. 733, Feb. 11, 1929 (shrub up to 3.5 m. high; flowers cream- colored). Psychotria nacdado Guillaumin, sp. nov. Seandens, ramis rubris, foliis glabris ovato-lanceolatis (usque ad 14 em. X 5 cm.) apice acute acuminatis basi cuneatis mem- branaceis, nervis circa 9-jugis tenuibus subtus tantum prominulis vel immersis, petiolo 1.5—2 cm. longo, stipulis citissime caducis, axillis puberulis, inflorescentia terminalis, pyramidali-paniculata, usque ad 5 em. longa et 7 cm. lata, multiflora, floribus albis, pedi- cello 2-3 mm. longo dense papilloso, calyx campanulatus, 1.5 mm. longus, sparse papillosus, dentibus minutis triangulari-acutis, corolla infundibuliformis, usque ad 7 mm. longa, lobis lanceolatis apice intus leviter uncinatis, tubo extra praecipue basin versus dense papilloso intus dimidio superiore dense barbato, lobis extra sparse papillosis intus glabris, stamina exserta, fauce inserta, filamentis antheris fere 2-plo longioribus, antheris 1 mm. longis linearibus apice basique truncatis, discus epigynus, tubulosus, stylus -antheris aequilongus, apice dilatatus 2-fidusque. Fructus nigri. 10 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Aneityum: Anelgauhat Bay, common in lower hills up 300 m., no, 979 (coll. J. P. Wilson), Aug. 1929 (vine covering tr leaves medium; flowers small white, in clusters; fruit small b berries). Banks Group: Vanua Lava, common in rain-fo at 500 m., no. 485, July 16, 1928 (vine climbing over rain-fo trees; flowers white). This species is fairly close to P. rupicola Schlechter of N: Caledonia.—Vernacular name ‘‘ Nacdado”’ (under no. 979). =. Psychotria sp. ; a Banks Group: Vanua Lava, not common in rain-forests ip to 100 m., no. 411, June 5, 1928 (small tree up to 10 m. high). Psychotria sp. Aneityum: Anelgauhat Bay, sea level, sea shore, comm no. 926, March 17, 1929 (small tree, 7 m. high; leaves dark glo green; fruit 8 mm. long, 6 mm. diam., red when ripe). COMPOSITAE Vernonia cinerea (L.) Lesser in Linnaea, rv. 291 (1829). Eromanga: Dillon Bay, common in rain-forest cleari at 300 m., no. 384, June 8, 1928 (common weed throughout Island).—Already found on Aneityum, Erronan, Tanna Eromanga; also New Guinea, Bismarck, Solomon, Fiji, Sam Cook, Marshall, Caroline and Mariana Islands, New Caledon Australia (Queensland, North Australia, New South Wa Moluccas and Malaysia.—Vernacular name “Sow-any-longa.”’ Zealand, Fiji, Samoa Cook, Marquesas, Marsha ) » wamoa, Cook, Ma 8, Marshall, Caroline, marek and Solo: mon Islands, New Guinea, Moluccas and J laysia.—Vernacular name “Ou lakkidine.” di : -Blumea densiflora De Candolle, Prodr. y. 446 (1836). Efate : Undine Bay, comm on in rain-forest at 200 m., no. April 27, 1928 (about 2 m. high growing in clearings; flo yellow).—Also New Caledonia, Fiji and Bismarck Islands, Guinea and Malaysia. is Ct ee eT a, VI. 160 (1831). Wedelia biflora De Candolle in Wight, Contrib. Boe Iggia, 18 (1834 ‘Bfate: Undine Bay, common in rain-forest soil of cocos 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 11 plantation at sea level, no. 196, April 24, 1928 (up to 3 m. high; flowers yellow; leaves used by natives to give flavor to fish). Banks Group: Vanua Lava, common in cleared rain-forest land at sea level; no. 460, July 12, 1928 (up to 2 m. high; flowers bright yellow; a troublesome weed).—Already found on Aneityum, Tanna, Eromanga and Efate; also New Caledonia, Loyalty Islands, Australia (Queensland, New South Wales, North Australia), Norfolk, Fiji, Tonga, Wallis, Samoa, Ellice, Marshall, Caroline, Mariana and Solomon Islands, New Guinea, Malaysia. GOODENIACEAE Scaevola cylindrica Schlechter & Krause in Engler, Pflanzenr. Iv.-277, p. 129 (1912). Aneityum: Anelgauhat Bay, | common in poor red _ soil country, no. 911, March 11, 1929 (bush up to 3 m. high; flowers white, sweetly scented). Eromanga: Dillon Bay, common in red soil of poor bracken country, at 300 m., no. 293, May 23, 1928 (flowers white).—Also New Caledonia.—Vernacular name “ Wibe Wibe”’ (under no. 293). Scaevola frutescens (Mill.) Krause in Engler, Pflanzenr. 1v.— 277, p. 125 (1912). neityum: Anelgauhat Bay, common in rain-forest at sea level, no. 786, Feb. 19, 1929 (common shrub up to 5 m. high, growing along the sea shore; flowers white; fruit white when ripe, 0.75 cm. long, 1 cm. in diam.).—Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland, North Australia), Fiji, Tonga, Santa Cruz, Samoa, Cook, Society, Pau- mota, Union, Ellice, Gilbert, Marshall, Caroline, Mariana, Hawaii, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia. VACCINIACEAE Vaccinium Macgillivrayi Seemann in Jour. Bot. 1. 77 (1864). Aneityum: Anelgauhat Bay, common in poor red soil country at 60 m., no. 701, Feb. 4, 1929 (shrub up to 4 m. high; _ petals white, stamens yellow; fruit purple when ripe, eaten by the natives). Eromanga, Dillon Bay, common in poor red soil of bracken country at 300 m., no. 301, May 24, 1928 (small shrub about 3 m. high; fruit black, 1.25 em. long, 1 cm. in diam., eaten by the natives) —Already found on Aneityum.—Vernacular name “Autarm-tell”’ (under no. 301). EPACRIDACEAE gon oymbulae Lebiltantiere, Sert. pees Cees a6, 1. Leucopo: 39 ase). 12 JOURNAL OF THE ARNOLD ARBORETUM [vor. xm | Aneityum: Anelgauhat Bay, common in rain-forest at 4 50 m., no. 702, Feb. 4, 1929 (crooked tree up to 10 m. high, growing in gullies; corolla white). Eromanga: Dillon Bay, common in | poor red soil of bracken country at 400 m., no. 351, June 1, 1928 (shrub up to 3 m. high; leaves bound to a pole are used as brooms).— _ Already found on Aneityum; also New Caledonia, Fiji Islands.— 4 Vernacular name ‘‘Nom-pul-low”’ (under no. 351). . MYRSINACEAE Maesa efatensis Guillaumin in Bull. Soc. Bot. France, Lxvi. 272 920). Aneityum: Anelgauhat Bay, common in rain-forest at | sea level, no. 792, Feb. 20, 1929 (small tree up to 6 m. high; fruit pink-cream when ripe, 0.5 em. long, 0.7 cm. in diam.). Tanna: _ Lenakel, common in heavy rain-forest soil at sea level, no. 9, Feb. 20, 1928 (tree up to 20 m. high, dark green with brown veins _ underneath).—Already found on Efate. a Rapanea modesta Mez in Engler, Pflanzenr. iv. 236, p. 367 (1902). A Efate: Undine Bay, common in rain-forest at 200 m., no. 237, April 28, 1928 (small tree up to 12 m. high; fruit purple when ripe; wood pink when freshly cut).—Also in New Caledonia. Rapanea sp. Aneityum: Anelgauhat Bay, common in rain-forest at 60 m., no. 751, Feb. 12, 1929 (large tree up to 12 m. high; leaves dark green with light yellow midrib; fruit black when ripe, 1 cm. long, 0.75 em. in diam.). Tapeinosperma Kajewskii Guillaumin, Sp. nov. Ramulis gracilibus (8 mm. diam.) primum sparse tomentellial deinde glaberrimis, foliis obovatis (usque ad 14 em. X 5.5 cm.) | apice obtusis basi longe attenuatis sessilibus rigide membranaceis- nervis a venis subtus reticulatis parum distinctis, punctis nigris — minimis; inflorescentiae 2-3 em. longae, ad ramulorum apicem — congestae, ramis 2-3 rubiginosis et nigro lineatis leviter furfuraceis — vel glabris, pedicellis 2 mm. longis, floribus 3 mm. longis sordide roseo-purpureis, sepalis ovatis apice obtusis basi tantum coalitis, margine furfuraceo-ciliatis, petalis late ovatis apice rotundatis basi ¥% coalitis bene punctatis, staminibus petalis minoribus, antheris — late ovatis epunctatis, filamentis nullis, ovario ovoideo, stylo ees distincto, stigmate peltatim discoideo. Fructus globosi (1.2 cm. diam.), apice umbonati, longitudinaliter brunneo-lineati. Aneityum: Anelgauhat Bay, common in rain-forest at 210 _ m., no. 822, Feb. 28, 1929 (petals a dirty pink-purple with small PIN Sees AE eh circ ee tate Paras) Alicea MP iy he Mee ty ats 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 13 purple spots on the outside; fruit 1 cm. long, 1 em. in diam. with elongated brown spots). Tapeinosperma netor Guillaumin, sp. nov. Arbor parva (ultra 7 mm.) vel alta (30 m.), ramulis sat gracilibus (circa 5 mm. diam.) glaberrimis, foliis lanceolatis (usque ad 20 cm. x 5 cm.) apice acutis basi in petiolum circa 1 cm. longum longe attenuatis membranaceis, nervis 16—19-jugis, tenuissimis immersis creberrime punctatis, inflorescentiae 8-9 cm. longae, graciles, ad ramulorum apicem congestae, pinnatim paniculatae, pedicellis circa 0.5 cm. longis, floribus 3 mm. longis, sepalis ovatis apice rotundatis tertia infima parte coalitis margine rigida breviterque ciliatis, petalis roseis ovatis apice obtusis basi 4 tantum coalitis apice manifeste punctatis basi lineatis, staminibus petalis minoribus, antheris ovatis dorso valde punctatis, filamentis brevibus, ovario subulato. Fructus maturitate rubri, 1.5 cm. diam. Aneityum: Anelgauhat Bay, common in rain-forest at 240 m., no. 768, Feb. 12, 1929 (small tree up to 6 m. high; fruit red when ripe, very similar to shape of a cherry, 1 cm. long, 2 cm. in diam.). Eromanga: Dillon Bay, common in rain-forest at 400 m., no. 313, May 28, 1928 (tree up to 30 m. high, growing about 6 miles inland among very tall rain-forest trees; petals pink).— Vernacular name “Net-or” (under no. 313). These two species are rather closely related though they can be distinguished at the first glance; they are near T. grande Mez of the Fiji Islands. SAPOTACEAE Sideroxylon (Planchonella) aneityense Guillaumin, sp. nov. Arbor 15-20 m. alta, ramis gracilibus primum sparse rufo-pilosis cito glabris, cortice brunneo abunde longitudinaliter striolato, gemmis rufo-pilosis, foliis lutescentibus membranaceis lanceolatis (usque ad 10 cm. x 3.5 cm.) utrinque glabris apice acutis basin versus longe attenuatis, nervis 7-9-jugis a venis reticulatis parum distinctis, petiolo gracili 1-2 cm. longo. Flores albo-virides, grave- olentes, in axillis supremis dense fasciculati, pedicellis tenuibus ad 1 em. longis sparse rufo-pilosis, calycis segmentis ovatis 1 mm. longis apice rotundatis extra sparse rufo-pilosis, corollae tubo” 1 mm. longo lobis ovatis aequilongo, staminodiis filiformi-subulatis, staminibus filamentis sub fauce insertis, antheris ovatis aequl- longis, ovario breviter lateque conico dense rufo-lanuginoso, stylo brevissimo cylindrico glabro. Aneit yum: Anelgauhat Bay, common in rain-forest at 300 m., no. 771, Feb. 14, 1929 (large tree up to 15 m. high; leaves light 14 JOURNAL OF THE ARNOLD ARBORETUM [vou xm petals greenish white, sweetly scented); common in rain-— 4 ; art at 300 m., no. 945, March 19, 1929 (large tree up to 18 m. high; — fruit immature; wood durable).—Vernacular name “Tnretchar™ a (under no. 945). Near S. acutum Krause of New Guinea. 4 Sideroxylon (Planchonella) tannaense Guillaumin, sp. nov. | bor 5 m. alta, trunco 75 cm. diam., ramis sat validis cinereis, foliis -membranaceis ovatis ae ad 15 ecm. x 7 em.) ntringaeay j dicellis kensibes 6-7 mm. Lesage Semelebonaealuate ‘iqvanbalis ovatis, — 1 mm. lenge dorso dense tomentosis, corollae tubo lobis breviore, March 6, 1928 (tree up to 15 m. high, 75 cm. in diam.; fruit eaten by natives or so reported). g Near S. “Sohne K. Schum. of New Guinea and perhaps ob 4 also of the Moluce Sideroxylon sp. | Aneityum: Anelgauhat Bay, common in rain-forest at. 150 m., no. 936, March 17, 1929 (large tree up to 15 m. high; lea dark green; fruit immature; a good commercial timber for wood 1 exposed to weather). Eromanga: Dillon Bay, common rain-forest at 300 m., no. 385, June 8, 1928 (large tree up to 25 high with a very hard wood; leaves light brown Peerere sg Vernacular name “ Ney-mor” (under no. 385). Aneityum: Anelgauhat Bay, common in rain-forest at | m., no. 941, March 1, 1929 bai x tree up to 15 m. high; fruit aaceEd: Sion kaen Dillon Bias. i common in poor red ail, bra country, at 300 m., no, 298, May 24, 1928 (small shrub up to 5 m. high; fruit purple when ripe).—Vernacular name ‘“‘Nor-sum.” Mimusops parvifolia R. Brown, Prod. Fl. Nov. Holl. 1. 531 (18) Aneit yum: Anelgauhat Bay, common in rain-forest at 30 no. 730, Feb. 11, 1929 (large tree up to 6 m. high; flower wh young stems covered with brown a eb etyberd cba A 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 15 Bay, common in rain-forest at 150 m., no. 266, May 17, 1928 (tree up to 20 m. high; fruit light red, bark rough and fissured). Efate: Fila Island, Vila, common in rain-forest at sea level, no. 179, April 12, 1928 (tree about 15 m. high reputed by natives to be poisonous).—Also New Caledonia, Loyalty Islands, Australia (Queensland, North Australia), and New Guinea.—Vernacular name “ Yatarhwu”’ (under no. 266). Manilkara dissecta (R. Br.) Dubard in Ann. Mus. Colon. Mar- seille, sér. 3, m1. 13 (1915). neit y u m:Anelgauhat Bay, scarce inthe lower hills at 150 m., no. 983 (coll. J. P. Wilson), Sept. 1929 (large tree up to 1.30 m. in diam.; leaves small, elongated; flowers pink; fruit small, brown, only on extreme tips of branches; timber very hard, dark red, too scarce to be of commercial value); common in the rain-forest at 150 m., no. 937, March 17, 1929 (large timber tree up to 21 m. high; fruit immature; one of the best durable hard timbers on the island). —Also Tonga Islands.—Vernacular name “ Niping.”’ Palaquium neo-ebudicum Guillaumin, sp. nov. Arbor 25 m. alta, ramis validis primum rubiginoso-villosis deinde glabris, foliis rigide membranaceis ovatis (usque ad 12 cm. x 7 cm.) primum rubiginoso vestitis deinde glabris apice rotundatis basi cuneatis, petiolo 2.5-3.5 em. longo, nervis 9—10-jugis bene distine- tis, venis oblique parallelis. Flores ad ramulorum apicem, infra folia, dense fasciculati, pedicellis 1.5-3.5 cm. longis rubiginoso- tomentosis, calycis lobis ovatis, exterioribus 4 mm. longis, interiori- bus duplo brevioribus dorso rubiginoso-vestitis, corollae glabrae tubo brevi, lobis oblongis obtusis, antheris fere sessilibus anguste lanceolatis, dorso dense rubiginoso-pilosis, ovario glabro, stylo subulato valde elongato. Aneityum: Anelgauhat Bay, common in rain-forest at 60 m., no. 756, Feb. 13, 1929 (large tree up to 18 m. high; outside of flower- buds covered with brown hairs). Eromanga: Dillon Bay, common in rain-forest at 400 m., center of the island, no. 344, June 1, 1928 (large tree up to 25 m. high; heart wood yellowish brown).— Vernacular name “Ney-more-yetu” (under no. 344). - Bassia Kajewskii Guillaumin, sp. nov. Arbor ultra 20 m. alta, trunco 35-45 cm. diam., ramis crassis, foliis obovatis (usque ad 15 cm. x 9 cm.) atro-viridibus pergamenta- ceis apice rotundatis basi sensim cuneatis, nervis 10—12-jugis validis cum totidem intermediis, venis reticulatis, petiolo valido 1.5 em. longo. Flores terminales fasciculati, pedicellis 1 cm. longis gracilibus appresse a calycis laciniis oblongis apice obtusis 16 JOURNAL OF THE ARNOLD ARBORETUM Bs. XI extra dense appresse pilosis $ mm. longis, corollae tubo 1 mm. long gO lobis ovatis fere aequilongo glaberrimo, staminibus 16, antheris fere = sessilibus anguste sagittatis, connectivo intus extraque dense lanu- ginoso, ovario conico stylo tereti coronato. : T anna: Lenakel, common in rain-forest soil at 150 m., no. Feb. 23, 1928 (tree up to 20 m. high, 35-45 cm. in diam.; leaves dark — green above, light green underneath; sap milky; fruit eaten by thee ) natives). ay Near B. bawun (Scheff.) Guillaumin, comb. nov., of New Guin Bassia obovata Forster, Fl. Ins. Austr. Prod. 35 (1786). Aneityum: Anelgauhat Bay, common in rain-forest at flesh white inside, seeds large, black; fruit very luscious, eaten by t natives). Banks Group: Vanua Lava, common in rain-fore . at 100 m., no. 464, July 10, 1928 (large tree up to 20 m. high; flowers cream-colored; fruit eaten by the natives).—Already fount on Tanna and Ffate. Bassia sp., affinis B. obovata Forst. Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 833, Feb. 1, 1929 (large tree up to 12 m. high; fruit brown, immature on specimens; wood durable). SAPOTACEARUM GEN.? Aneityum: Anelgauhat Bay, common in rain-forest at 1 m., no. 935, March 17, 1929 (large tree up to 18 m. high; leaves dé glossy green; fruit immature, covered with dense brown bine EBENACEAE Maba buxifolia Persoon, Syn. Pl. 11. 606 (1807). : Eromanga: Dillon Bay, common in rain-forest at 300 m., 391, June 8, 1928 (tree up to 15 m. high; natives say fruit is red ripe; kernel of the nut is eaten by natives). Aneityum: A gauhat Bay, common in rain-forest at 150 m., no. 932, March 1 1929 (medium-sized tree up to 15 m. high; fruit immature; woe used for native houses).—Also New Caledonia, Loyalty, Tongo at Caroline Islands, New Guinea, Malaysia.—Vernacular name “ Mung-lei” (under no. 391). SYMPLOCACEAE Symplocos ancityensis Brand in Engler, Pflanzenr. rv.-242, p- Ane thy wa: Anelgauhat Bay, common in rain-forest at sea 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 17 level, no. 720, Feb. 11, 1929 (large tree up to 18 m. high; flowers white; wood used by natives for canoe paddles). Eromanga: Dillon Bay, common in poor red soil, bracken country, at 300 m., no. 295, May 24, 1928 (tree up to 10 m. high; flowers white; wood used in making paddles for canoes; this tree is found 4 or 5 miles inland).—Already found in Aneityum.—Vernacular name “ More- lease”’ (under no. 295). OLEACEAE Jasminum didymum Forster, Fl. Ins. Austr. Prod. 3 (1786). Eromanga: Dillon Bay, common in rain-forest at sea level, no, 259, May 15, 1928 (fruit black when ripe; used by natives as rope or cordage for binding their houses as it lasts a long time).— Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland, North Australia), Fiji, Tonga, Samoa, Cook and Society Islands, New Guinea.—Vernacular name *“No-sori.” Jasminum simplicifolium Forster, Fl. Ins. Austr. Prod. 3 (1786). Eromanga: Dillon Bay, common in rain-forest at 300 m., no. 404, June 8, 1928 (vine growing up to the tops of rain-forest trees; fruit black when ripe).—Already found on Aneityum; also New Caledonia, Loyalty Islands, Australia (Queensland, New South Wales, North Australia), Norfolk, Fiji, Tonga and Samoa Islands.— Vernacular name “ No-sori.” APOCYNACEAE Melodinus neo-ebudicus Guillaumin, sp. nov. Alte scandens, ramis gracilibus 2 mm. diam. rubris, foliis oppositis lanceolatis (usque ad 9 cm. x 3 em.) membranaceis integris apice longe acuminatis basi. obtusis, nervis tenuibus crebris parallelis, petiolo gracili ad 4 mm. longo. Inflorescentiae axillares, cymosae, 3-5-florae, 1.5-3.5 cm. longae, pedicellis gracilibus usque ad $ mm. longis apice 2-bracteolatis, sepalis 5 ovatis obtusis 1 mm. longis, corollae tubo roseo 5 mm. alto extra glabro intus tertia iafima parte glabra, media sparse pilosa, suprema dense barbata, lobis 5 patulis ovatis (4 mm.x 2 mm.) albis, staminibus 5 anguste lanceolatis 1mm. longis sessilibus ad 2/8 tubi insertis, ovario anguste ovoideo — attenuato, stylo tereti i stigmateque conico apice 2-aristato inclusis 3 mm. longo. Fructus globosus 2 em. diam. maturitate luteus, sicco atro-rubro-brunneus, pruinosus. A neit yum: Anelgauhat Bay, common in rain-forest at 30 m., no, 729 (type of the flower), Feb. 11, 1929 (vine growing over rain- forest trees; petals white, corolla stem [? tube] pink). Eroman- ga: Dillon Bay, common in rain-forest at 300 m., no. 392 (type of 18 JOURNAL OF THE ARNOLD ARBORETUM [vor. xm fruit), June 8, 1928 (vine growing over rain-forest trees; fruit yellow , when ripe). To be placed near M. buzifolius Baill. and M. celastroides Baill, : of New Caledonia which probably constitute only one species.— _ Vernacular name “‘ Nosarak” (under no. 392). Alyxia efatensis Guillaumin, sp. nov. Alte scandens, ramis sat crassis, foliis 3-nis lanceolatis (usque aa 10 cm. x 3.5 em.) rigide membranaceis apice obtusis vel acutis rarissime acuminatis basi cuneatis, venis immersis tenuibus crebis — fere rectis, petiolo circa 1 em. longo. Inflorescentiae ad ramulorum | 4 apicem axillares, ad 5 cm. longae, umbellatim cymosae, foliis Q ; parvis (2.5-3 cm. x 0.9-1 cm. ) oppositis, apice bracteatae, floribus i luteis 3 in pedunculo communi 5 mm. longo sessilibus vel 5 exteriori- _ bus pedicello 2.5 mm. longo suffultis, tantum inapertis 2 cm. longis, — bracteola 1 lanceolata ima calycis basi, sepalis 5 lanceolatis 2.5 mm. _ longis apice molliter ciliolatis, corollae tubo 1.2 cm. longo cylinem drico apice leviter strangulato et intus sub staminibus breviter bar-_ bato, lobis brevibus (5 mm. longis) anguste ovatis, staminum fila- — mentis antheris 2-plo brevioribus sub fauce insertis, antheris tubi apicem attingentibus, locis fere parallelis apice obtuse mucronatis, — ovario conico apice attenuato, stylo longo staminum basin attine gente tereti apice leviter dilatato (?). E fate: Undine Bay, common in rain-forest at 500 m., no. 231, April 28, 1928 (vine climbing up to tops of rain-forest trees; flowers a yellow). : | 2 The most closely related species, though differing in its pubescent — ovary seems to be A. laurina Gaudich. from the islands Rawak ant i Gebeh of the Moluccas. : Alyxia, sp. nov.? , Banks Group: Vanua Lava, common in rain-forest at 300 My no. 477, July 12, 1928 (vine growing over rain-forest trees). Anotee te PL 208 (1753). @ [oss Anelgauhat Bay, common in rain-forest at sea ria no. 806, Feb. 21, 1929 (common serub tree up to 12 m. high; "a ; wers sweetly scented, petals dirty brown with white edges; . ruit 9 cm. long, 5.5 cm. in diam. , purple when ripe).—Already found — : = Tanna and Efate; also New Caledonia, Loyalty, Fiji, Tonga, Te ered: ma an Gambier, Marshall, Caroline, Mariana m, Bismarck and A. svat ae dmiralty Islands, New Guinea, a Ochrosia aebties Guillaumin, sp. noy. : Arbor parva, 7 m. alta, hictsedens: ramis gracilibus, foliis sts aren shat ee dere ine is" PL ee eS ae Pat nines 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 19 rigide membranaceis linearibus (usque ad 12 cm. x 1 em.) apice attenuatis basi longe attenuatis, nervis immersis creberrimis rectis, petiolo circa 1 cm. longo gracile. Inflorescentiae axillares, circa 3 cm. longae, ramosae, papillosae, bracteis ovatis, pedicello subnullo, sepalis ovatis 1.5 mm. longis extra papillosis dorso carinatis, corolla (tantum juvenili) sepalis aequilonga ad medium 5-loba, lobis ovatis, antheris sagittatis sessilibus tubo fere aequilongis, ovario conico lateraliter compresso medioque sulcato, stylo subulato aequi- longo. Fructus (2.5 cm. x 2 cm.), maturitate rubri, drupis 2, epicarpio tenuissimo, endocarpii una facie convexa, altera applanata medioque sulcata. Aneityum: Anelgauhat Bay, common on sea shore at sea level, no. 925, March 17, 1929 (small tree up to 6 m. high; fruit 2.5 cm. long, 1.75 cm. in diam., red when ripe; sap milky). Very remarkable on account of its flowers and leaves which quite resemble those of A. bracteolosa Rich. ex A. Gray var. angustifolia A. Gray from the Tonga Islands. Ochrosia elliptica Labillardiére, Sert. Austro-Caled. 25, t. 30 (1824). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 693, Feb. 4, 1929 (small tree up to 6 m. high, sap milky; leaves dark green; fruit red when ripe, 3.75 cm. long). Ero man- ga: Dillon Bay, common in rain-forest close to sea beach, at sea level, no. 360, June 4, 1928 (large shrub up to 6 m. high; fruit red).— Also New Caledonia, Australia (Queensland), Fiji, Tonga and So- ciety Islands.—Vernacular name “‘ Naugh-comb” (under no. 360). Ochrosia parviflora Henslow in Ann. Nat. Hist. 1. 345 (1838). Ochrosia elliptica K. Schumann, non Labillardiére. Aneityum: Anelgauhat Bay, common on seashore at sea level, no. 947, March 19, 1929 (medium sized tree up to 9 m. high; flowers white; fruit 8 cm. long, 5.5 cm. in diam.). Eromanga: Dillon Bay, common in rain-forest at sea level, no. 405, June 8, 1928 (large tree up to 15 m. high; flowers white; seed eaten by the natives). E fate: Undine Bay, common in rain-forest at sea level, no. 207, April 25, 1928 (tree up to 20 m. high; white sap flows — freely from branches but not from the trunk).—Also Fiji, Solomon, Bismarck and Admiralty Islands, New Guinea, Aru Islands, Malaysia.—Vernacular name “ Yefat” (under no. 405). Alstonia Seem. form. calvescens Markgraf in Bot. Jahrb. LxI. 198 (1927). Aneity um: Anelgauhat Bay, common in rain-forest at 60 m., no. 740, Feb. 12, 1929 (small tree up to 9 m. high; flowers white), 20 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Eromanga: Dillon Bay, common in rain-forest and red soil, bracken country, no. 292, May 23, 1928 (small tree up to 8 m. high; — flowers white; fruit hangs down like a long appendage, 25 cm. long). —Also Bismarck and Key Islands, North Australia, Malaysia.— Vernacular name “ Ne-yev-vi-are”’ (under no. 292). | Ervatamia orientalis (R. Br.) Turrill in Jour. Linn. Soc. XLim. — 32 (1915). . Tabernaemontana orientalis R. Brown Fl. Nov. Holl. Prodr. 468 (1810). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 692, Feb. 4, 1929 (small tree up to 10 m. high, sap milky; _ leaves bright green; flowers white; fruit brownish red when ripe). _ T anna: Lenakel, common in heavy rain-forest soil at 45 m., no. 13, Feb. 20, 1928 (small tree with numerous branches; leaves — i" bright green; flowers white).—Already found on Aneityum, Tanna, — Eromanga and Efate; also Loyalty Islands, Australia (North _ Australia, Queensland, New South Wales), Fiji, Tonga and Society ae Islands, New Guinea, Malaysia. a Parsonsia neo-ebudica Guillaumin, sp. nov. Scandens, foliis oppositis rigide membranaceis vel sub-coriaceis — ovatis (usque ad 12 em. x 6 cm.) apice breviter acuteque acuminatis — basi rotundatis glaberrimis, nervis immersis 8-9-jugis arcuatis, — petiolo 2-3 em. longo. Inflorescentiae axillares, 4-6 cm. longae, 4 laxiflorae, %11-florae, pedunculo 3-4 em. longo, ramis 0.5-1.5 em. longis, bracteis minimis subulatis, pedicellis usque ad 5 mm. longis ut pedunculo glaberrimis, calycis lobis acute triangularibus, 1.5 mm. — longis glaberrimis, corolla 8 mm. longa, lobis anguste lanceolatis — tubo sub-2-plo longioribus infra supraque faucem dense deorsum — 2 . ta, staminum filamentis 2 mm. longis, antheris filamentis — -plo longioribus basi longe sagittatim appendiculatis, disco 5-— crenato ovario aequialto, ovario glabro, stylo gracili, stigmate paeveye dilatato antherarum tertiam supremam partem atten- : < Eromanga: Dillon Ba: co = 306. Ju : > mmon in i -» NOs ! 19 re rain-forest at 300 m no. 3 ASCLEPIADACRAR a 1266 (9) Mebarirrcc it Pi 215 ‘ecciemtibaraaat 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 21 Aneityum: Anelgauhat Bay, common in open country at sea level, no. 930, March 17, 1929 (small weed up to 1 m. high; flowers orange and yellow; the natives say this is indigenous to the islands as they have a native name for it). Eromanga: Dillon Bay, common in open grassy country at 300 m., no. 388, June 8, 1928 (weed up to 1 m. high; flowers red; suspected poisonous weed). —Already found on Aneityum; also New Caledonia, Loyalty Islands, Australia (Queensland), Tonga, Samoa, Cook, Society, Marquesas, Marshall and Mariana Islands, Malaysia, Hawaii (introduced).—Vernacular name “‘Te-vess”’ (under no. 388). Hoya australis R. Brown apud Traill in Trans. Hort. Soc. vu. 28 (1830). Aneityum: Utgi, plentiful on seashore at 90 m., no. 1002, (coll. J. P. Wilson), Sept. 1929 (vine; leaves medium, light green, fleshy ; flowers small, white; fruit long spike-like pods, 12.5 cm. long, 1.2 cm. in diam.).—Already found on Tanna and Efate; also Queensland, Fiji, Samoa, Solomon? and Bismarck?? Islands.— Vernacular name “‘ Napalht.” The specimen cited differs from the description of R. Brown only in the petals being entirely papillose on their whole upper surface. The young parts (branchlets and the leaves on both sides) are copiously villous which corresponds with Seemann’s remarks (Fl. Vit. p. 163) concerning his H. pilosa. It would be very desirable to make a complete revision of the specimens referred to this species. Schlechter (in Bot. Jahrb. x. 118) has already drawn attention to the confusion particularly in regard to the specimen of Naumann from the Solomon Islands which he considers as the type of a distinct species: H. Naumannii Schlechter, while Engler refers it to H. australis R. Br. Tylophora aneityensis Guillaumin, sp. nov. sae Scandens, ramis gracilibus flexuosis teretibus glabe , foliis (in sieco) rigide membranaceis ovatis (usque ad 6 cm. x 4 cm.) apice attenuatis mucronatisque basi rotundatis et saepius glandu- losis, petiolo 1.5-2 cm. longo. Inflorescentiae extra-axillares, 5-9 em. longae, paniculatae, parte florifera 0.5 cm. longa racemosa, — bracteis numerosis minimis ciliatis imbricatis, floribus subflavis numerosis, pedicello capillari, corolla fere aequilongo, calycis lobis ovatis obtusis 1.5 mm. longis margine sparse ciliola - basi tentem coalitis, corollae lobis ima basi tantum coalitis lineari-ellipticis (8 mm. x 3 mm.) apice rotundatis 7-nervis, coronae squanus eee angulari-ovatis 1 mm. altis obtusiusculis, carnosis antherarum basin superantibus, antheris parallelipipedis ree Ae 22 JOURNAL OF THE ARNOLD ARBORETUM [voL. XI ‘3 4 om pendice hyalina semi-orbiculari, polliniis globosis, translatoribus _ brevibus, retinaculo ovoideo-elongato polliniis aequilongo, stigmatis — capite plano stamina haud superante. Folliculus (tantum imma- turus) globosa-compressus (2 cm. x 1.8 em.), apice apiculatus, — pedicello 6 mm. longo. a Aneityum: Anelgauhat Bay, common on sea shore at sea level, no. 796, Feb. 21, 1929 (vine growing over undergrowth on f sea shore; petals cream-colored). ; The most closely related species seems to be T. Brackenridgti A. Gray from the Fiji Islands. The genus was already known fromm a the Torres Islands by a specimen not ay determined. 4 Tylophora tannaensis Guillaumin, sp. n Scandens, ramis gracilibus flexuosis aie. glaberrimis, fois (in sieco) membranaceis ovatis (usque ad 6.5 cm. x 4 cm.) apice — mucronatis rotundatis vel etiam emarginatulis basi rotundatis glandulosisque, petiolo circa 2 cm. longo. Inflorescentiae extra-_ -axillares, usque ad 8 cm. longae, umbellatae, rarius paniculatim — q | umbellatae, pedunculo communi 1.5—2.5 em. longo, glabro, bracteis minimis extra puberulis floribus albis numerosis, pedicello copillasiay corolla 2-plo longiore vel ultra sparsissimeque puberulo deinde — glabrescente, calycis lobis late ovatis apice rotundatis 2 mm. longis | dorso sparse puberulis basi tantum coalitis, corollae lobis ima basi tantum coalitis lineari-ellipticis (1 cm. x 3.5 mm.) apice rotundatis — a 7-nervis, coronae squamis ovato-acutis, 1.5 mm. altis acutis carnosis, — ad antherarum medium attingentibus, antheris parallelipidedis basi A obtusis apice appendice hyalina semi-orbiculari, polliniis globosis, — translatoribus brevibus, retinaculo ovoideo-elongato, stigmabill u capite plano stigma non superante. o Tanna: Lenakel, common in rain-forest at sea level, no. 20, Feb. 21, 1928 (vine climbing up on rain-forest trees; flowers white) 3 Extremely close to the preceding species, but different in its inflorescence, the much longer pedicels and the presence of soft — ¢ scattered hairs on the pedicels and the back of the sepals and in the > acute apex of the scales of the corolla. : oe balsa See fo} “eerie LOGANIACEAE Geniostoma rupestris Forster, Char. Gen. 24, t. 12 (1776). a : bach: febes : Anelgauhat Bay, common in rain-forest at 30 M-»_ 0. 728, Feb. 11, 1929 (small tree up to 9 m. high; leaves dull, dark green; petals white).—Already found Samos, Cook and Society Islands ound on Tanna; also Fiji, th Aneityum:Anelg ‘hat i siege ‘y, common in rails forest at 100m 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 23 no. 734, Feb. 11, 1929 (large tree up to 18 m. high; leaves dark green, flowers yellow). Couthovia neo-ebudica Guillaumin, sp. nov. Arbor parva, 12 m. alta, foliis ovatis (13-20 cm. x 6-10 em.) basi cuneatis apice rotundato-obtusis vel latissime acuminatis, sat crassis, petiolo pro genere gracili 1.5-3 cm. longo, nervis 5-6-jugis, venis immersis, stipulis interpetiolaribus majusculis coriaceis bre- vissime obtusis, cum petiolis in cupulam usque ad 1 cm. altam coalitis nunquam fissis. Inflorescentia composite cymosa, peduncu- lo robusto 1.5 em. longo, ramis I. ord. 7—7.5 cm. longis brevissime sparsissimeque tomentellis, II. ord. 2-3 cm. longis et. III. ord. 3-8 mm. longis densius longiusque tomentellis, pedicellis nullis vel brevissimis et dense tomentellis, bracteis minimis suborbicularibus margine dense ciliatis 1 mm. longis corolla. . . staminibus : , ovario ovoideo-elongato in stylum indistincte attenuato, stigmate capitato. Fructus élongati (fere 4 cm. x 1 cm.), inferne in stipitem curvatum attenuatis, apice sensim longe acuteque at- tenuati. Aneityum: Anelgauhat Bay, common in rain-forest up to 180 m., no. 774, Feb. 18, 1929 (small tree up to 12 m. high; flowers sweetly scented; fruit when gathered 4 cm. long, 1 cm. in diam.). To be placed near C. densiflora K. Schum. and C. terminalioides Gilg & Bened. from New Guinea. BORAGINACEAE Cordia Myxa Linnaeus, Sp. Pl. 190 (1753). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 691, Feb. 4, 1928 (small tree up to 15 m. high; fruit pinkish rown when ripe, 4 cm. long, 3 em. in diameter). Eromanga: Dillon Bay, rain-forest, 300 m., common, no. 397, June 8, 1928 (fruit red when ripe; bark used for grass skirts and berries to feed pigs). Tanna: Lenakel, common in rain-forest at 100 m., no. 42, Feb. 22, 1928 (tree 7 to 10 m. high; fruit pink when ripe).— Already found on Tanna; also New Caledonia, Australia, (Queens- land), Society and Marquesas Islands, New Guinea, Malaysia.— Vernacular name “Yalehoi” (under no. 397). Tournefortia argentea Linnaeus f., Suppl. 153 (1781). | Aneityum: Anelgauhat Bay, common on sea shore at sea level, no. 795, Feb. 20, 1929 (small tree up to 9 m. high; flowers white). Tanna: Lenakel, common on sea shore at sea level, no. 78, March 3, 1928 (tree about 10 m. high with very spreading top). Eromanga: Dillon Bay, common along sea coast at sea level, no. 406, June 8, 1928 (small, well-branched tree up to 10 m. high, only ts 24 JOURNAL OF THE ARNOLD ARBORETUM [voL. XIII along sea shore). Banks Group: Vanua Lava, common on sea shore at sea level, no. 425, July 6, 1928 (small tree up to 8 m. high; common on all the shores of the New Hebrides).—Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland, North Australia), Fiji, Tonga, Samoa, Cook, Society, Paumota, Union, Ellice, Gilbert, Marshall, Caroline, Mariana, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia, Hawaii.—Vernacular name “‘Ney-in-pori” (under no. 406). CONVOLVULACEAE Ipomoea congesta R. Brown, Fl. Noy. Holl. Prod. 485 (1810). Banks Group: Vanua Lava, common in rain-forest at sea level, no. 449, July 9, 1928 (climber; flower very pretty, puce- colored with a white base; grows freely both in scrub and planta- tions).—Already found on Tanna and on Efate; also New Caledonia, Australia (Queensland), Norfolk, Fiji, Tonga and Mariana Islands, New Guinea, Hawaii. ? Ipomoea denticulata Choisy in Mém. Soc. Phys. Genéve, v1. 447 (1838). Eromanga: Dillon Bay, common in open grass country at 300 m., no. 320, May 28, 1928 (vine growing over small bushes; patter: ie bgree —Already found on Aneityum and Efate; also tralia (Queensland), Tonga, Fiji, Society, Marshall, Solomon laa Bismarck Intends, New Guinea, Malaysia. —Vernacular name **Niye-eni-usak-wavi. Ipomoea pes-caprae Roth, Nov. Pl. Sp. 109 (1821). Aneityum: Anelgauhat Bay, common on sea shore at sea level, no. 815, Feb. 26, 1929 (creeping vine growing on the sand close to the sea beaches hout the entire group). Eromanga: Dice’ has « common on sea beach at sea level, no. 332, May 29, 1928 (vine growing close to the sea and in some plantations adjoining the sea it is a pest; flowers purple).— Also New Caledonia, Australia (Queensland, New South Wales, North Australia, West Australia), New Zealand, Fiji, Tonga, Samoa, Cook, Marquesas, Ellice, Marshall, Mariana, Solomon, Bismarck and — Islands, Malaysia, Hawaii ii Vernacular name “Novi- vu’’ (under no. 332. - Ipomoea turpethum R. Brown, Fl. Nov. Holl. Prod. 485 (1810). Banks Group: Vanua Lava, common in rain-forest at sea level, no. 450, July 9, 1928 (common in the New Hebrides; flower large, snow white).—Already found on Ancityum, Tanna and Ero- _ manga; also New Caledonia, Loyalty Islands, Australia (Queens- a | land), OL, Toman. Samoa, Bociny aed Merman neem) lars 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 25 SOLANACEAE Datura arborea Linnaeus, Sp. Pl. 179 (1753). T anna: Ikiti, common in rain-forest at 100 m., no. 175, March 16, 1928 (shrub up to 4 m. high; flowers cream). South American species already introduced to Australia (West Australia) and to Tahiti. GESNERIACEAE Cyrtandra aneiteensis C. B. Clarke in De Candolle, Monog. Phan. v. 282 (1883). _Descriptioni adde: Corolla sub-flava extra ad tubi apicem lobisque ula, ceterum glabra, tubo cylindrico 12 mm. longo, lobis 5 rotundatis, 3 posterioribus 4 mm. longis, 2 anterioribus 3 mm. longis, staminibus 2 ad apicem tubi insertis, filamentis antheribusque aequilongis annulum horizontalem formantibus, antheris albis apice coherentibus. Aneityu m:Anelgauhat Bay, common in rain-forest at 180 m., no. 773, Feb. 18, 1929 (plant up to 1.80 m. high, growing in rain- forest clearings; petals creamy green).—Already found on Aneityum. Except that the calyx is still persistent on the young fruit, all the other characters agree with the description by Clarke. The hitherto unknown corolla is described above. cymosa J. & G. Forster, Char. Gen. 6 (1776). Bia ks G rou p: Vanua Lava, common in rain-forest at 200 m., no. 452, July 10, 1928 (small tree up to 3 m. high; flowers creamy yellow) -—Already found on Aneityum and Tanna. The specimen differs from the type in the much longer petioles (up to 7 em. long) scarcely undulate on the margin. | Cyrtandra Kajewskii Guillaumin, sp. 2 : Frutex 2 m. alta, ramis novellis dense ‘ide Sieoas, deinde glabris, foliis c oppositis aequalibus elongato-ellipticis (usque ad 30 cm. x 11 3 em.) basi longe attenuatis apice obtusis margine serratis membrana- - Ceis, novellis dense rufo-hirsutis deinde margine et costa subtus re cred aky glaberrimis, nervis circa 12-jugis, petiolo 2-3 em. longo, as Inflorescentiae petiolo aequilongae, pedunculo sub- — null, haacteis Giaelecs (1 cm. x 0.4 cm.) margine pilosis, Pagina inferiore hirsutis, pedicellis fasciculatis 1 cm. longis sparse _hirsutis, calyce 7 mm. alto ad tertiam partem fisso intus extraque _ glaberrimo, lobis 5 lanceolatis acuminatis, corolla extr a ad tubi apicem lobisque sparse villosa, ceterum glabra, tubo 2 ate longo _ apicem versus dilatato, lobis rotundatis, 5 mm. longis, staminibus 2 ad tabi ghee: insertis, ‘filamentis brevissimis horizontaliter 26 JOURNAL OF THE ARNOLD ARBORETUM [vou. xmT annularibus, antheris. . . , ovario glaberrimo in stylum attenu- ato 1.5 em. longo, apice sparsissime hirsuto, stigmate discoideo peltato, disco tubuloso 2 mm. alto apice 5-undulato. Fructus, tantum novelli, fusiformes, basi calyce cincti. T anna: Lenakel, common in rain-forest soil at 200 m., no. 62, Feb. 24, 1928 (shrub about 2 m. high, found on banks of gullies). This species is similar to C’. aneiteenis C. B. Clarke but is clearly distinct by its much larger flowers with a tube enlarged above and by the perfectly glabrous calyx. ACANTHACEAE Hemigraphis reptans T. Anderson pee Hemsley, Rep. Sci. Res. Voy. Challenger, Bot. 1. pt. m1, 173 (1885 Aneityu m:Anelgauhat Bay, common in rain-forest at 240 m., no. 903, March 11, 1929 (small plant up to 80 or 100 cm. high; flowers pink).—Already found on Aneityum, Tanna, Efate, Ero- manga and Espiritu Santo; also New Caledonia, Loyalty, Tonga, Marshall, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia. Acanthus ilicifolius Linnaeus, Sp. Pl. 639 (1753). Ef ate: Undine Bay, common at sea level in rain-forest on salt water creek, no. 195, April 24, 1928 (growing in large clumps or _ patches, the canes of which have a sprawling habit; flowers puce- ; colored; fruit brown when ripe).—Already found on Tanna; also New Caledonia, Australia (Queensland, North Australia), Bismarck Islands, New Guinea, Malaysia. : SIG Litt, Pe EOIN 4 Pseuderantt laxiflorum (Gray) Hubbard in Rhodora, xvm. ; 159 (1916). < Tanna:Lenakel, common in native gardens, rain-forest at 150 m. ee ; estes epee 3, ar (shrub about 4 m. high; flowers white _ with violet centers).—Already found E i Bi bien Di dy found on Eromanga (ined.); also Pseuderanthemum sp. : 00 oA, uly 1, I Lava, common in rain-forest at ry m., no 12, 1928 (plant up to 75 em. high; petals white VERBENACEAE Stachytarpheta indica Vahl, Enum. 1. 206 (1805). Tanna: Lenakel, common in rain-forest soil at 100 m., on all — tie cas ree a ra (common weed 1—1.5 m. high; flowers brigh — ound on New recive Samoa, — Caroline Islands, Malaysia. ragtond ee a 1932] -GUILLAUMIN, FLORA OF THE NEW HEBRIDES 27 Vitex Negundo Linnaeus, Sp. Pl. 638 (1753). Aneityum: Anelgauhat Bay, common on seashore at sea . level, no. 801, Feb. 21, 1929 (shrub up to $ m. high, growing on the beach; leaves silvery underneath; flowers blue; fruit yellow when ripe).—Also New Caledonia, Mariana Islands, Malaysia. Vitex rapinioides Guillaumin, sp. nov. Arbor usque ad 20 m. alta, trunco 60 em. diam., ligno luteo, ramis luteis primum puberulis, foliis usque ad 15 cm. longis, 3-5 foliolatis, foliolis ovatis (9 cm. x 5 cm., 7 cm. x 4 cm., 3 em. x 1.5 cm.) basi obtusis vel cuneatis apice acuminatis rigide membranaceis, nervis 5—9-jugis, petiolulis 1 cm., 0.5 cm., 0.25 em. longis, petiolo 2.5-5 em. longo, primum lamina, petiolis, petiolulo rufo-puberulis, deinde lamina glabra infra glanduloso-punctata, petiolulis et petiolo glabrescentibus. Inflorescentia cymosa, ad 4 cm. longa, 3-flora, pedunculo gracili 1-1.5 em. longo sparse rufo-puberulo, apice brac- teis 2 filiformibus 2-3 mm. longis rufo-puberulis munito, pedicellis 1.5-2 em. longis, lateralibus medio bracteolis 2 oppositis rufo- glanduloso-puberulis 1 mm. longis munitis, floribus roseis, calyce campanulato integro 4 mm. alto extra dense rufo-glanduloso margine rufo-lanuginose 4-penicellato, corolla circa 2 cm. longa extra basi excepta dense rufo-puberula, tubo circa 1.5 cm. longo intus supra minum insertionem lanuginoso, lobis rotundatis, majore 6 mm. longo, intus lanuginosis, staminibus exsertis, filamentis 3 mm. supra basin insertis lanuginosis, ovario globoso glabro 4-loculari, stylo _elongato, apice 2-fido. Fructus rubri, depresse globosi (1.3 cm. x 1.5 em.), calyce patellato 6 mm. diam., putamine circa 5 mm. crasso 4-loculari. Aneit yum: Aname, west coast, common on seashore to 15 m., no. 992 (coll. J. P. Wilson) (type), Sept. 1929 (large tree to 60 cm. in diam.; leaves medium; flowers small, pink, bell-shaped; fruit red 1.2 em. long, 1-5 em. in diam.). Eromanga: Dillon Bay, com- mon in rain-forest at 300 m., no. 299, May 24, 1928 (large tree up to 20 m. high; fruit red; common in the southern group; wood very hard and used by the natives in house building). E fate: Undine Bay, common in rain-forest at 100 m., no. 211, April 26, 1928 (tree with yellow wood and very faulty, up to 20 m. high; flowers pink; fruit pink). Vernacular names “Incdic” (under no. 992) and “Nay-mofsi” (under no. 299). 3 ; _ Nearest to V. Rapini Beauv. of New Caledonia which has the leaves much thinner and of different shape. The specimen from ‘Efate is remarkable for its more elongated oval-lanceolate or - 28 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm Vitex trifolia Linnaeus, Sp. Pl. 638 (1753). A neity um: Anelgauhat Bay, common on the sea shore at sea level, no. 690, Feb. 4, 1929 (small plant up to 25 cm. high growing on sand beach; leaves silver-green; flowers light blue). Tanna: Lenakel, not common, in rain-forest soil at 200 m., no. 66, Feb. 24, 1928 (small tree about 4 m. high); common in native gardens of rain- forest soil at 150 m., no. 84, March 3, 1928 (shrub about 4 m. high; leaves variegated with white; flowers blue).—Already found on Aneityum, Tanna, and Epi; also New Caledonia, Australia (Queens- land, North Australia), Norfolk, Fiji, Tonga, Samoa, Cook, Marquesas, Caroline, Mariana and Bismarck Islands, New Guinea, Malaysia, Hawaii. Faradaya neo-ebudica Guillaumin, sp. nov. _ Alte scandens, cortice fulvo, foliis oppositis late vel latissime ovatis (usque ad 14 cm. x 11 em.) apice brevissime subitoque acumi- natis vel subrotundatis basi cuneatis, leviter coriaceis, nervis 4-6- jugis, venis reticulatis ut nervis subtus tantum prominulis, petiolo 1-4 cm. longo. Inflorescentiae axillares, 5~7 cm. longae, centrifuge corymboso-dichotome cymosae, sparse puberulae, bracteis lanceo- latis (1 cm. x 0.3 cm.) basin versus longe attenuatis in utraque pagina rufo-puberulis, floribus albis, centrali pedicello rufo puberulo lem. longo suffulto, calyce primum clauso sparse puberulo deinde in lobos 4 valvatim fisso tunc 8 mm. longo, corollae lobis 1 exteriore majore, 1 intermedio leviter minore, ut exteriore margine integer- rimo, 2 interioribus minoribus margineque suberosis, tubo 3-plo brevioribus, staminibus 4 medio tubi insertis leviter exsertis, an- theris in alabastro inflexis ellipticis, loculis parallelis, ovari iter : : ‘icls, parallelis, ovario breviter 4-lobo 4-loculari, loculis l-ovulatis, ovulis alte insertis, stylo sta- minibus breviore, stigmate 2-lamellato. Aneityum: Anelgauhat Bay, common in rain-forest at 120 a m., no. 813, Feb. 29, 1929 (vine growing to the tops of rain-forest — i trees; flowers white) I believe that on account of the calyx and the me tenses hk he calyx an ovary, notwith- — g the fact that the scarcely exserted stamens are inserted at : 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 29 the New Hebrides; flowers cream).—Already found on Efate; Also New Caledonia, Loyalty, Marshall, Mariana, and Bismarck Islands, New Guinea, Malaysia. Clerodendron inerme Gaertner, Fruct. 1. 271 (1788). E fate: Fila Island, Vila, common on seashore at sea level, no. 185, April 13, 1928 (shrub about 3 m. high; flowers white).—Also New Caledonia, Loyalty Islands, Australia (Queensland, New South Wales, North Australia), Fiji, Tonga, Samoa, Caroline, Mariana, Santa Cruz, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia. LABIATAE Ocimum sanctum Linnaeus, Mant. 85 (1767). Tanna: Lenakel, common in native gardens of rain-forest, at 200 m., nos. 64 and 65, Feb. 24, 1928 (herb. about 33 cm. high).— Already found on Efate and Espiritu Santo; also Australia (North Australia), Solomon, Bismarck and Mariana Islands, New Guinea, Malaysia. Coleus scutellarioides Bentham in Wallich, Pl. As. Rar. 1. 16 (1831). Eromanga: Dillon Bay, common in poor red soil, bracken country, at 400 m., no. 316, May 28, 1928 (small plant up to 75 em. high; flowers blue).—Already found on Tanna, Eromanga, Efate and Espiritu Santo; also New Caledonia, Australia (North Aus- tralia) Bismarck Islands, New Guinea, Malaysia.—Vernacular name “Qu-lair-ung.” Plectranthus Forsteri Bentham, Lab. Gen. et Sp. 38 (1832). Eromanga: Dillon Bay, common on sea shore at sea level, no. 258, May 15, 1928 (small plant about 50 cm. high; flowers pale blue; used by natives for sores, sap taken and mixed with salt water).—Already found on Aneityum, Tanna and Eromanga; also Isle of Pines, Fiji and Samoa Islands. (To be continued) _ Museum nartionat D'HISTOIRE NATURELLE, Parts. 30 JOURNAL OF THE ARNOLD ARBORETUM [vox. xm : FS KALMIOPSIS, A NEW GENUS OF ERICACEAE FROM & NORTHWEST AMERICA ALFRED REHDER a Plate 40 3 : THE DISCOVERY in Oregon by Mr. and Mrs. Leach of a new : ericaceous shrub subsequently described by L. F. Henderson as — Rhododendron Leachianum' is highly interesting, particularly as — this shrub turns out to be a new genus related to Loiseleuria and Kalmia and also to Rhododendron for which I propose the name Kalmiopsis referring to its general resemblance to Kalmia polifolia Wangenh. In its inflorescence it agrees closely with Kalmia poli- folia Wangenh. except that it has alternate instead of opposite — bracts. The inflorescence in both species is a short raceme terminat- — inglast year’s branchlets; the bracts decrease in size toward the Apex, the lowest being more or less leaf-like; the persistent bractlets sur-_ round the pedicels at the base and together with the bract act ase protecting scales for the flower-buds during the winter, The calyx | 2 in both species is 5-parted, rather large, colored and gibbous at — the base, forming a ring around the immersed apex of the slender — pedicel. The corolla is similar in size and color in both species, but in Kalmiopsis more campanulate and lacking the peculiar pouches 3 of Kalmia. The capsule i is subglobose and thin-walled and the valves bifid at the apex in both genera. In Kalmiopsis the under — side of the leaves and the inflorescence are dotted with lepidote_ glands absent in the genus Kalmia. Loiseleuria has the same in- florescence as Kalmia polifolia though reduced to fewer flowers, the De eB a Ti en bo 1 9 aa Sesion are dehiscent by longitudinal slits and the ovary is 2-3-celled. “The E capsule is similar to Kalmia and Kalmiopsis. Like Kalmia it lacks the glandular dots on the leaves and inflorescence. Rhododendron differs chiefly in its seeds which are enclosed in a loose and thin elongated testa lacerated or fringed at the ends and in the capsule which is more or less elongated, has woody walls and the valves 0 bifid at the apex. Its inflorescence, too, is very different, being @ umbel-like raceme sometimes reduced to one flower, enclosed i scaly winter-buds before anthesis except in the subgenus Therorho- dion which has a one- to two-flowered inflorescence termina leafy shoots of the current year; it may well be considered a dist genus as proposed by Small, connecting Rhododendron with Rho 1 L. F. Henderson in Rhodora, xxxm. 205 (1931). 1932] REHDER, KALMIOPSIS A NEW GENUS OF ERICACEAE $1 thamnus. ‘The flowers in Rhododendron are more or less zygomor- phous while in Kalmiopsis as in the two genera mentioned above they are actinomorphous. The peculiar lepidote glands are copi- ously present in the numerous species of the subgenus Eurhododen- dron Endl. except its section Leiorhodion Rehd. In its inflorescence and in the structure of its flowers Kalmiopsis shows a rather close relation also to Phyllodoce Salisb. and Rhodothamnus Reichenb. Both differ in the absence of lepidote glands on the under side of the leaves and on the inflorescence; Phyllodoce also in the crowded linear strongly reflexed leaves, the ovoid to campanulate corolla and the elongated style, and Rhodothamnus Reichenb. in the slightly zygmorphous flower, rotate corolla, long exserted style and long- ciliate leaves. The new genus represents apparently, like Loiseleuria and Rhodo- thamnus, a phylogenetically old type of the Tertiary. The age of these three monotypic genera is also shown by the fact that they do not exhibit any variation, while Rhododendron, Kalmia and Phyl- lodoce, genera probably derived from these monotypic genera, have split up into many species, Rhododendron being apparently the youngest and most unstable in its specific characters. Loiseleuria being an arctic plant has still a wide circumpolar distribution, while amnus is a relict of the European Alps and Kalmiopsis of the higher mountains of Northwest America. The last named is the most restricted, having been found so far only in two or three locali- ties in the Siskiyous, a mountain range which shelters some other interesting local relicts as Picea Breweriana S. Wats., a species whose nearest relations are found now in southeastern Europe and Eastern Asia. Also Quercus Sadleriana R. Br. Campst. which is restricted to the same region has no close relation in western North America. Kalmiopsis, gen. nov. Calyx 5-partitus, persistens, basi gibbosus, segmentis oblongo- Ovatis; corolla late campanulata, 5-loba, lobis late ovatis obtusis; stamina 10, corollae subaequilonga, antheris dorsifixis poris apicali- bus dehiscentibus, filamentis filiformibus basi leviter dilatatis et ciliolatis; discus tenuis, obscure 10-lobatus; stylus brevis rectus, stigmate capitato leviter lobulato; ovarium hemisphericum, 5-locu- lare, placentis angulo interiori medio adnatis bilobis multi-ovulatis; capsule subglobosa, crustacea, septicide 5-valva, valvis apice bifidis axi placentifero solutis; semina ovoidea, utrinque obtusa, testa firma leviter reticulata.—Frutex parvus, sempervirens, gemmis pau- ciperulatis, foliis coriaceis breviter petiolatis subtus glanduloso- lepidotis » margine minutissime spinuloso-ciliolatis, floribus in axillis | $2 JOURNAL OF THE ARNOLD ARBORETUM tron Si ‘| My bractearum in apice ramulorum in racemum congestis, graciliter : pedicellatis, pedicellis basi bracteolis persistentibus suffultis, fructu — ; erecto. Genus monotypicum, proximum Loiseleuriae Desv., Kalmiae L. ae $ Rhododendro L. A primo praecipue differt staminibus 10, antheris poris apicalibus dehiscentibus, ovario 5-loculari, foliis alternis sub- — tus glandulosa-lepidotis; a secundo staminibus ab initio liberis, non antheris in foveolis corollae inclusis, foliis subtus glandulosa-lepi- — dotis; a tertio capsula subglobosa, parietbus tenuibus, valvis me bifidis, seminibus ovoideis utrinque obtusis, testa firma. is Ss distare videtur a Phyllodoce Salisb. quae foliis congestis linearibus valde revolutis et corolla ovoidea vel campanulata differt, e Rhodothamno Reichenb. qui floribus leviter zygmorphis, core rotata, stylo exserto, foliis longe ciliatis et ut genus praced defectu glandularum lepidotarum recedit. Species unica Americae boreali-occidentalis incola. Kalmiopsis Leachiana sincere comb. no Leachianum L. enderson in ih odorkd © fo" Me s Satve'sae a Oe. 82% OF eo — deatiscine a Od a ee " eo % gee be oe Pee Se Fig. 1 Fig. 2 Fig. 3 (Explanation in the text) Some pollen grain measurements were made and practically 1 t difference in size was found between diploid and tetraploid species. rrewia with only n = 9 chromosomes had pollen grains which wer equal to or larger than those of Tilia, Gaiser (1930) lists a number of species from related families Tiliaceae. The original references were studied to find out whe these species showed any cytological relationship to Tilia or “6 In the Malvaceae family Malva moschata (n = ?), Lavi thuringiaca (n = 20?), Althaea sulphurea (n = ?), Malvastrum © capense (n = 21), Sidalcea neo-mexicana (n = 13), and of the © Tiliaceae family, Entelea palmata (n = 8) and Sparmannia afi 13, 26) chromosomes are somewhat larger, while in Hibiscus r 180 sinensis (@ = 72?) they are slightly smaller as compared with ‘ : tha. The only striking diversity was found in the size of | — tinea | (n = 8, 10, 13) chromosomes, which are shown : several times larger than any of the above named species. Lee » one another, the only striking exception being Thespé / . 3 F : F a ne . with very large chromosomes. A case similar to this. 1932] DERMEN, THE GENUS TILIA 51 found in my studies of the genus Verbena of the Verbenaceae family (unpublished data). Here the genus was divided into two distinct groups, one with a basic number n = 5 chromosomes and one with n = 7, the latter group having chromosomes about 1/10 the size of the former. LITERATURE CITED Berry, E. W. (1923). Tree ancesto Williams and Wilkins Co., Baltimore. GAISER, 3 O. (1980). }s bceneneend numbers in angiosperms Il. (Bibl. REHDER, x sen. Manual of cultivated trees and shrubs. The Mac millan New Ferree G. (1931). Pflansliche Chromosomen-Zahlen (Tabulae Biol. Per. 109-226) dadabuek LaporaTory, ARNOLD ARBORETUM, RVARD UNIVERSITY. 52 JOURNAL OF THE ARNOLD ARBORETUM [vou. xmT STUDIES ON THE PRECIPITIN REACTION IN PLANTS I. THE SPECIFICITY OF THE NORMAL PRECIPITIN REACTION KennetH §. CHESTER Plate 42 : INTRODUCTION THE AppLicaTION of the theories and methods of animal immu- nology to plant pathology has been a subject which up to the present has received very little attention. In large part the failure of pa- thologists to attack their problems from the standpoint of acquired immunity lies in the theoretical and technical obstacles inherent in a biochemical study of plant disease. Such obstacles are due in part to the differences in structure and development between plants — and animals, and in part to differences in the type of infection cus- _ tomarily produced. Thus in plants one is dealing with organisms _ of indeterminate growth in comparison to the determinate growth 4 of the higher animals, there is no circulatory system in plants closely | comparable to the blood system of vertebrates, and finally the majority of plant infections are more strictly localized than are many of the animal infections. ; Nevertheless, bearing in mind the close ultimate physiological — relationship between the plant and the animal cell, their presumably _ mean vias and their essential functional similarity, it would — 2 deed ; mene bad find the plant cell reacting to infections in 3 — REET! res distinct from that of animal cells. The work re- ays * Present paper was undertaken in an attempt to throw pag hee pets the question of the reactions of the plant cell in — oe as aa ae ae. protoplasm or its constituents, and is a con- _ NR RB Ly: writer’s previous i Hi ook ria keke Sa eran reported studies of the pre- ; he precipitin reaction has long been an essential technique of aN Revere rng essential technique Of immunology. Briefly the theory of the reaction is as fol- lows. A foreign protein is inj : | | Arh r ne nte e Hd ofe mamma 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 53 reaction is specific against the protein originally employed but is weak or negative against other foreign proteins. In addition to such acquired precipitating power, mammalian blood frequently possesses the ability to precipitate certain proteins with which it has never been sensitized. The substances or properties in the blood which induce such precipitation of foreign proteins with which the animal has been previously sensitized may be called acquired precipitins, whereas the substances or properties of the blood in- ducing the precipitation of proteins against which the animal has not been immunized are called normal precipitins. The present paper reports a study of the normal precipitins in a number of families of woody plants. Its purposes are fourfold, namely, to determine whether in extracts of woody plants one may obtain phenomena comparable to the reactions of mammalian blood in the presence of foreign proteins, to discover whether such reactions are in any manner correlated with the systematic rela- tionships of the plants studied, to aid in the interpretation of the earlier published data upon the precipitin reaction in plants, and finally to clarify to some extent our conception of the nature of the precipitin reaction in plants and its bearing on the processes of plant immunity. | HISTORICAL The precipitin technique was first applied to plants by Kostoff in 1928 and 1929 (3,4). The latter, working with a number of species of the solanaceae, observed that in numerous combinations of extracts of Solanaceous species positive reactions were obtained, whereas negative normal precipitin reactions resulted from other extract combinations. The alphabetical order in which Kostoff arranged his tables of reactions does not bring out well the sig- nificance of the reactions which he obtained, nor did he attempt to analyze them in the light of possible correlation with systematic Position. A rearrangement of Kostoff’s data is given in Table 6 for comparison with the results of the present study. From the stand- point of plant immunology this first paper makes a number of im- portant contributions. The author found that after grafting two Species of Solanaceae the precipitin reaction of the extracts of ree -intergrafted plants was markedly increased, that the increase in -precipitin potency was strongest in the tissues nearest the graft union and weakened progressively with increasing distance from the graft union, and that the precipitin potency increased from the _ time of grafting for thirty or forty days, after which ve reached _ 48 equilibrium. In addition to the increased precipitin reaction _ after grafting, Kostoff also observed certain cytological changes, : 54 JOURNAL OF THE ARNOLD ARBORETUM [voL. xur all of which were interpreted in terms of an acquired immunity as the result of sensitization by the foreign protein of the graft biont. The only other study of the precipitin reaction in plants following the technique of direct testing of plant against plant is reported by Chester in 1931 (2). In this paper are considered the results of about a thousand tests in the Oleaceae. No normal precipitins ’ were reported among the species used, but a very strong acquired precipitin reaction was obtained in Syringa vulgaris hybrids grafted upon Ligustrum species as tested against various oleaceous species. The grafted plants were displaying symptoms of disease due to an incompatibility between the lilac scion and the privet stock, and it was found that the morbid processes resulting from such incom- patibility so profoundly modified the components of the lilac ex- tract as to alter markedly the precipitin reaction. The morbidity resulting in this case from graft incompatibility was closely re- — sembled by a morbidity from other causes, and in either case there _ was a marked increase of precipitin potency accompanying the appearance of morbid symptoms in the leaf. z Brief mention should be made of two other bodies of experi- mentation somewhat related to the work at hand. Mez and his col- leagues have published extensive work dealing with an application of the precipitin reaction in animals to plant materials. The re- sults are the basis of the Kénigsberg phylogenetic tree. The work of Mez and his collaborators has been published in numerous papers in Botanisches Archiv, to which the reader is referred for @ complete account of the Kénigsberg studies. The theory of Mez — work is fundamentally different from that of the work of Kostoff and Chester, and hence need be considered no further at this time _ beyoee remarking that the reactions obtained are the reactions of an animal injected with plant proteins and do not represent im- munological reactions in which the plant is the organism acquiring — i A second body of plant immunological research deals with ex- _ pits of these experiments, and a fai Jete 3 bibliography of the work on acqui uired im1 sg : eer ran : 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 55 TECHNIQUE In the summer of 1931, at the suggestion of Dr. Karl Sax, a set of experiments was undertaken to determine the relationships among the genera of the Pomoideae as indicated by the normal precipitin reaction. The results proved so suggestive that other subfamilies of the Rosaceae and other families of the woody plants were eventually tested to the same end. In all about five hundred reactions were performed, involving twelve genera of the Pomoideae, four genera of the Prunoideae, fifteen species of the genus Prunus, two other genera of the Rosaceae, nine genera of the Caprifoli- aceae, eight genera of the woody Saxifragaceae, and one genus each of the Leguminosae and the Platanaceae. Except as indicated be- low all the experiments were performed under uniform conditions, with extracts of the same concentration, tested in the same manner, and with the employment of numerous controls. The technique has been fully described in an earlier paper (2) and only brief men- tion will be made of it here beyond pointing out the modifications which have been developed. Fresh leaves of the plants to be tested were collected, weighed, washed in tap water and distilled water, dried, and ground to a fine paste in an unglazed porcelain mortar. To the paste thus ob- tained was added the required amount of distilled water (four times the weight of the leaves in all the experiments below) and the mixtures were placed in an electric refrigerator at 2° C for twenty- four hours. At the end of this time each mixture was filtered until crystal clear through progressively finer filters, and finally placed in an ice bath. Two to four tenths of a cubic centimeter of the liquid of greater specific gravity was next introduced into a special- ly-made small test-tube by means of a capillary pipette, and the second extract to be tested was pipetted above so as to form a re- fractive zone between the two. Readings of the reaction were taken at intervals of one, five, ten, twenty, thirty, and forty minutes, and in most of the experiments reported below readings were In- dependently made by two observers. The utmost care was taken at every step to avoid contamination, and all instruments and glass- "Ware were cleaned with a sulphurie acid-potassium bichromate mixture for twenty-four hours followed by repeated washings m The positive tests were strong and well marked. In comparison . With the results earlier obtained it may be said that all the readings _ ar minimal. There might have been justification for calling some _ Of the plus two reactions plus three or even plus four, but an at- ___ tempt was made to increase the significance of the results by using 56 JOURNAL OF THE ARNOLD ARBORETUM [vou xm extreme caution in not overestimating the readings. A study of Plate 42 will indicate the scale employed. Figure 1 represents a negative reaction. The delimitation of the two liquids in the tube is clearly indicated, but there is no trace of a white precipitate at the zone of contact. In Figure 2 is seen a reaction indicated in the tables as a ‘“‘trace” (t). Figures 3 and 4 illustrate plus one reactions, in Figure 3 the precipitate representing the ultimate intensity of the reaction between the two extracts employed while in Figure 4 the plus one reaction illustrated being merely an early stage of a reaction which after some minutes would have increased to plus two or greater. Figures 5 and 6 represent plus two reactions, Fig- ure 6 being a later stage of the same reaction as pictured in Figure 4. Figures 5 and 6 show well the penetration of the precipitate in- to the lower extract in little white tortuous rootlets. Later stages of these reactions would show only a uniform cloudiness of the lower extract finally extending upward and involving all the liquid in the tube. : The only essential modification of technique that has been made in the experiments herein reported as compared with the earlier precipitin testing in plants has been that a weaker concentration of extract has been used. Heretofore the ratio of plant tissue to dis- tilled water has been 1:2, while for the purposes of the experiments — . o present paper a dilution of 1:4 was found more satisfactory. _ change was made necessary by the ont percentage of water — eRe ore earlier studied 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 57 dried in a Riker plant press for two weeks at a temperature of about 30°C. At the end of this time the leaves were brittle and quite dry. Half of the leaves of each species were then placed in an oven at 60° C for twenty-four hours. Finally each of the six lots of leaves thus dried was ground, extracted, filtered, and tested against the other five extracts. The results were wholly satisfactory. Strong reac- tions were obtained between the Prunus extracts and those of the other species in every case, while the Platanus and Robinia ex- tracts remained negative to each other. The results of this experi- ment are shown in Table 1. TABLE 1. A STUDY OF THE EFFECT ON THE PRECIPITIN REACTION OF THE PREVIOUS DRYING OF THE PLANT TISSUES EMPLOYED Explanation in the text. g cee ag a oe cer Vege fe ties 3 225/38 22/822 oe ete eS etre o aid e “2 oo! 4 Pat lee elias SSe ea sole a6 nm wm @ Se ee ee hn Ss = 3 3 s m m we = ar ey ‘Sa “8S .-a Pe ee £22\/522/ 2 3 3 Eke Pe aie ieee Platanus acerifolia - - 3 (Fresh leaves) Platanus acerifolia - - ohor 38 3 (Air dried leaves) Platanus acerifolia : es aie Be ete 25°49 Creed ee cic ieege anes Robinia #: ‘ 3 | guia 3 bina fertilis | ~ = er 2 2 3 (Air dio lenves) Robinia fertilis Soe aie a4 (Oven dried leaves) eet tea eee Prunus A Armeniaca Peg tie 3 rs ba 58 JOURNAL OF THE ARNOLD ARBORETUM [Vou. xm It will be seen that there is a slightly higher reaction accompany- ing the use of air dried leaves in one case, and although the distine- tion was too fine to be indicated by the difference of a plus sign, it should be noted that the extracts from air dried leaves always tested slightly more strongly than those from the oven-dried leaves. The reactions as a whole were slightly weaker than those where fresh leaves were used, but the difference is not considered sig- nificant. It was impossible to equalize the concentration of the two types of extract with certainty, since the percentage of water in the fresh leaves was not known. In the case of the fresh leaves the weight ratio of leaf tissue to water of extraction was as 1 :4, while with the dried leaves the corresponding ratio selected was 1:10. The ability to utilize dried leaves in precipitin testing has, as has been indicated, a number of advantages. It is thus possible to continue work with the deciduous woody plants throughout the winter; moreover one may work with specimens shipped in a dried — condition from all parts of the globe; it is possible to employ dried herbarium specimens in testing; manipulation of dried material is much easier than of fresh leaves with especial respect to grinding; the extracts made with dried leaves filter much more readily than — those made with fresh leaves; and finally adjustment of the water L concentration of the extracts may be made with much greater ac- curacy. The fact that experiments may be performed with essen- tially the same results whether one uses dried or fresh leaves does not argue against the hypothesis of the protein nature of the re- _ EE according to the evidence of Osborne and Wakeman as cited _ above. bee: EXPERIMENTAL DATA | e The actual experimental results are shown in the following 4 tables (Tables 2-6). With them is included for comparison a synop sis of the results obtained by Kostoff on the Solanaceae, rearranged — according to the systematic position of the genera. Ae READING AND INTERPRETATION OF EXPERIMENTAL DATA A word of interpretation is a necessary preliminary to a study of . the tables. In contrast to the work of Mez and others, an increasing — positive reaction here indicates an increasing degree of divergence : : from the type. The negative reaction (—) appears to have a double — Significance. A study of Table 2 will bring this out. In the uppet t-hand corner there is a block of negative reactions indicative — _ of the homogeneity of the Pomoideae. Traces of reactions appeat in a few instances, but for the main part the results are uniformly 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 59 a acl = i ' 14 ej}oj}s i \ A ae ~ _— er 1 1 £6 PR. ' | 0 oO t eran os l- 1 ~~ aio} i | t - | B Wy Sigpatras. a n+ o ata oe} 1 | oy oo | oe = oS esosni Bsoy| — ar _ Rr i | | o | oe] a) a Q ; ae eee 8 cb) ek Ee A ea 2 -syeara he 2 Oe ee Ga ee ta Be 4 Tis adie + ~ ee Ja bel Sate bat iy < gh een t er ED Pad oe Ae Ee ao] o i ' ! 5 PER aieerte oper hae REP Tet 2 E aera ge PE Eee sie eee eo endl | E s ae - ee cS ies ate Poh 2 a8 me | oo) | \ 1 zZ = Tarj0dd ws 5 Spree 55? EO itt : : i * cere! 1h ee oe Awe ek ta 5 a" Fs saat ' 1! foe oe ee ee eee gq | Gegieouce, og oe ee ee ee ok ie ie E = ; cena 14 1G! een Wine gk | ir eS a : Ro aa ltiwti tere i Pa ss z ‘ acme 3 i rot ~ ea mm | oe ~ 4 | eel Lae ober tak ig Naa Wii oe se aes 4 ; SI) és | |Seae a de 8] lelsl 13 ee Fier tie ila : : =|2| 9 a) aeess Aasss |g e)2 Es Bg aage2 Bi 2 33 24: = 8 23 g- lZ EI Bis 2 eo i = = B18 ERE :: Ss x. \2 1a a 60 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm negative. The reaction steadily increases as one passes farther from the Pomoideae, beginning with Spiraea which is very closely allied to the Pomoideae, passing to Rosa which is somewhat more distantly related to the Pomoideae, and finally reaching Prunus with the strongest reaction of all, as might well be expected since of all the subfamilies of the Rosaceae the Prunoideae are farthest removed from the Pomoideae. As we pass outside the family, however, the reaction again diminishes. This phenomenon has been observed in a number of cases (cf. Tables 3 and 4). In other words, reasoning from the data thus far available, there appears to be a degree of divergence which results in a maximum reaction. If this degree of divergence be either exceeded or decreased the reaction appears to diminish in intensity. One is immediately struck by the analogy to the “zone phenomenon” in animal immunology, where there is an optimum concentration for reaction, and where if this optimum be passed in either direction the reaction diminishes. There is one significant difference in the plant phenomenon, however, namely that whereas in the zone phenomenon one has to do solely with a quantitative optimum, in the case at hand it appears to be a qual- itative difference which determines the optimum of reaction. Thus in the normal plant precipitin reactions the optimum condition for reaction, according to the hypothesis above, implies a qualitative difference in the reacting substances which must be neither too _ great nor too small. Hence, one would infer that a negative reac- tion might imply a very close relationship or a very great divergence, whereas a positive reaction would involve a definite degree of affin- ity. It would thus be impossible from a single reaction to judge the degree of affinity of two plants; such judgement could only follow from a consideration of the reactions of the plant in question in re- lation to the reactions of a number of related species. As a case in point, the first reactions performed with Prunus involved eight species. From the reactions thus obtained it was very difficult to arrive at a logical interpretation. Seven additional species of Prunus were chosen and tested against the original species and nal against one another. The result was that the seven additio. species offered connecting links and transitions of such value that : the fifteen species at once fell into a logical order, in fact an order which with but one exception agrees with the order of arrangement accepted by present-day taxonomists of the genus. One other fact must also be taken into consideration. As is well known to taxonomists, groups of plants vary among one another in variability. A given character pecgei may ot very uniform in one Bali ee bat Br atl hoo = Totes et Se ee is ies ners esas us Lae 62 JOURNAL OF THE ARNOLD ARBORETUM - [von xm geneous group. Such is the case with the precipitin reaction. Be- cause one finds the Caprifoliaceae to be rather uniformly negative among themselves whereas the genus Prunus exhibits a high degree of variability does not necessarily imply that the Caprifoliaceae as a group are more homogeneous than the genus Prunus. Prunus may be homogeneous in many characters, but in its precipitin re- TABLE 4. NORMAL PRECIPITIN REACTIONS IN THE SAXIFRAGACEAE ‘ Explanation in the text. i z : 3g z 3 : 3 3 i = S465 | ’ ae ga 2 g a3 BS 2e¢ G4 « &£l/als| tees 3 u £ 2 & EB 3 4 ee = ms a oe E a z E S| > 4 ¢§& eed ae & 2|-2\2 eo oe ae . 2/2/32 eeaehe2 ¢ SlEleizle Cee anmaa& és a |e |e Philadelphus grandiflorus | — & = ¢ @|e@)1 |= teem Fendlera Wrightii ee a ie ei ee oe ge Bol B® | eee Jamesia americana - § = - t q¢is iti tae Deutzia scabra plena cope) i eR ee eee ee | 4 2S ae ee | Hydrangea paniculata oR (oe ct Se wae Bee | 2 ~ ee ydrangeoides Ao 0 td = 8 | 8] 8 | ee ; Itea virginiana 18 3 8 4-2 |e]. |e Ribes petraeum #442 ¢ @ e -|-|2| See Prunus Armeniaca Pe Pie be @ $28 = - 1 8 > i Robinia fertilis ee ee eee 8 RR OE 8 eS Lead : Photinia villosa Pee as 1 le | - |e Platanus acerifolia = 1 ey See bee oe ie es i oe it exhibits a remarkable variability. Whatever the character : - | etaud sence of conditions vital to the development of the rerraat 8 a of the principle of incompatibility 35. font ca may well afford an answer to some of the eae xplica se of the parasitology of plants. | _ SUMMARY 1. The pres lescril series of tests of the direct normal precipitin reactions i in a number of families of woody plants for the purpose of determining the specificity of the reaction and its conse quent esa to apace 4 of the method in power? and 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 73 dried leaf tissues in the tests, is described, and its advantages over the use of fresh tissues are pointed out. 3. In general the results of the experiments performed indicate a clear-cut parallel betweeen the systematic relationships as indicated by the precipitin reaction and the main trends of relationship as indicated by the conventional methods of taxonomy. The diver- gences from such a parallel are no greater than are found in the com- parison of the relationships indicated by any two systematic meth- ods. Uniform groups of plants are in general characterized by homogeneity in reaction and absence of mutual reactivity, whereas the reactions steadily increase as one passes farther from the type originally selected, reaching a maximum at a given distance from the type, and then again disappearing as the divergence of the plants hypothetically becomes too great to be expressed in reaction. 4. An analysis is made of the nature of the reaction on the basis of the evidence yielded by the study of normal precipitins. The hypothesis that the reaction may be due to purely physical variables is shown to be untenable, and attention is hence directed to the influence of specific components of the extracts studied. 5. The parallel obtained between the results of the precipitin tests and the accepted systematic relationships indicates the value of an application of the precipitin technique to intensive studies of circumscribed plant groups both in confirming the results of Sheena investigations employing other methods of attack and in shedding light upon the more debatable relationships between certain genera and species. . The specificity manifested by the normal precipitin reaction definite indication of the value of an application of the tecnicos in a study of the nature of immunity to disease in plants. ACKNOWLEDGEMENT The s suggestions and criticisms of a number of the members of the _ Staff of the Arnold Arboretum have assisted in the study herein reported. Acknowledgement is particularly due to Dr. Karl Sax who ; originally proposed that the work be undertaken and who aided lly in the selection of plants for investigation, to Professor 1 H Faull, and Dr. Edgar Anderson for numerous helpful sugges- is, and to Dr. Ivan M. Johnston for valuable assistance in t ae the experimental data from the systematic viewpoint. LITERATURE CITED © ‘ rae. & C. Arnavunt (1930). L’immunita nelle piante. (Monogr. Sierotera KS. 8. Cias. “Catt blight: Ga ae ee | 74 JOURNAL OF THE ARNOLD ARBORETUM [voL. XII employment t of certain understocks in propagation. (Jour. Arnold Arb. an 78-146.) 3. Kostorr, D . (1928). Induced immunity in plants. (Proc. Natl. Acad. Sci. 14: 236-237.) ; 4, (1929). : Acquired immunity in plants. (Genetics, 14: 5. Cemces T. B. & A. J. Waxeman (1 vol Proteins of green leaves. inach leaves. (Jour. Biol. Chem. 42 6. Pa, A. (1927). Manual of cultivated eons aad shrubs. (Macmillan Co., New York.) 7. WELLS, A G.&T.B. Ossornz (1913). Is the specificity of the anaphyl- axis reaction dependent on the chemical co aeroue re . the proteins or on their biological relations? (Jour. Infec. Dis. 1 41-348.) dy peneegs Lasporatory, ARNOLD ARBORETUM, VARD UNIVERSITY, September 10, 1931. Jour. ARNOLD Ars. XIII Plate 42 = ; bs “ wie ’ + L ITS SruprEes ON THE PRECIPITIN REACTION IN PLANT 1932] EHRLICH, CRYPTOCOCCUS FAGI (BAER) DOUGL. 75 THE OCCURRENCE IN THE UNITED STATES OF CRYPTOCOCCUS FAGI (BAER) DOUGL., THE INSECT FACTOR IN A MENACING DISEASE OF BEECH JoHN ERLICH Tue Brrecu sca.x, Cryptococcus fagi, has long been known in Europe as a common pest on Fagus sylvatica and its varieties. It has been found in practically every country of western Europe. Sporadically infestations have at times been followed by an exten- sive killing of Beech, but elsewhere no important damage has re- sulted. Studies on destructive epidemics have yielded some evi- dence that the immediate killing agents were certain fungi, species unable to secure a foothold on healthy uninjured bark, but capable of causing much damage if entrance has been made possible by the scale. The first discovery of the scale in America was reported by Hewitt in 1914 on ornamental European Beeches and our native Fagus grandifolia in the vicinity of Halifax, Nova Scotia. Evidence was presented which indicated that it had been there since about 1890. Subsequent notes on its spread indicate that it is at present general throughout the Maritime Provinces of eastern Canada. But until now there has been no report of its having reached the United States. In the earlier stages of the outbreak in Nova Scotia no serious damage to the Beech was noted, but eventually there set in a destruction of Beech on a wide scale and in alarming proportions. A pathological reconnaissance of the forests of Nova Scotia made in 1929 by Dr. J. H. Faull of the Arnold Arboretum, Harvard Uni- versity, led him to conclude that the death of the Beech was immedi- ately due to fungal action. A study of the problems involved was turned over to the writer in 1929, and is still in progress under the direction of Dr. Faull, supported by the National Research Council of Canada and the Arnold Arboretum. Rance i rue Unirep States j _ Since the scale had not, apparently, been reported in the United States, an examination was begun of Beeches in the metropolitan district about Boston. In November, 1929, a light infestation of What appeared to be the Beech scale was found on the native Beeches in the Arnold Arboretum. In December, a more severe attack was discovered on some Beeches sheltered by Hemlocks in the o Boston city park at Jamaica Pond. Some of this material was sent 76 JOURNAL OF THE ARNOLD ARBORETUM [voL. XI to Dr. E. M. Patch and to Dr. Harold Morrison, both of whom independently stated that the insect was Cryptococcus fagi. The search was continued in the vicinity of Boston and infestations were located in several other places within the city, also in adjacent Brookline, and in Middlesex County. In the spring of 1931, the presence of the Beech scale in metropolitan Boston came to the attention of the United States Entomological Laboratory, Forest Insects Division, at Melrose Highlands, and the survey was con- tinued by their staff, with the result that several additional out- breaks were located, ranging in Massachusetts from Gloucester on the north to the Blue Hills in Milton on the south, and westwards to Belmont and Newton. On November 5, 1931, a heavy infestation was found in Liberty, Maine, by members of the Melrose Highlands Laboratory. The age of the Boston attack cannot be stated with certainty; but its severity in isolated places and the comments of gardeners indicate that the insect has been present for approxi- mately ten years. IMPORTANCE It cannot be prognosticated how widespread or serious the scale is likely to become; but the general fatality of infested Beeches in Nova Scotia and New Brunswick resulting from fungal attack sug- gests a potential danger of the first importance. Slime fluxes are not infrequent on the infested trees in Massachusetts as they are im the Maritime Provinces, but these seem to be of only minor sig- nificance. The species of fungus responsible for the death of the Maritime beeches has not been found in Massachusetts although the — writer has examined a large proportion of all the trees in the region — known by him to be infested. ee Lire History or Cryprococcus FAGI eines history of Cryptococcus fagi was studied in Germany by ; ee. ler (1915, 1922). He found that oviposition occurs from the middle of June to the end of October. The motile form I larvae = after a month, or longer in cooler weather. They are slender, 24-0.33 ees in length, pale yellow in color, and equipped with renee legs and fiv e-membered antennae. These crawlers roam nine in an upward direction, until a suite : Evy: 1s found in a crevice of the bark, on the lower side of ranches, or under the curly-threaded canopy of preceding genem™ ! ) | 1982] EHRLICH, CRYPTOCOCCUS FAGI (BAER) DOUGL. 17 the late autumn or the following February. Between April and July the form II larvae molt and become the nymphal form ITI, whose body length at first is 0.88—0.4 mm., whose legs are lacking, and whose antennae are two-membered. A second molt occurs soon after, followed by development into the egg-laying females, circular in outline and 0.5-0.8 mm. in size. Winged forms and males are lacking. The females oviposit periodically through the late sum- mer and early autumn and die with the coming of the frosts. The life-cycle is thus annual. The development of C. fagi in eastern America coincides with that in Europe. In Nova Scotia and New Brunswick, however, hatching does not begin until about the end of July and the crawlers stop roaming by the end of September. In Boston the period of larval activity is somewhat longer. Local dissemination of the insect is effected by wind-carriage of the crawlers and of bits of wool containing eggs, and within a beech stand by the travelling of the crawlers over the ground from one tree to another. Conrroi The beech scale does not seem to be a serious pest unless accom- panied by certain fungi. Although that danger is apparently not yet present in the eastern United States, the surest way to forestall the possibility is to eradicate the scale. Obviously when once an insect such as this one becomes widely established in the woodlands, as it has in the Maritime Provinces, eradication is a very serious problem. But in Massachusetts, with a range still limited in area and restricted largely to ornamental situations, control seems not only feasible but greatly to be desired. With a view to selecting an easily obtainable and effective insecti- cide for New England use, preliminary tests were made in the late winter of 1931 with different strengths of commercial Sunoco Oil, home-made kerosene-soap emulsion, nicotine sulphate (Black Leaf | 40), and lime-sulphur. The tests were made by soaking an infested area of bark with the aid of a hand spray gun and removing samples of the bark at once and at intervals of several days for microscopic examination. The efficacy of the various materials was dete - by placing the bark on the stage of a binocular and gently raising individual nymphs from the bark with a needle so as not to injure them while thus forcibly withdrawing their stylets from the bark. 1ey were then rolled over so that their ventral sides were upper- most. Those not killed by the insecticide indicated their end by waving the stylets above their bodies. Others, raised slightly, : but not sufficiently to cause complete withdrawal of the stylets from 78 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm the bark, would wave their bodies about, pivoted only on the stylets. A sufficient number of nymphs was examined in this way from every bark sample so as to leave no doubt as to the effect of a particular treatment. The results of these preliminary tests are summarized in Table I. TABLE I. PRELIMINARY TESTS WITH CONTACT INSECTICIDES FOR CONTROL OF THE BEECH SCALE Material | Strength Place Effective | fiteetive | Effective Kerosene-soap emulsion 25% Arnold Arboretum . Kerosene-soap emulsion 25% Jamaica Pond ed Sunoco Oil 1-25 Jamaica Pond # Sunoco Oil 1-5 Middlesex Fells ? Black Leaf40 | 1-50 | Jamaica Pond ¢ Black Leaf 40 | 1-50 Middlesex Fells " Black Leaf 40 1-25 Middlesex Fells ¥ re tecee On te Middlesex Fells | , Lime-sulphur | app.5° B:| Jamaica Pond _| | These results indicate that lime-sulphur and nicotine sulphate © are not satisfactory materials but that Sunoco Oil and kerosene-soap . emulsion are suitable. : Later in the spring of 1931 field tests were made with commercial ; equipment operated by regular park employees in the Arboretum, the Boston Parks, and the Middlesex Fells Reservation, — Sunoco Oil, kerosene-soap emulsion, and nicotine sulphate pertinent results of these tests are brought together in Table T It was concluded from these results that Sunoco Oil, 1-15, is 8 ye most are material. The critical factor in these — a tests i. nal ability of the operator to cover the entire surface e fluffy sufficient thoroughness so that the material wet the” This canopy protecting the insects and penetrated to their bodies: : was possible only with the oils. A few trees were also scrubbed ; kerosene-soap emulsion, was much the 1 ERE But itis a dificu matin and the labor and == 1 : q j : 2 1932] EHRLICH, CRYPTOCOCCUS FAGI (BAER) DOUGL. 79 involved make it applicable only in estates where a small number of valued shade trees are to be kept completely free of the scale. Acknowledgments of assistance in the work done by the writer in Massachusetts are due to Professor J. H. Faull for direction; Mr. Richard Hayden, Superintendent of Parks for the City of Boston, for generous help in scouting and in making field control tests; Mr. . W. Collins, in charge of the Melrose Highlands Laboratory, for scouting information and other assistance; Mr. L. V. Schmitt of the Arnold Arboretum and Mr. W. H. Mollins of the Middlesex Fells Reservation for aid in making control tests and for other courtesies extended TABLE II. FIELD TESTS WITH CONTACT INSECTICIDES FOR CONTROL OF THE BEECH SCALE Number Sprayed Free! Sunoco Oil 1-15 Arnold Arboretum 15 15 100 Kerosene-soap Willow Pond Road, emulsion 25% Boston 18 12 6624 a ee. 1-10 | and Blac 14 pint per| Middlesex Fells | 69 69 100 Leaf 40 40 gallons Met Black 50 gallons Leaf 40 | 1pint | Jamaica Pond, a 234 Ibs Boston 9 0 oc. Me OER _' Number of these trees on which no living scale srededsng dpapusii7siitl tion of accessible parts with a hand-lens a month after spra} SUMMARY The beech scale of Europe, present and spreading in eastern Canada for many years, has recently been found to be abundant on Beeches in the vicinity of Boston, Massachusetts, and has just been discovered in one locality in Maine. Beeches in eastern — Canada are dying i in vast quantities through the attacks of fungi which grow in scale-infested bark; but these fungi have not been —_ on Beeches in Massachusetts. The insect has an annual le; dissemination is effected by eggs and larvae during the Summer and autumn. Experiments with contact earn point to the efficacy of thoroughly spraying ornamental trees wit aye becca Hep emulsion. 80 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm REFERENCES CITED Hewirr, C. Gorpon. Note on the occurrence of the felted beech coceus C fade? (Baerens) Dougl. in Nova Scotia. (Canadian Entom, 1914.) ryp beet 15-16. rs fir Landwirtschaft vom 19. Dezbr,, ter, L. Ve g des Ministe "1913111. _"12109—_betr. Pachareclane. (Jahrb. ebb Forst- ver, 1914: 189-199. i. Buchen- Wollschildlaus (Coccus (Cryptococcus) fagi). In Ntss L. Ruumeter. Forstinsektenkunde. pp. 125-139, fg 81, 2 Pail tary. Berlin, 1922 ParHoLoGIcAL LABORATORY, Nee ARBORETUM, IVERSITY. JOURNAL OF THE ARNOLD ARBORETUM Votume XIII APRIL, 1982 NuMBER 2 CONTRIBUTION TO THE FLORA OF THE NEW HEBRIDES PLANTS COLLECTED BY S. F. KAJEWSKI IN 1928 AND 1929 A. GUILLAUMIN 3 Plate 43 and two text figures NYCTAGINACEAE Calpidia excelsa Heimerl in Oesterr. Bot. Zeitschr. Lx. 284 (1913). Aneityum: Anelgauhat Bay, common in rain-forest up to 600 m., no. 974 (coll. J. P. Wilson), Aug. 1929 (large tree up to 0.60 m. diameter; flowers small, pink; fruit a small round berry).—Al- ready found on Tanna; also Fiji, Society? and Bismarck Islands, New Guinea, Timor and Malaysia.—Vernacular name “Moca.” Cited erroneously from Norfolk Island, Australia and New Zea- land by confusion with C. Brunoniana Heimerl. : AMARANTACEAE Achyranthes aspera Linnaeus, Sp. Pl. 204 (1753). Tanna: Lenakel, common in native gardens, rain-forest, at 150 m., no. 86, March 3, 1928 (plant about 1 m. high, with pretty variegated leaves of pink and brown).—Also New Caledonia, Loy- alty Islands, North Australia, Norfolk Island, Fiji, Tonga, Samoa, Cook, Society, Marquesas, Union, Ellice, Mariana and Bismarck Islands and Malaysia. : POLYGONACEAE Polygonum minus Hudson, Fl. Angl. ed. 1, 148 (1762). _. Eromanga: Dillon Bay, common in rain-forest at sea level, ‘No. 339, May 29, 1928 (a plant growing in semi-cleared patches of _ Tamn-forest; mixed with Nunpar-lell, Ney-wass and Ne-cit-ersif for _ Medicine for sickness on left side of stomach).—Also New Caledonia, Australia (Queensland, New South Wales, Victoria, Tasmania), , Malay: sia (Java) and Philippines (Luzon) —Vernacular name f Neta-pea.”’ : ce ‘Continued from Vol. x11. 30. 82 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Polygonum subsessile R. Brown, Prodr. Fl. Nov. Holl. 419 (1810), Eromanga: Dillon Bay, common on banks of creeks and in swampy places at sea level, no. 367, June 5, 1928 (large plant up to 1.25 m. high; flowers white)—Also New Caledonia and Australia (Queensland, New South Wales, Victoria, Tasmania).—Vernacular name “Ukut-ukut.” PIPERACEAE Piper latifolium Forster f., Fl. Ins. Austr. Prodr. 5 (1786). Aneityum: Anelgauhat Bay, common in rain-forest at se level, no. 718, Feb. 9, 1929 (plant up to 2.5 m. high; fruit brown when ripe; this is the wild Kava and is not used for drinking purposes). Tanna: Lenakel, common in heavy rain-forest soil at sea-level, | no. 3, Feb. 20, 1928 (broad leaved plant about 1.25 m. high; leaves bright green)—Banks Group : Vanua Lava, common in rail- forest at sea level, no. 436, July 9, 1928 (plant up to 1.5 m. high; fruit red when ripe).—Already found on Tanna, Efate and Epi, also Society, Tonga and Fiji Islands and in Timor.—Vernacular | oa “Wild Kava” (under no. 718) and “ Wild N’Kava” (under 10. 436). | Piper methysticum Forster f., Pl. Esc. 76 (1786). Eee | Tanna: Lenakel; common in rain-forest at 150 m., no. 119, March 6, 1928 (plant 1.5 m. high; the roots make the native ne toxicant N’Kava).—Already found on Efate; also New Guinea, Bismarck, Fiji, Society, Tonga, Marquesas and Wallis Islands and | Hawaii.—Vernacular name “N’Kava.” Be Piper miniatum Blume in Verh. Batav. Genostsch. x1. 166 (1826), . Banks Group: Vanua Lava, common in rain-forest at 500) m., no. 463, July 10, 1928 (parasite on rain-forest trees, fruit red | when ripe).—Also New Guinea, Moluccas, Malaysia and Philippines Peperomia leptostachyoides C. De Candolle in Bull. Herb. Boiss | ser. 2, vit. 330 (1908). ie Er omanga: Dillon Bay, common on rocks in shade of scrub or rain-forest at sea level, no. 359, June 4, 1928 (plant about 7 les pre sap of leaf used to make the faces of the natives glossy ; : ate : Fila Island, Vila, common in rain-forest of sea shore, ®° | 188, April 14, 1928 (small plant growing on rocks) _—Already fou" | = | pris Hebrides.—Vernacular name ‘‘Nimtoro-orah”’ ee qe. es, - CHLORANTHACEAE : reseR ea: lanceolata Hooker f. in Jour. Linn. Soc. 1. 127 (1856). pide na: Mt. Tokosh Meru, common in rain-forest at 19 no. 152, March 15, 1928 (tree up to 10 m. high). AneitY 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 83 Anelgauhat Bay, common in rain-forest at 360 m., no. 863, March 5, 1929 (small tree up to 9 m. high; flowers yellow; bark brown and fis- sured). The plant from Tanna differs from the typical form of the Fiji and Samoa Islands in the larger leaves (up to 10.5 em. by 4.5 em.) and the longer petioles (up to 2.5 cm.), while the plant from Aneity- um has narrower lanceolate leaves (up to 12 cm. by 3 cm.). MYRISTICACEAE Myristica aff. M. Hollrungii Warburg. Banks Group: Vanua Lava, common in rain-forest at sea level, no. 422, July 6, 1928 (tree up to 25 m. high; fruit brown, 4.5 em. long, 4 cm. diam.) No Myristica had been found in the New Hebrides, though War- burg (Monog. Myrist. p. 486) indicates that M. subulata Mig. of the Aru Island and New Guinea might exist perhaps also in the New Hebrides. MONIMIACEAE Hedycaria neo-ebudica Guillaumin, sp. nov. Arbor parva, 10 m. alta, ramis gracilibus glabris, foliis oppositis vel 3-nis atro-viridibus papyraceis oblongis vel oblongo-elongatis (usque ad 16 cm. X 5.5 cm.) apice acute acuminatis basi obtuse cuneatis vel sub-rotundatis integerrimis utrinque glabris, nervis 7-10-jugis a venis parum distinctis procul a margine arcuatis tenu- issimis subtus tantum prominulis, petiolo 2-2.5 cm. longo; inflo- Tescentiae 3-4.5 cm. longae, receptaculo fructifero plano supra puberulo, drupis maturite nigris ovoideis (0.8 cm. X 0.6 cm.). Aneityum: Anelgauhat Bay, common in rain-forest at 125 m., no. 811 (typus), Feb. 23, 1929 (small tree up to 9 m. high; leaves dark green; fruit 8 mm. long,6 mm. in diam., black when ripe). Eroman ga: Dillon Bay, common in rain-forest at 400 m., no. 322, May 28, 1928 (tree about 8 m. high; fruit 1.25 cm. long, tur- binate with blunted point, very scarce).—Vernacular name Ney- yar-riverum” (under no. 322). This new species approaches H. denticulata Perk. & Gilg of the Solomon and perhaps the Tonga Islands, but is readily distinguished by the receptacle being pubescent above. The specimen from nanga differs from the type in the oval leaves (not exceeding 10 X 4.5 cm.) with shorter acumen and broader base and in the | shorter, 1-2 cm. long petiole. Hedycarya, sp. nov.? oe he t Peaakal common in rain-forest at 200 m., n0. 90, March 5, 1998 (specimens from tree about 6 m. high). 84 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm LAURACEAE Cryptocarya Wilsonii Guillaumin, sp. nov. Arbor parva, trunco 30 em. diam., ramis glabris, foliis ovatis (usque ad 19 cm. X 10cm.) breviter acuminatis basi subito cuneatis glabris, nervis lateralibus 5-8-jugis, petiolo 1-1.5 cm. longo. Pani culae usque ad 6 cm. longae axillares, ramulis pedicellisque breviter rufo-velutinis, floribus minimis (1 mm. longis) breviter pedicellatis” (sub 1 mm.) vel sessilibus, bracteis minimis lanceolatis in utraque- pagina rufo-velutinis, perianthio campanulato fere usque ad med um 6-lobato, segmentis ovatis carinatis extra dense intus sparsils — rufo-velutinis, staminibus generis, subsessilibus, connectivis pilosis, | antheris lanceolatis apice muticis, staminodiis omnino sessilibus” staminibus similibus sed latioribus apiceque acutis, ovario glabro, stylo subulato staminum apicem subattingente; fructus transverse ellipsoidei (2 cm. X 2.6 cm. X 2 cm.), pericarpio tenui. Aneityum: Anelgauhat Bay, scarce in rain-forest at 175-500 m., no. 951 (coll. J. P. Wilson) Sept. 1929 (low tree to 30 cm. in diam., leaves large; flowers small, yellow; fruit round 2.5 cm. im m., seeds eaten by natives).—Vernacular name “ Inceohp.” The fruit resembles particularly that of C.! obcordicarpa Lecard ex Guillaumin of New Caledonia. Cryptocarya sp. Aneityum: Anelgauhat Bay, scarce in rain-forest at 300 My no. 950 (coll. J. P. Wilson) Sept. 1929 (small tree; leaves small, pointed, with yellow midrib; flowers white; fruit round, 1.8 diam.)—Vernacular name “Ingeyho.’ 1 ‘sp. | bie net : y um : west coast, common in rain-forest up to 450 By 0. 964 (coll. J. P. Wilson) Sept. 1929 (tall tree up to 60 cm. it m., leaves medium; flowers small, white; fruit red, 2 em. Bie a4 em. in diam. )—Vernacular name “Nipicgow.” i aneityensis Guillaumin, sp. nov. cortice ay 10 m. alta, trunco 22 em. diam., ramis com cm. X Seu nrg ellis rufo-puberulis cito glabris, foliis ovatis ¢ mentaceis va te nte viridibus apice basique cuneatis pet nerviis, in utra. er coriaceis costae basi excepta glabris == ‘longo pba he dense reticulato-nervosis, petiolo 1 om puberulae, fi aniculae axillares, circa 5 cm. longae; spat betes oribus coeruleis nent (1-3 mm.) iets pert © a typograp rng > ie ia) have machine genus Cassytha, eer y 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 85 anthio sphaerico 2 mm. diam. extra sparsissime puberulo intus glabro, apice lobis 6 ovato-triangularibus 0.5 mm. longis reflexis glaberrimis, staminibus 3 perianthii ore erectis fere 1 mm. longis glabris, fllamentis basi utrinque glandula parva globosa donatis, antheris ovatis, staminodiis 3 glandulis simillimis, ovario glabro ovato in stylum attenuato, stigmate 2-lobo. Fructus nigri, ellip- soidei (3.5 em. X 2.5 cm.). Aneit yum: Anelgauhat Bay, common in rain-forest at 25 m., no. 704 (typus), Feb. 4, 1929 (large tree up to 20 m. high, leaves ight green); common in rain-forest at 175-500 m., no. 955 (coll. J. P. Wilson), Sept. 1929 (low tree up to 21.5 cm. diam., flowers small, blue; fruit 3 em. long, 2.5 em. in diam., edible, eaten by natives).— Vernacular name “Incitray”’ (under no. 955). This species is very remarkable on account of the spherical rather than campanulate shape of the tube of the perigone. Litsea aneityensis Guillaumin, sp. nov. Arbor magna, 13 m. alta, trunco 25 cm. diam., innovationibus fulvo-tomentosis cito glabris, foliis ovatis (5.5-10 cm. X 3-6 cm.) leviter coriaceis apice acutis vel subacuminatis basi cuneatis penni- herviis, nervis circa 6-jugis, venis subtus dense reticulatis, petiolo 1-1.5 cm. longo. Umbellulae fasciculatae, 1.2 cm. longae, axillares, albae, 4-5-florae, pedunculo 5-8 mm. longo, involucri phyllis 4-5 mm. longis ovatis margine parce ciliatis, pedicello usque ad 5 mm. longo, perigonii tubi lobis 0, staminodiis circa 6, antheris ad laminam lineari-lanceolatam reductis, omnibus filamentis 2-glandulosis dorso sparse ciliatis, ovario glabro. Fructus rubri, ovoidei (2 em. X 1.5 em.) in tubo staminodifero calycem simulante 7 mm. diam. in- sidentes. Aneit yum : Anelgauhat Bay, common in rain-forest at 70 m., no. 748 (typus), Feb. 12, 1929 (large tree up to 12 m. high; stamens and anthers white); southwest, common in rain-forest at 60 to 275 m., no, 960 (coll. J. P. Wilson), Sept. 1929 (low tree up to 40 cm. diam. ; flowers small, white; fruit red, 2 cm. long, 1.8 em. in diam.)— Vernacular name “Incipet” (under no. 960). i This species resembles most L. minor Teschn. of New Guinea. Litsea tannaensis Guillaumin, sp. nov. previa Arbor circa 7 m. alta, innovationibus fulvo-tomentosis nataess glabris, foliis ovatis (4.5-7.5 cm. X 2.5-4 em.) apice obtusis vel obtuse acuminatis basi late cuneatis leviter coriacels ; penninervils, nervis 5-6-jugis, venis immersis fere inconspicuis, petiolo 0.5-1 cm. longo. Umbellulae singulae vel fasciculatae, ad axillas foliorum } vel foliorum delapsorum, 1 cm. longae, albae, 5-florae, pedunculo 86 JOURNAL OF THE ARNOLD ARBORETUM one 5 mm. longo, involucri phyllis 4, 3 mm. longis ovatis margine pate. ciliatis, pedicello brevi vel 0, perianthii tubi lobis 0, staminibus 6-9, _ 8-6 longioribus filamentis antheris 3-plo longioribus glandulis 2 globosis pedicellatis ad apicem munitis, 3-0 interioribus brevioribus filamentis antheris 2-plo longioribus, nonnunquam glandulis pedi cellatis ad apicem destitutis, pistillo 0. Tanna: Lenakel, common in rain-forest at 200 m., no. 9, March 6, 1928 (tree about 7 m. high; flowers white). This species seems nearest to L. maluensis Teschn. of New Guinea from which it is chiefly distinguished by the perianth having no lobes. HERNANDIACEAE Hernandia cordigera Viellard in Ann. Sci. Nat. sér. 4, XVL a (1861). Aneityum :Anelgauhat Bay, common in rain-forest at 50m, no. 703, Feb. 4, 1929 (large tree up to 20 m. high; petals white, used by the natives for canoe-making). Tanna : Lenakel, com mon in rain-forest at 150 m., no. 125, March 7, 1928 (tree 20 m high, 70 cm. in diam.). Eromanga: Dillon Bay, centre 0 island, common in rain-forest at 400 m., no. 341, June 1, 1988 (large tree up to 20 m. high; petals dirty Diacila ockicedl: fruit 4.5-54 | em. long tapering to a blunt point; wood white, soft and used for canoe building) —Also in New Caledonia. son dia peltata Meisner in De Candolle, Prodr. xv. pt. Aneityum: Utgi, common along seashore at 90 m., n0. (coll. J. P. Wilson), Sept. 1929 (large tree to 1 m. diam.; fl yellow, clustered; fruit round, red, clustered). Eroma® Dillon Bay, common in rain. forest and along seashore, Se no. 307, May, 2, 1928 (large straight tree up to 25 m. high, 7 iam. ; white, stamens yellow, stigma pale purple; this a a very soft wood and is used for canoe-building). Bal ul 0up : Vanua Lava, common in rain-forest at sea level, 20 ¥ 10, 1928 (large tree, wood used by natives for canoes). aed pera on Efate; also New Caledonia, Fiji, Tonga, ¥ Bee a lety, Marquesas, Union, Ellice, Marshall, Ma Gui tuZ, Solomon, Bismarck and Admiralty Islands, unea and Malaysia.—Vernacular names “*Nogogu’ ( 1001) and “‘Nehele” (under no. 307). PROTEACEAE : OSE Guillaumin, sp. nov. - Arbor magna, 20 m. alta, ramis validis primum dense fulv a ae, aera re H erett Mepoechtin & insane ad 14 cm. fr 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 87 primum dense fulvo-pilosis deinde glabris apice acute attenuatis basi cuneatis supra lutescentibus coriaceis, costa supra lutea, nervis 4-jugis subtus prominentibus, venis reticulatis subtus prominenti- bus, petiolo usque ad 4.5 cm. longo. Inflorescentiae axillares, foliis subaequilongae primum dense fulvo-pilosa, deinde glabrae, race- mosae, floribus luteis in ramo circa 3 mm. longo 2-nis oblique sessili- bus, bracteis 0, perigonii tubo recto basi oblique dilatato extra fulvo-piloso 1 cm. longo per anthesin uno latere fisso, segmentis concavis lanceolatis dein solutis, antheris ovatis sessilibus, disco unilaterali carnoso, ovario oblique sessili glabro ut stylo cylindrico, stigmate leviter incrassato. Fructus racemosi, in pedicello robusto 1 em. longo singuli, curvatim ovati, leviter compressi (3-4 cm. X 2-3 em. X 1.5-2 em.), apice apiculati, nigri, exocarpio tenuis- simo, endocarpio 2 mm. crasso osseo, semine 1 valde compresso. neit yum: Anelgauhat Bay, common in rain-forest 275 m., no. 901 March 11, 1929 (large tree up to 18 m. high; leaves with light yellow midribs; fruit 4 cm. long, 2.75 em. in diam.; wood beautifully grained similar to the Queensland silky oak); no. 828, Feb. 28, 1929 (tall tree up to 12 m. high; fruit 3 cm. long, 2.5 cm. in diam., yellow when ripe).—T anna: Mt. Tokosh Meru, rain- forest, 400 m., no. 167 (typus) March 15, 1928 (flowers yellow; fruit black).—Vernacular name “Silky Oak’’ (under no. 167). This species resembles, particularly in its leaves, Adenostephanus austro-caledonicus Brongn. & Gris of New Caledonia, which, as also Bentham and Hooker believe, must be a Kermadecia, but its flowers are unknown. Grevillea elaeocarpifolia Guillaumin, sp. nov. a Arbor magna, 12 m. alta, ramis crassis cinereis primum rubiginose tomentosis deinde glabris, foliis lanceolatis (usque ad 14 em. X 3.5 em.) apice acutis mucronulatisque basi acutis, nervis 15—18-jugis tenuibus infra prominulis, venis reticulatis immersis, petiolo circa 1 em. longo primum rubiginose tomentoso mox glabro. Inflores- centiae e ramis veteribus ortae, racemosae, usque ad 15 cm. longae, sparse tomentosae, floribus luteis pedicello 1 em. longo suffultis, Perigonii tubo basi oblique dilatato extra sparse puberulo 1.5 cm. ngo per anthesin uno latere fisso, segmentis ovatis concavis solutis, antheris sessilibus ovatis, disco unilaterali carnoso, ovario stipitato stipite 3 mm. longo, stylo 13-14 mm. longo apice Mecrassato, stigmate terminali conico. ructus indehiscentes, maturitate lutei, curvatim ovoidei, lateraliter compressi (2.5 cm. X 2.3 em. X 1.8 em.), apice apiculati, pericarpio usque ad € 108 lignoso, seminibus 2 plano-compressis orbicularibus margin alatis. 88 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm | Tanna: Lenakel, not common in rain-forest at 200 m., no, 9 | (typus) March 5, 1928 (tree about 12 m. high; flowers yellow; fmuit | yellow when ripe; nuts eaten by natives). Eromanga: ‘llon | Bay, not common in rain-forest at 400 m., no. 350, June 1, 1993 (tree up to 10 m. high; fruit yellow when ripe; nuts eaten by natives), —Vernacular names “Ngye-ngye” (under no. 95) and “ _ ukom” (under no. 350). | The genus is essentially Australian and New Caledonian (with the exception of the Loyalty Islands), but has two representatives in | New Guinea. THYMELACEACEAE Wikstroemia viridiflora Meisner in Denkschr. Bot. Ges. Regeash 1. 286 (1841), Tanna: Lenakel, common in rain-forest at 200 m., no. 107, March 6, 1928 (small shrub 2 m. high; flowers cream-colored). Eromanga: Dillon Bay, common in poor red soil, bracket country at 300 m., no. 302, May 24, 1928 (small shrub 2-3 m. high flowers cream-colored; bark thrown into pools to stupefy fish so they can be caught by hand).—Also New Caledonia, Loyalty, Fiji and Cook Islands and Australia (Queensland, New South Wales, North Australia)—Vernacular name “Tao-wap” (under no. 302). LORANTHACEAE Elytranthe banksiana Guillaumin, sp. n Glaberrima, ramis gracilibus, foliis spathulatis (5-8 cm. X aah cm.) apice rotundatis basin versus in petiolum indistinctum ° tenuatis crassis, nervis immersis. Inflorescentiae racemosae (’ peers 5 mm. baa bractea late ovata 1 mm. longa, calycis ra tubum 1 mm. longum leviter patulum formante, P _ apice rubris, tubo 4 cm. longo, supra medium sensim dilatat arc as costato, lobis 5-6 linearibus reflexis 1.5 cm. airy 20% 5-6 erectis corollae lobis brevioribus, antheris fila sirpel: revioribus linearibus, ovario omnino infero, stylo Bi ants griverenh stamina superante, stigmata capitato. oped pa ' roup: Vanua Lava, common in rain-forest 3°" uly 10, 1928 (parasitic plant growing on a tree Pest’ sunlight is available; very pretty golden petals with red ends | Perse aneityensis Guillaumin, sp. nov. ) mis gracilibus cinereis, foliis obovatis (3-4 em. X 22.55 apice rotundatis basi aa plus minusve subito in petiolum indisti -Coriacei longse, umbe _ venis inconspicuis. Inflorescentiae cymosae, ramis langue | 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 89 triangulari acuta vel ovata extra dense rubiginos-opuberula, calycis parte libera tubum brevissimum campanulatum formante, petalis aurantiacis liberis 5 lineari-lanceolatis 2 cm. longis extra sparse puberulis, staminibus 5 erectis corollae lobis aequilongis, antheris filamentis aequilongis apice muticis, ovario omnino infero, obconico, dense rubiginoso-puberulo, stylo filiformi petalis aequilongo, basi disco annulari cincto, stigmate capitellato. Fructus ovoidei (5 mm. X 4mm.), minimi, rubiginose puberuli. Aneityum: Anelgauhat Bay, common in rain-forest at 300 m., no. 753, Feb. 12, 1929 (parasite growing upon trees in the open partly cleared scrub land; flowers orange-colored). Loranthus ficivorus Guillaumin, sp. nov. Ramis sat robustis dense rugose lenticellatis cinereo-fulvis, foliis late obovatis (5-7 cm. X 3.5-5 cm.) apice rotundatis basi in petio- lum indistinctum cuneatis valde coriaceis, venis 2-jugis immersis. Inflorescentiae 4 cm. longae, umbellatim cymosae, ramis fulvo- puberulis, bractea ovata extra fulvo-puberula, calycis parte libera tubum brevissimum extra fulvo-puberulum formante, petalis basi roseis apice luteis liberis 5 linearibus ad 2.5 em. longis, staminibus corollae lobis leviter brevioribus, antheris filamento 1.5-plo breviori- bus, connectivo apice leviter globose producto, ovario omnino infero turbinato fulvo-puberulo, stylo filiformi petalis aequilongo, basi disco annulari cincto, stigmate capitellato. Tanna: Lenakel, common in rain-forest at 60 m. no. 94, March 5, 1928 (parasitic on Ficus; flowers pink at base with yellow tips). The specimen no. 87, March 3, 1928, from Tanna: Lenakel, common in rain-forest at 150 m. (commonly parasitic on rain- forest trees) with less thick leaves and ellipsoid fruit (1 em. X 0.6 cm.) probably belongs to the same species. SANTALACEAE Santalum austro-caledonicum Vieillard in Ann. Sci. Nat. sér. 4, Xvi. 61 (1861). : Aneityum: Anelgauhat Bay, common in rain-forests at 180 m., no. 814, Feb. 23, 1929 (small tree up to 9 m. high; flowers cream- colored; fruit purple when ripe; this is exported as the san of commerce). Eromanga: Dillon Bay, common in rain-forest at 300 m., no. 287, May 23, 1928 (tree up to 20 m. high; leaves light green above, silvery underneath; flowers cream-colored; fruit 15 _ em. long, 1 em. in diam., black when ripe; the sandalwood of Hee merce).—Already recorded from Aneityum and Eromangs; In” New Caledonia and Loyalty Islands.—Vernacular name Worlu . 90 JOURNAL OF THE ARNOLD ARBORETUM [vou xm EUPHORBIACEAE A Euphorbia obliqua Endlicher, Prodr. Fl. Norf. 85 (1833). . Eromanga: Dillon Bay, common on rocky beach at sea level, no. 257, May 15, 1928 (small plant about 25 cm. high, growing in the crevices of rocks; flowers white; sap used in conjunction with charcoal for tatooing, producing blue marks).—Already found on Aneityum and on Eromanga; also New Caledonia, Loyalty, Norfolk and Tonga Islands.—Vernacular name “‘ Uripatepu.”’ $4 Phyllanthus, sp. nov. ? Aneityum: Anelgauhat Bay, common in rain-forest up (0 150 m., no. 912, March 11, 1929 (small tree up to 6 m. high; flowers orown). | Phyllanthus sp. Aneityum: Anelgauhat Bay, common in rain-forest at e level, no. 908, March 11, 1929 (small tree up to 6 m. high; fruit 0.8 cm. long, light pink when ripe). Tanna: Lenakel, common i rain-forest at 100 m., no. 73, March 1, 1928 (small tree about $m ). Eromanga: Dillon Bay, common in rain-forest at 30 m., no. 274, May 17, 1928 (small tree up to 10 m. high; fruit red; leaves crushed and used by natives for fevers). iy This seems to belong to the same group as P. Gaudichaudu Muell._ Arg. of the Samoa and Tonga Islands, New Guinea, Aru Island ant the Mariana Islands.—Vernacular name “ Narmlee”’ (under no.%74). Glochidion tannaense Guillaumin, sp. nov. 4 oe ramulis leviter compressis glabris, foliis lanceolatis ( " ios X 3 em.) apice basique acutis rigide membranaceis B pallidioribus, nervis circa 7-jugis tenuissimis, petiolo 4-5 mm. 7 lineari-lanceolatis cito caducis petiolo 2-plo bre bi ovario glabro 10-loculari, columna stylari conica superne a0 ist et ovario continua. Tanna: Len akel, ans 3 _ Marcl common in rain-forest at 200 m., BO i ax es (tree of large dimensions with numerous roots of Tee enon). 2 dows ps - staminate flowers are not known, there cal Bl boa . ade Glochidion which should be placed near G, buen Could ese ane G. novo-guineense K. Schum. of New & 1s be the Glochidion sp. found on Tanna by Forster? min in Ace ate puanehei (Brongn. & Gris) Baillon apud 6 ENE SNS. Cl. Marseille, sér. 2, 1x. 224 (1911). 1982] © GUILLAUMIN, FLORA OF THE NEW HEBRIDES 91 Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 933, March 17, 1929 (large tree up to 18 m. high; fruit 1.4 em. long, 1 cm. in diam., yellow when ripe).—Also New Caledonia. Bischofia javanica Blume, Bijdr. 1168 (1826). Eromanga: Dillon Bay, common in rain-forest at sea level, no. 310, May 26, 1928 (large tree up to 25 m. high; bark is boiled in salt water and applied to cuts).—Also New Caledonia, Australia (Queensland), Fiji, Tonga, Cook and Society Islands and Ma- laysia.—Vernacular name- “ No-ghor.” Aleurites moluccana Willdenow, Sp. Pl. rv. 590 (1805). Eromanga: Dillon Bay, common in rain-forest at sea level, no. 242, May 14, 1928 (very fine tree with a straight barrel up to 20 m. high; kernel of fruit threaded on a cocoanut fibre used as a can- dle).—Already found on Efate; also New Caledonia, Loyalty Is- lands, Australia (Queensland), New Zealand, Fiji, Tonga, Samoa, Cook, Society, Marquesas, Gambier and Mariana Islands, New Guinea, Malaysia and Hawaii.—Vernacular name “ Candle-nut.” Croton insularis Baillon in Adansonia, 1. 217 (1861-62). Aneityum: Aname, locally common in lower ranges up to 150 m., no. 988 (coll. J. P. Wilson), Sept. 1929 (small tree to 23 cm. in diam.; leaves broad; flowers small, brown; fruit small 0.6 cm. in diam.). Eromanga: Dillon Bay, common in rain-forest at sea- level, no. 267, May 17, 1928 (small tree up to 15 m. high; leaves green above, silvery brown underneath).—Also New Caledonia, Loyalty Islands, Australia (Queensland, New South Wales).— Vernacular names “‘Nalipes” (under no. 267), “Imrath” (under no. 988). Codiaeum variegatum (L.) Bl. var. moluccanum (Decne.) Mueller Arg. in DeCandolle, Prodr. xv. pt. 1. 1119 (1866). Banks Group: Vanua Lava, common in rain-forest at sea level, no. 409, June 5, 1928 (small tree up to 12 m., high; flowers white) ; no. 423, July 6, 1928 (small tree up to 7 m. high; leaves dark green), _ Codiaeum variegatum (L.) Bl. var. pictum (Lodd.) Mueller Arg. in DeCandolle, Prodr. xv. pt. m1. 1119 (1866). Tanna: Lenakel, common in rain-forest at 100 m. no. 31, Feb. 21, 1928 (shrub up to 6 m. high, with pretty variegated leaves). : This species has been found already on Aneityum, Tanna, Efate, and Mallicolo; also New Caledonia, Loyalty Islands, Australia (Queensland), F iji, Tonga, Samoa, Cook, Marshall, Caroline, ari Santa Cruz, Solomon, Bismarck and Admiralty Islands, New Guinea and Malaysia. 92 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Alphandia furfuracea Baillon in Adansonia, x1. 86 (1873). Aneity um: Anelgauhat Bay, common in rain-forest at 90 m. no. 809, Feb. 23, 1929 (large tree up to 12 m. high; leaves dark green, midrib yellow; flowers creamy yellow; fruit 2.5 cm. long, 2 cm. in diam., light yellow).—Also New Caledonia. Fontainea Pancheri (Baill.) Heckel, Thése Inaug. Montpell. 1870, apud Baillon in Adansonia xr. 80 (1873). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 906, March 11, 1929 (small tree up to 9 m. high; fruit 5 cm. long, 4 cm. in diam., orange-colored when ripe; this tree con- tains a remarkable poison throughout and is used for poisoning fish.) —Also New Caledonia and Loyalty Island. Claoxylon insulanum Mueller Arg. in Linnaea xxxiv. 164 (1865-66). A neityum : south west, common in rain-forest at 60-300 m. no. 958A (coll. J. P. Wilson), Sept. 1929 (small tree, trunk 22.5 cm. in diam.; flowers small, white; fruit a small berry). Efate : Un- dine Bay, common in rain-forest at 500 m., no. 232, April 28, 1928 (tree 10-15 m. high).—Also New Caledonia and Loyalty Islands.— Vernacular name “Namchrai” (under no. 958A). Claoxylon taitense Muell. Arg. var. neo-ebudicum Guillaumin, var. nov. A planta taitensi differt racemis brevioribus (2.56 cm. longis), floribus 3-5-fasciculatis, glabris, petioli glandulis subulatis glandulis minimis adjunctis. Eromanga: Dillon Bay, common in rain-forest at 400 m., no. 347, June 1, 1928 (tree about 15 m. high; flowers white).—Vernacu- lar name “Ney-emptey.”’ | The type of the species occurs in Tahiti and New Caledonia. Acalypha grandis Bentham in Lond. Jour. Bot. m. 232 (1843). _ T anna: Lenakel, common in heavy rain-forest soil at sea level, no. 16, Feb. 21, 1928 (small tree growing in semi-cleared land).— Already found on Aneityum; also New Caledonia, Loyalty, Fiji Wallis, Tonga, Samoa, Bismarck and Admiralty Islands, New | Guinea, Moluccas and Malaysia. a Acalypha neo-caledonica Mueller Arg. in DeCandolle, Prodr. X¥- pt. 1. 812 (1866). T anna: Lenakel, common in rain-forest soil at sea level, 9% 21, Feb. 21, 1928 (small shrub up to 4m. high), Eromang®’ Dillon Bay, common in rain forest at $00 m., no. 383, June 8, 1928 © shrub up to 5 m. ne ee oe on Aneityum; also nt 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES ~ 93 Caledonia and Loyalty Islands.—Vernacular name “ Nau-nompe pura-puri’” (under no. 383). Acalypa sp., an A. Forsteriana Muell. Arg.? Eromanga: Dillon Bay, common in rain-forest at sea level, no. 248, May 15, 1928 (small tree up to 5 m. high).—Vernacular name “Nau-numpey.”’ Identical with no. 1 of Levat from Efate, vernacular name “ Noc- Acalypha Forsteriana, endemic in the New Hebrides, has been al- ready found on Tanna and Efate. _ Cleidion angustifolium Pax & K. Hoffmann in Engler, Pflanzenr. _ W.-147. pt. vir. 293 (Euphorbiac.) (1914). Aneit yum: south west, common in lower hills up to 60 m. no. 959 (coll. J. P. Wilson), Sept. 1929 (small tree up to 25 cm. diam.; flowers very small, white; fruit a small berry) .—Also New Caledonia.—Vernacular name “Nijivit.” The character “ovarium sparsissime adpresse pilosum” brings the species not near C. spathulatum Baill. but near C. Vieillardii : Baill. var. acutifolium Muell. Arg. _ Cleidion Vieillardii Baill. var. acutifolium Mueller Arg. in De _ Candolle, Prodr. xv. pt. 1. 986 (1866). ae Aneityum: Anelgauhat Bay, common in rain-forest at 240 ™,, no, 909, March 11, 1929 (small tree up to 9 m. high, flowers | minute, white).—Also New Caledonia. _ Macaranga Tanarius (L.) Mueller Arg. in DeCandolle, Prodr. xv. tem. 997 (1866). . | _ £anna-: Lenakel, common in heavy rain-forest soil at sea level, _ 20.17, Feb. 21, 1928 (tree up to 15 m. high).—Already found on _ Tanna; also New Caledonia, Australia (Queensland, North Australia, __ New South Wales), Bismarck Islands, New Guinea, Moluccas and 44 oo plant from Tanna corresponds to var. genuina Muell. Arg.; _ Wile the plant of New Caledonia is pubescent, but as indicated by : be Smith and later by Pax and Hoffmann, all intermediate stages are found. | | : Ttis very probable that Ricinus Mappa Forst. also collected on _ “4tna (Forster, no. 213) belongs to this species. ae Sp. », 22a: Lenakel, common in rain-forest soil at 100 m., ene , 1928 (tree 12-15 m. high, about 20-40 cm. diam.). © 0 2 : Dillon Bay, common in rain-forest at 300 m., 10- , 94 JOURNAL OF THE ARNOLD ARBORETUM - [vou xm May 28, 1928 (tree about 10 m. high; dried leaf bandaged over sore to heal it)—Vernacular name ‘‘ Norvo-among” (under no. 319). Macaranga sp. Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 708, Feb. 10, 1929 (small tree up to 10 m. high; leaves dark — green above, silvery underneath; stamens and anthers cream- colored). Banks Group: Vanua Lava, common in rain-forest at sea level, no. 440, July 9, 1928 (small tree up to 7 m. high). These four specimens all staminate possibly belong to the same species. Homalanthus ebracteatus Guillaumin, sp. nov. Arbor 8 m. alta, trunco 15 cm. diam. omnino glabra, ramis tor- tuosis, foliis rhomboideo-ovatis (2-3.5 cm. X 2-4 cm.) apice r0- tundatis vel brevissime acutis basi obtuse truncatis, petiolo 1-3 cm. longo apice 2-glanduloso. Racemi erecti, circa 4 cm. longi, floribus singulis ¢ ebracteatis, 9 bractea naviculari apice cuspidata basi cordata breviter pedicellata involucratis, floribus <= patentibus, pedicello 2 cm. longo supra medium vel ad apicem glandulis 2 hemi- sphaericis notato, sepalo parvo reniformi, antheris circa 20, floribus 2 ad racemi basin 1-2 nutantibus, pedicello 3 mm. longo, tepalis 3 semi-circularibus bene distinctis, stylo brevi, stigmatibus 3-plo longioribus apice 2-lobis. Fructus ovoideo-compressi, carpellorum — dorso carinatis. | Tanna: Lenakel, common in rich rain-forest soil at 200 m. : no. 47, Feb. 24, 1928 (tree about 8 m. high, about 15 cm. diam.). . The pedicels of the staminate flowers without bracts at the base 4 but furnished with two glands below the middle represent an entire- ly new type in the genus which is worthy to constitute a new series _ EBRACTEATI. Homalanthus longipes Pax & K. Hoffmann in Engler, Pflanzent. Iv.-147. pt. v. 51 (Euphorbiac.) (1912). Eromanga: Dillon — Bay, common in rain-forest at sea level, no. 263, May 17, 1928 (tree up to 15 m. high; leaves silvery underneath).—Already found 0D Eromanga.—Vernacular name “‘ Nemtar-bwar.” : The specimen lacks pistillate flowers. Homalanthus nutans (Forst.) Pax var. 4 Aneityum: Anelgauhat Bay, common in rain-forest at 600 m., no. 976 (coll. J. P. Wilson), Aug. 1929 (tree to 0.30 m. diam; flowers very small, yellow; fruit flattened 1 cm. long, 0.7 cm. 0 _ diam.).—Already found on Tanna, Eromanga and Efate; also New _ Caledonia, Loyalty, Fiji, Tonga, Samoa, and Society Islands— _ taing.” ? 4 Vernacular name “ Et 1932] ‘GUILLAUMIN, FLORA OF THE NEW HEBRIDES 95 Excoecaria Agallocha Linnaeus, Syst. Nat. ed. 10, 1288 (1759). Aneityum: Anelgauhat Bay, common along seashore, no. 694, Feb. 4, 1929 (tree up to 15 m. high; sap milky). Tanna: Lenakel, common in heavy sandy soil of seashore, no. 1, Feb. 20, 1928 (spreading tree about 12 m. high, with numerous spreading stems or coppice growth; fruit dark brown when ripe; sap milky).— Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland, North Australia), Norfolk Island, Fiji, Tonga, Caroline, Mariana, Solomon and Bismarck Islands, New Guinea and Malaysia. BALANOPSIDACEAE Trilocularia pedicellata Guillaumin, sp. nov. Arbor parva, ultra 6 m. alta, ramis erectis teretibus glabris, foliis alternis ad ramulorum apicem congestis ovatis (4-7 cm. X 2-4 cm.) valde obtusis basi rotundatis leviter coriaceis, marginibus recurva- _ tis, nervis vix conspicuis concoloribus, petiolo 0.5-1 cm. longo. Fructus 1 em. pedicellati, bracteis 4 minimis sparsis, bracteis in- volucrantibus 8, interioribus sensim majoribus et 5 mm. longis, orbicularibus margine ciliatis, glandi simillimi, cirea 1.5 cm. X 1 em., stylorum 3 usque ad basin 2-fidorum reliquiis coronati, loculis 3. Banks Group : Vanua Lava, common in rain-forest on tops _ of high mountains at 600 m., no. 476, July 12, 1928 (small tree up _ to6 m. high). This species on account of its pedicelled fruit is very distinct from ___T. sparsiflora Schlechter of New Caledonia, the only species of the _ 8enus hitherto known. ae The family thus was represented outside of New Caledonia only bya single species found only once in Queensland. ee CELTIDACEAE a Celtis paniculata Planchon in Ann. Sci. Nat. sér. 3, x. 305 (1848). ‘jag 4 neityum: Anelgauhat Bay, common in rain-forest at sta ih ‘evel, no. 927, March 19, 1929 (large tree up to 15 m. high; fruit a when ripe).—Also New Caledonia, Loyalty Islands, A pate (Queensland, New South Wales, North Australia), Norfolk Island and Society Islands. | fas Vieillardii Schlechter in Engler, Bot. Jahrb. xxxtx. 96 ‘ + A neit yum: Anelgauhat Bay, common in rain-forest at sea ce ee No. 717, Feb. 9, 1929 (small tree up to 15 m. high; eee ack when ripe). Eromanga: Dillon Bay, common mn aad at sea level, no. 398, June 8, 1928 (tree up to 10 m. high; 96 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm used for rafters of native houses).—Also New Caledonia and Loyalty Islands.—Vernacular name “‘ Nendog” (under no. 398). MORACEAE Pseudomorus Brunoniana Bureau in Ann. Sci. Nat. sér. 5, x2. 371 (1869). Aneityum: Anelgauhat Bay, common on seashore, no. 926, . March 17, 1929 (small tree up to 9 m. high).—Also New Caledonia, Australia (Queensland, New South Wales), Norfolk Island, New Guinea and Hawaii. ARTOCARPACEAE Ficus L. Determined by V. 8. SUMMERHAYES Sect. PALAEOMORPHE Ficus Decaisneana Miquel, Fl. Ind. Bat. 1. pt. 11. 312 (1859). Ficus philippinensis var. sessilis Bureau in Ann. Sci. Nat. sér. 5, XIV. 253 (1872). Banks Group: Vanua Lava; sea level, rain-forest, common, no. 493, July 18, 1928 (large tree 20 m. high). Eromanga: Dillon Bay, sea level, rain-forest, common, no. 265, May 17, — 1928 (Fig with straight barrel instead of the usual contorted type; fruit yellow to purple when ripe); alt. 300 m., rain-forest, common, no. 395, June 8, 1928 (small tree 10 m. high).—Vernacular _ pera “*Nevelisi” (under no. 395) and “ Nervelisen”’ (under 20. : Careful comparison of these specimens with some of those — cited by Bureau from New Caledonia and with others from New ye ee ee ee See ae 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 97 in F. Decaisneana and cannot in my opinion be used for separating a yariety from the type. There are specimens at Kew with almost sessile receptacles from almost the whole range of the species although they are commonest in New Caledonia and the New Hebrides. Ficus neo-ebudarum Summerhayes, sp. nov. Arbor parva usque 10 m. alta vel frutex scandens; ramuli cortice brunneo primo sparse scabridulo leviter suleato demum fere laevi obtecti. Folia alterna, breviter petiolata, oblique et inaequaliter ovata, apice breviter acuminata, basi late cuneata vel saepius rotundata, 7-14 cm. longa, 3.5—7 cm. lata, coriacea, omnino glabra, laevia, costa supra vix prominula subtus prominente, nervis lateralibus utrinsecus 6-8 e costa angulo 55-70° exeuntibus prope marginem arcuatim conjunctis supra vix prominulis subtus pro- minentibus, rete venularum distincto saepe siccitate distinctissimo _ sed vix prominulo; petiolus crassiusculus, 6-15 mm. longus, mox cortice in laminas parvas tenuissimas decorticante obtectus; stipulae lanceolatae, acuminatae, glabrae. Receptacula axillaria, solitaria vel gemina, alia flores 8 et 9 cecidiophoros alia flores ? includentia, pedunculata, subsphaeroidea, rubra, 10-12 mm. di To, sparse scabridula, umbicilo prominulo, ostioli bracteis subprominentibus; pedunculus falsus (stipes receptaculi) gracilis, +6 mm. longus, basi bracteis tribus ovatis acutis 1 mm. longis -Instructus, pedunculo vero interdum brevi. Flores & prope ostio- lum, sessiles vel saepissime pedicellati, perianthii segmentis 4 — <4 inearibus vel anguste lanceolatis acutis, ovarium et antheram Superantibus basi sparse pubescentibus; stamen 1, anthera 1 mm. longa; ovarium cecidiophorum ellipsoideum, stylo infra- apleali brevi; pedicellus pubescens. Flores 2 cecidiophori sessiles eS Usque 3 mm. longe pedicellati, perianthii segments 3-4 eis “rum 8 similibus; ovarium ellipsoideum, 1.5 mm. longum, laeve, : Stylo infra-apicali, stigmate parvo clavato. Flores 9 sessiles vel _ ‘Mque 1.5 mm. longe pedicellati, perianthio gamophyllo demum "regulariter fisso quam ovario longiore inferne pubescente; ovarium = _ Neniformi-ellipsoideum, 2 mm. longum, rugulosum, stylo infra- a Seal vel laterali, stigmate clavato. ig Ta nna: Lenakel, alt. 150 m., rain-forest, common, 4 no. 79, arch 3,1 928 (small tree 10 m. high). Aneit y um : west ee oP. Wilson) (type), Sept. 1929 (large vine on forest trees; on red, small; fruit red, 12 mm. diameter).— Vernacular name : ” (ander no. 995). eae ee ee “close relative of F. tinctoria Forst. from which it differs 7 ame, alt. 150 m., lower ranges and sea-shore, common, No- ( 98 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm in the smaller, definitely acuminate leaves and the larger flowers, of which the perianth is only sparsely pubescent at the base. There are specimens of F. tinctoria at Kew from Tahiti, Tonga, Samoa and Fiji and these are remarkably constant in the features mentioned above. Sect. UROSTIGMA Ficus acrorrhyncha Summerhayes, sp. nov. Arbor magna, trunco 2-3 m. diametro. Ramuli teretes, glabmi, demum cortice pallide brunneo leviter ruguloso obtecti, lenticellis + rotundatis pallidis instructi. Folia petiolata, elliptica, oblongo- elliptica vel ovato-elliptica, apice subito anguste et retrorsum acuminata, basi obtusa, latissime cuneata vel subrotundata, — 5-10 em. longa, 3-6 cm. lata, utrinque glabra, supra praesertim — juniora subnitida, costa supra impressa subtus prominente, nervis _ lateralibus utrinsecus 7-10 rectis vel levissime curvatis basalibus angulo 50-60° ceteris angulo 60° e costa exeuntibus supra pr0- minulis subtus prominentibus nervo submarginali leviter curvato conjunctis, nervis secondariis parallelis crebris, rete venularum subtus distinctissimo; petiolus gracilis, 1-2.5 cm. longus, supra canaliculatus, glaber; stipulae lanceolatae, acuminatae, extra adpresse sericeo-pubescentes. Receptacula axillaria, solitaria, sessilia, ellipsoidea vel subglobosa, circiter 15 mm. longa, 10-12 mm. _diametro, siccitate rugosa, glabra, lenticellis rotundatis, ostioli bracteis paulo prominentibus, basi bracteis tribus late reniformibus arcte adpressis instructa. Flores ’, 2 et 9 & cidiophori commixti. Flores longe pedicellati, perianthii segmentis 4 liberis obovatis vel obovato-orbicularibus 1 mm™-_ Es glabris stamen singulum arcte includentibus. Flores ° cecidiophori pedicellati, perianthii segmentis eis florum simili- bus, ovario sessili ellipsoideo 1.5 mm. longo, stylo 0.5 mm. long® stigmate breviter clavato. Flores 9 breviter pedicellati vel saepius i sessiles, perianthii segmentis 3-4 ovatis vel lanceolatis 0.7 mm. longis _ gist ns, Ovario ovoideo 1.3 mm. longo, stylo 2 mm. longo, stigmate eylindraceo-clavato 0.7 mm. longo. ; An e1tyum : Umage, alt. up to 460 m., rain-forest, comma pee 999 (coll. J. P. Wilson), Sept. 1929, (large tree up to $4) diam.; flowers small, yellow ; fruit 15 mm. diam.).—Vernaculat name “‘Nepluth.” - This typical member of sect. Urostigma is allied to F. dicty™ phlebia F. Muell., F. retusa L. and F. benjamina L., from all # which it is easily distinguished by the very abrupt and narrow : acumen to the leaves. F. retusa has shorter and stouter peti? while the fruit possesses a distinct ealyculus. In F. bengamme ; ~— SA 2 zy Res eupulam circiter 1 cm. longam breviter pubescentem recep” ! "stam connatis, receptacula ergo glandes Querci specierum a nen tia; pedunculus usque 4 mm. longus, 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 99 and F. dictyophlebia the leaves are more gradually acute with the veins much more prominent on the upper surface and the petioles _ are shorter, while the fruits are smaller in F. benjamina. Those o F. dictyophlebia are unknown. Ficus austro-caledonica Bureau in Ann. Sci. Nat. sér. 5, xiv. 267 (1872). Aneityum: Anelgauhat. Bay, alt. 240 m., rain-forest, com- mon, no. 862, March 5, 1929 (small tree 6 m. high; fruit 18 mm. long, 22 mm. diam., purple when ripe); s. w. coast, alt. up to 600 m., rain-forest, common, no. 975 (coll. J. P. Wilson), Aug. 1929 (small tree 45 cm. diam.; flowers red; fruit 18 mm. diam., red).— Vernacular name “‘ Natinaia’’ (under no. 975). This species strongly resembles F. granatum Forst. in leaf char- acters. It may, however, be distinguished by the slightly different venation of the leaves, the smaller receptacles with an easily detachable skin, the inside of the receptacles and the stalks of the flowers being almost or quite glabrous, and by the perianth of the female and gall flowers splitting irregularly into usually broad portions. In F. granatum the female perianth consists of five free, linear, acute segments which usually considerably overtop the Ovary or achene. Ficus glandifera Summerhayes, sp. nov. : _ Arbor magna, habitu F. indicae L. similis. Ramuli crassi, juniores sparsiuscule pubescentes, demum glabrescentes, cortice brunneo obtecti, cicatricibus foliorum et stipularum delapsorum valde notati. Folia petiolata, late ovata, apice breviter acuminata, _ 8cuta, basi obtusissima vel saepius rotundata, 7-14 cm. longa, 48 cm. lata, costa supra impressa subtus prominente, nervis : E ° lateralibus utrinsecus 20-30 parallelis e costa angulo 70-80° exeunti- bus: utringue distinctis supra prominulis nervo submarginali _ “irvato conjunctis, nervis secondariis mnumerosis subparallelis 3 crebris, rete venularum subtus distincto, coriacea, supra subni- | tentia, utringue glabra; petiolus pro rata gracilis, supra sheet “analiculatus, 34.5 cm. longus, glaber; stipulae non ans ‘ur caducae. Receptacula axillaria, singula vel gemina, sessiila : “a Subsessilia, oblongo-ellipsoidea, circiter 4.5 cm. longa, 2.5 cm. oe tro, apice in mammillam 5 mm. diametro producta, ostiolo 80 leviter depresso bracteis haud manifestis, bracteis 7 as n. diametro. Flores <7, 9 et 2 cecidiophort commixti. 100 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Flores < longipedicellati, perianthii segmentis 4 liberis ellipticis vel ovatis valde concavis vix 1 mm. longis glabris stamen singulum - arcte includentibus, pedicello bracteis duabus lanceolatis instructo. — Flores 2 cecidiophori pedicellati vel rarius sessiles, perianthii — segmentis eis florum < similibus, ovario sessili, stylo brevi, stig. — mate clavato. Flores @ sessiles, ovario ovoideo 1.5 mm. longo, | stylo infra-apicali 2.5 mm. longo, stigmate minuto. Tanna: Lenakel, alt. 200 m., rain-forest, common, no. 80, | March 8, 1928 (large tree of grand proportions with many roots, — “4m. in diam.”; fruit orange-colored). Aneity um: Anelgauhat | Bay, sea level, rain-forest, common, no. 802, (type), Feb. 21, 1929 (large tree of spreading banyan type, very beautiful; fruit — orange-color when ripe, 4.5 em. long, 2.5 cm. in diam.). The species, with its remarkable acorn-like fruit, is allied toa — group of four species occurring in Queensland, viz: F. cylindrica — Warb., F. Baileyana Domin, F. crassipes F. M. Bailey and F. Watkinsiana F. M. Bailey. The two last-named differ in having — the tip of the fruit drawn out into a point, as well as in the shape — the leaves and other features. F. cylindrica is the most closely — related to F. glandifera, but differs in the possession of large relatively narrower oblong leaves with different venation and longer — petioles, in the receptacles being definitely stalked and the cupuk at the base being telatively smaller. F. Baileyana has leaves vey — similar to those of our species but the receptacles are much smaller, peste, a poorly developed bracteal disc at the base and a slender { o Ficus granatum Forster, Pl. Esc. 37 (1786). ai E romanga: Dillon Bay, alt. 400 m., rain-forest, commo?, — no. 324, May 29, 1928 (large tree 25 m. high; leaves bright grees inside bark used to make cloth as it is glutinous). Tann4: Lenakel, alt. 100 m., rain-forest, common, no. 124, March 7, 1998 rae a m. high; fruit 3-4 at end of each branch, pink when ripe, / by long, » bans diam.). Aneityum: Anelgauhat Bay, alt. i ? m., rain-forest, common, no, 732, Feb. 11, 1929 (large tree 18 ™ on a red when Tipe) ; alt. 15 to 450 m., common, nO. ey, 2; Wilson), Sept. 1929 (low tree 60 cm. diam.; leav ‘a Pu —— flow rs small, red; fruit brown, edible).—V& a ieibe seg: Nating” (under no. 956) and “Nah-tong” (under Pr A striking species of sect. Urostigma which is closely allied ® ~ callosa Willd. and P. austro-caledonica Bureau. The differene between it and the latter are mentioned under that species- hee LEST) F. iA ? a 2> 9 wie z : 2 eallosa ia, F. —, is distinguished by its gradually po as ©-CaLéeado 4 d . Lie Mesnites : 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 101 leaves with the main lateral nerves joining near the margin more simply, the receptacles being red or purple when ripe, and being thickly hairy inside, this also applying to the pedicels of the flowers. In general floral and vegetative characters the species are otherwise very similar. Ficus obliqua Forster, Prodr. Fl. Ins. Austral. 77 (1786). Eromanga: Dillon Bay, sea level, rain-forest, no. 253, May 15, 1928 (common throughout group growing into large tree of many branches and roots; fruit orange-color). Tanna: Lenakel, alt. 200 m., rain-forest soil, common, no. 67, Feb. 24, 1928 (large tree 15-20 m. high, 60 cm. to 1 m. in diameter; fruit brown when ripe). Aneit yum: Anelgauhat Bay, sea shore, rain-forest, common, no. 781, Feb. 19, 1929 (large tree with spreading head; fruit orange-color when ripe).—Vernacular name “Nar-evirepp” (under no. 253). Ficus prolixa Forster, Prodr. Fl. Ins. Austral. 77 (1786). Ficus prolixoides Warburg in Fedde, Rep. Nov. Spec. 1. 79 (1905).— Synon. nov. Eromanga: Dillon Bay, alt. 300 m., rain-forest, common, no. 273, May 17, 1928 (giant Banyan of many trunks and roots; fruit black when ripe). Tanna: Lenakel, alt. 100 m., rain- forest, common, no. 29, Feb. 21, 1928 (large tree of many trunks sending down many roots, up to 25 m. high; fine shade tree); alt. 200 m., rain-forest, common, no. 113, March 6, 1928 (largest Banyan on Tanna). Aneityum: Anelgauhat Bay, sea level, rain-forest, common, no. 900, March 15, 1929 (large spreading Banyan, 18 m. high; fruit 10 mm. long, 8 mm. in diam., black When ripe).—Vernacular name “Nepang” (under no. 273) and Banyan” (under nos. 29 and 900). j 4 I cannot distinguish F. prolizoides Warb. from this bg eee we the material at my disposal shows that the differences given by Warburg are not constant. ae bas Sect. SYCIDIUM £ Ficus aspera Forster, Pl. Esc. 36 (1786). Tanna: Lenakel, alt. 200 m., rain-forest, common, no. 49, Feb. 24, 1928 (small tree 8 m. high; fruit red when ripe); alt. 200 '™, rain-forest, common, no. 103, March 5, 1928 (tree S12 m. Mg; fruit yellow, eaten by natives). -‘Ricus ciliata Warburg in Bot. Jahrb. xxv. 615 (1898). : Anecityam; Anelgauhat Bay, sea level to 100 m. alt., +d _ “rest, common, no, 731, Feb. 11, 1929 (large tree 18 m. high; 102 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm fruit red when ripe, 14 mm. diam., flattened at the ends); sea level, sea shore, common, no. 788, Feb. 20, 1929 (small tree 6 m. high; fruit red when ripe, 15 mm. diam., flattened at top). This species, which has been previously recorded only from Samoa, differs from F. Kajewskii Summerhayes in the larger receptacles and ciliate perianth lobes. Ficus copiosa Steudel, Nomencl. ed. 2, 635 (1840). Banks Group: Vanua Lava, sea level, rain-forest, common, no, 469, July 12, 1928 (small Fig 10 m. high, with fruit growing on trunk). Tanna: Lenakel, alt. 100 m., rich rain-forest soil, common, no. 24, Feb. 21, 1928 (tree 10-12 m. high; fruit red when ripe and growing on side of wood). This species is also represented at Kew by New Hebridean specimens collected by Miss Cheesman and by Dr. A. Morrison. It is easily recognized by the fact that cork is formed at an early stage at both ends of the petioles but not in the middle. The specimens, in addition to agreeing with the description and the plate in King’s monograph (the latter a reproduction of Roxburgh’ — original drawing), match the specimens collected in Amboina MY | C. B. Robinson and distributed by Merrill as Ficus wassa Roxb. In the writer's opinion these latter specimens are better referred to F. copiosa Steud. (F. polycarpa Roxb., non Jacq.). It is interest ing to note from Roxburgh’s own descriptions that the two species _ are probably allied, since apart from other features they both i said to possess a green gland in the axil of the lowest nerve on each side of the leaf. Ih his description of F. polycarpa Roxburgh likens - his species in this respect to F. laciniata. There is, however, ek F.. laciniata either in Flora Indica or elsewhere, and a glance tt the description and plate (in Wight’s Icones) of F. wassa, with its leaves sometimes laciniate, suggests that Roxburgh originally _ are this species F. laciniata and afterwards changed the 29™° to F’. wassa forgetting, however, to make the corresponding alter ; tion in his manuscript under F. polycarpa. : F. copiosa and F. wassa are put into different groups by Ro * urgh on account of the receptacles being borne differently, bul at F. copiosa may have its receptacles either axillary or on the as : ranches or trunk Roxburgh’s segregation can scarcely be ma¥ tained on the original grounds alone. For the moment, however, the descriptions and plates: ct ) too much in other respects to justify uniting the two spe errill’s identification of Robinson’s specimens as F. wassa 3 5 pends to a great extent on the identity of the native name 80 to the plants described by Rumphius and Roxburgh with ae 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 103 given to Robinson’s plant. Until we know, however, that the names “wassa”’ and “giho”’ are applied by the natives to one species of Fig only, we cannot be sure of the identity of the species concerned. I am, therefore, maintaining the name F. copiosa Steud. for the present species, although if the two species were united F. wassa would have to be adopted in view of its priority. Ficus Kajewskii Summerhayes, sp. nov. Arbor parva, 10-15 m. alta. Ramuli teretes, graciles, leviter flexuosi, primo scabridulo-pilosi, demum glabri cortice brunneo vel cinereo-brunneo longitudinaliter ruguloso obtecti. Folia pro genere parva, petiolata, oblique ovata, oblongo-ovata, oblongo- lanceolata vel lanceolata, apice breviter obtuse vel rarius subacute acuminata, basi inaequaliter cuneata usque subcordata, usque 11 em. longa et 5 cm. lata sed saepius satis minora, chartacea vel rarius subcoriacea, utrinque asperula, glabra, costa supra Prominula subtus prominente, nervis lateralibus utrinsecus curvatis subtus prominentibus juxta marginem arcuatim conjunc- tis, nervis tertiariis prominulis, rete venularum quadrato mani- festo; petiolus gracilis vel subgracilis, 4-8 mm. longus, primo seabridule pilosus demum glaber, supra canaliculatus; stipulae lanceolatae, acutae, 2-3 mm. longae, extra adpresse pilosae vel fere glabrae. Receptacula axillaria, solitaria vel gemina, ped- unculata, globosa vel subglobosa, 5-6 mm. longa, 6-8 mm. dia- metro, breviter scabrido-pilosula vel fere glabra, ostiolo parvo is vix prominentibus, intus inter flores setis brevibus hyalinis Rumerosis instructa; pedunculus gracilis, 2-6 mm. longus, sca- brido-pilosulus, supra medium vel apice bracteis tribus ovatis obtusis instructus. Flores 2 sessiles vel rarius usque 1 mm. pedicellati, perianthii segmentis 4-5 lineari-oblongis vel oblongis obtusis vel subacutis superne + coalitis ovarium includentibus slabris; Ovarium sessile, 1-1.5 mm. longum, complanatum, obovoi- deum vel ellipsoideum, stylo laterali 0.6-1.2 mm. longo, stigmate t minuto leviter dilatato. Flores o' et 9 cecidiophori non visi. Efate: Undine Bay, alt. 200 m., rain-forest, common, No- 216, April 27, 1928 (tree about 15 m. high; fruit red when Tipe) ; Mt. McDonald, alt. 500 m., rain-forest, common, No. 233, April 28, 1998 (tree 15 m. high) Eromanga: Dillon Bay, Centre of island, alt. 400 m., rain-forest, common, no. 343, June Ly: 1998 (small fig, 10 m. high, with very small straight trunk). Anel- 'Yyum: Anelgauha in-forest, common, no. SS ; t Bay, alt. 34 m., rain-forest, ©" | _ 37 (type), Feb. 12, 1929 (small tree 10 m. frig leaves dar 104 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm green, midrib light yellow; fruit purple black when ripe).—Ver- nacular name “Neme-sue” (under no. 343). The nearest relative of this species seems to be F. Barclayana Summerhayes, comb. nov. (Covellia Barclayana Miq., Freus Bar- clayi Seem.), a native of Fiji. Although that species is placed by Miquel in sect. Covellia and the female flowers are drawn as pos- sessing no perianth, F. Barclayana seems to me to belong mor properly to sect. Sycidium. Seemann’s dissections from specimens which seem identical with the type show the female flowers as possessors of typical Sycidium perianths. F. Kajewskii differs from the Fiji plant in its less hairy receptacles, which are however furnished with colorless setae inside, in the perianth being quite — glabrous and in minor characters of the leaves. Both species have the rough leaves and the small rough axillary receptacles char- acteristic of sect. Sycidium. Ficus trichoneura Summerhayes, sp. nov. Arbor parva vel mediocris usque 15 m. alta; ramuli hornotini a adpresse hirsuti, annotini glabrescentes, cortice leviter longitudina- liter sulcato castaneo-brunneo obtecti. Folia alterna, longiuscule petiolata, rhombeo- vel oblongo-elliptica vel ovata, superne late acutata apice ipso emarginato-obtusa, basi rotundata vel sub- rotundata, 5-14 cm. longa, 2-6 cm. lata, chartacea, supra geht: laevia, siccitate griseo-viridia, subtus costa et nervis praesertim infimis adpresse hirsutis demum glabrescentibi ceterum glabra, pallide viridia, saepius albido-punctulata, costa et nervis supra prominulis subtus prominentibus, nervis primariis utrinsecus 8-10 infimis angulo acuto ceteris angulo circiter 60° e costa exeuntibus prope marginem arcuato-conjunctis, rete vent mabe) crebro — petiolus gracilis, 1-4 cm. longus, bast ie canaliculatus, primo dense adpresse hirsutus, demu? Laneheqeie s cortice ei caulis simili obtectus; stipulae linea’ , glabrae, castaneo-brunneae, cito decidua 1982 GUILLAUMIN, FLORA OF THE NEW HEBRIDES 105 9 cecidiophori sessiles, brunnei, perianthio irregulariter fisso segmentis saepius 2-3 lanceolatis ovario aequilongis glabris; ovarium compresse ellipsoideum vel ovoideum, circiter 1 mm. longum, stylo laterali brevi, stigmate breviter clavato. Flores Q sessiles, brunnei, perianthii segmentis tribus linearibus 1.25-.5 mm. longis; ovarium compresse reniforme, 1.5 mm. longum, rugulosum, stylo laterali, stigmate clavato ovarium paulo superante, bractea basali spathulato-lineari. Banks Group: Vanua Lava, sea level, rain-forest, common, no. 417, July 6, 1928 (medium sized tree up to 15 m. high), and no. 426 (type) July 6, 1928 (small tree up to 6 m. high). Ero- manga: Dillon Bay, sea level, rain-forest, common, no. 264, May 17, 1928 (smallertype of Fig averaging 10 m.high). Tanna: Lenakel, alt. 200 m., rain-forest, common, no. 89, March 5, 1928 (large tree 20 m. high); alt. 100 m., rain-forest soil, common, no. 23, Feb. 21, 1928 (tree 12-15 m. high; leaves dark green). Aneityum: Anelgauhat Bay, sea level, rain-forest, common, no. 711, Feb. 9, 1929 (Fig tree growing 20 m. high; fruit green when ripe, 6 mm. in diam.).—Vernacular name “Bongnute” (under no. 264). This species is apparently one of the commonest Figs in the New Hebrides, as in addition to the above gatherings there are at Kew several more from the southern islands of the group. Ac- cording to Miss L. E. Cheesman the species is also found in Male- kula, which is quite probable, although I have not seen a specimen from that island. The species has been collected up to 300 m. altitude. There is also a specimen at Kew collected by Kajewski _ Inthe Santa Cruz Islands. F. trichoneura is most closely allied to F. chrysolaena K. Schum., _ & native of New Guinea and the Solomon Islands, which differs _ Inhaving the veins more hairy below while the lamina is also partly : hairy, the receptacles larger and shortly pubescent, and the bracts _ -Mmmediately at the base of the receptacles so that there 1s ag _ ‘Stipe. ts is also interesting to note that our TT He axergid : _ Fesembles F. adenos a Mig. which, however, by gales oy hermaphrodite etna dge placed in sect. Palaeomorphe. Since | 3 _, both F. trichoneura and F. chrysolaena vary considerably from the _ Rormal type of sect. Sycidium it is possible that they are really types of sect. Palaeomorphe in which the hermaphrodite flowers | lave been lost. In this connection it is worthy of notice that ng 4 F. Md a wholly male flowers are present as well as hermap x q : _Todite and gall flowers, the hermaphrodite flowers contaiming 4 ___ Properly developed gall ovary and an abortive stamen. : 106 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Sect. COVELLIA Ficus verrucosa Vahl, Enum. 11. 192 (1806). Ficus septica Forster, Prodr. Fl. Ins. Austral. 76 (1786).—Non Burm.f, — Ficus casearia F. v. Mueller apud Bentham FI. Austral. v1. 177 (1873)— Summerhayes in Jour. Arnold Arbor. x. 148 (1929). Eromanga: Dillon Bay, alt. 300 m., rain-forest, common, no. 271, May 17, 1928 (Fig with single stem up to 15 m. high; leaves dark glossy green; fruit yellow when ripe). Aneityum: — Anelgauhat Bay, sea level, rain-forest, common, no. 793, Feb. 20, 1929 (large tree 12 m. high; fruit 12 mm. long, flattened at both sides, cream-color when ripe).—Vernacular name ‘“‘ Ubariherhon” (under no. 271). F This species is widely spread from the New Hebrides to New Guinea and Australia. There are specimens at Kew from the © Admiralty and Solomon Islands. It is closely allied to F. leucanta- toma Poir. ] Sect. EUSYCE ! Ficus Moseleyana King in Ann. Bot. Gard. Calcutta, 1. 14 t. 181 (1888). : Ba nks Group: Vanua Lava, sea level, sea shore, common, no. 433, July 7, 1928 (medium tree up to 15 m. high). Previously known only from eastern Malaya. Artocarpus incisa Linnaeus f., Suppl. 411 (1781). } es roman ga: Dillon Bay, common in rain-forest at sea level, | no. 394, June 9, 1928 (large tree up to 20 m. high; the most usefil tree of the islands as it supplies large quantities of fruit and the : bag is used for canoes).—Also New Caledonia, Samoa, Society: Ellice, Gilbert, Marshall, Caroline, Mariana, Santa Cruz, Solomo® — Bismarek and Admiralty Islands, New Guinea and Hawaii— Vernacular name ““Ne-marl,” +e = URTICACEAE aa sportea crenulata (Roxb.) Gaudichaud in Arch. Mus. Nat Paris, 1x. 188, t, 2c, fig. 5 (1856). a gee. be ipl gs Vanua Lava, common in rain-forest at se : Guinea and Malaysia, : £ sea epi that the plant found on Efate by Levat should be® — = ites Species and not to L. photiniphylla. ee ee macroph yiitire rongniart in Duperrey, Voy: Co i quille 207, t. 45 (1829). . 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 107 Aneit yum: Anelgauhat Bay, common in rain-forest up to 450 m., no. 968 (coll. J. P. Wilson), Sept. 1929 (small shrub about 1.30 m. high; flowers small, white; fruit at base of flower, about 0.4 em. in diam.; stem pithy).—Vernacular name “TIlbuvin.” Procris pedunculata Weddell in De Candolle, Prodr. xv. pt. 1. 191 (1869). Aneit yum: Anelgauhat Bay, common in rain-forest at 300 m., no. 852, March 2, 1929 (plant up to 1 m. high; fruit with red pulp and white seeds).—Already found on Efate; also New Caledonia, Fiji, Samoa, Society, Marquesas, Marshall, Mariana, Solomon and Bismarck Islands and New Guinea. Boehmeria anisoneura Guillaumin, sp. nov. Arbor parva, trunco 15 cm. diam., ramis rufo-velutinis, foliis oppositis inaequalibus, uno valde asymmetrico ovato (usque ad 25 em. X 10 cm.), petiolo 6-8 em. longo, altero fere symmetrice ovato (usque ad 12 cm. X 8 cm.), petiolo circa 0.5 em. longo, membrana- ceis margine a basi ad apicem dense serratis apice attenuatis vel longe acuminatis basi oblique rotundatis leviterque cordatis, nervis principalibus 3, uno in parte angustiore fere apicem attingenti, altero in parte latiore ad tertiam supreman partem vergente, secun- ris @ venorum exteriorum exteriore parte regulariter parallelis numerosis, a veno centrali ad partem latiorem arcuatis paucis, nervis valde reticulatis, pagina superiore hispido-scabra, inferiore velutina, petiolo velutino, stipulis intrapetiolaribus lanceolatis 28 cm. longis tarde deciduis, extra praecipue secundum lineam Jongi- tudinaliter centralem argenteo-hispidis. Flores albi, in axillis glo- merati, glomerulis circa 1 cm. diam. multifloris, bracteis glabris, ¢ Perigonio ore breviter 2-3-dentato extra hirsuto ellipsoideo, ovario styloque glabris. Aneit y um: west coast, common in rain-forest up to 600 m., no. 1005 (coll. J. P. Wilson), Sept. 1929 (small tree to 15 em. i diam. ; flowers very small, white; fruit very small, at base of petiole). —Vernacular name ‘‘Nowau.” : A very remarkable species among those with axillary glomerules and opposite leaves. It seems nearest to B. monticola Bl. of Sumatra. | a platyphylla (Buch-Ham.) D. Don, Prodr. Fl. Nep. 60 Aneit yum : west coast, common in rain-forest up to 450 eee 20. 961 (coll. J. P. Wilson), Sept. 1929 (small tree to 15 cm. ™ diam. ; wers small; fruit very small).—Already found on Espiritu Santo; __4lso New Caledonia, Australia (Queensland), Fiji, Samoa ee ty Islands, New Guinea, Malaysia —Vernacular name “Nahwat- 108 JOURNAL OF THE ARNOLD ARBORETUM [vou.xm Cypholophus macrocephalus Weddell in Ann. Sci. Nat. sér. 4, — 1. 198 (1854). Aneityum: Anelgauhat Bay, common in rain-forest at 300 m., no. 766, Feb. 14, 1929 (shrub up to 3.60 m. high; fruit light green). —Also Fiji, Samoa and Society Islands, Moluccas and Malaysia. Pipturus albidus (Hook. & Arn.) A. Gray apud Mann in Proc. — Am. Acad. Sci. viz. 201 (1867). Tanna: Mt. Tokosh Meru, common in rain-forest at 1000 m., no. 148, March 15, 1928 (small plant 1 m. high on crest of the moun- tain).—Also New Caledonia, Society Islands, New Guinea and — Hawaii. Pipturus argenteus Weddell in DeCandolle, Prodr. xv. pt. 1 935 (1869). | cleaning hair)—Also New Caledonia, Loyalty Islands, Malays and Hawaii. : Leucosyke 235% (1869). | ai sy ity um : west coast, common in lower forest ranges "P to 180 m., no. 963 (coll. J. P.. up to 0.3 m. diam., flower Bay, common in rain-forest capitellata Weddell in De Candolle, Prodr. xvi. pt! New Guinea, Moluccas fig” (under no. 963), 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 109 CASUARINACEAE Casuarina equisetifolia Linnaeus, Amoen. Acad. rv. 143 (1759).— Forster, Char. Gen. 103, fig. 52 (1776). Aneityum: Anelgauhat Bay, common on seashore, no. 697, Feb. 4, 1929 (tall tree up to 20 m. high, common throughout the New Hebrides) Banks Group: Vanua Lava, common on seashore, no. 451, July 9, 1928 (tree up to 20 m. high growing on the average much larger than the Queensland coastal ‘‘She-oak’’).— Already found on Aniwa; also New Caledonia, Australia (Queens- land, North Australia), Fiji, Tonga, Cook, Society, Marquesas, Mariana, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia. ORCHIDACEAE Determined by Oakes AMES Corymborchis veratrifolia (Reinw.) Blume, Fl. Jav. nov. ser. 1. 105, t. 43, fig. 1 (1858). Tanna: no. 146 (see p. 127). Malaxis neo-ebudica Ames in Jour. Arnold Arb. x1. 128 (1932). Tanna: no. 137 (see p. 128). Malaxis xanthochila (Schltr.) Ames & Schweinfurth in Ames, Orch. v1. 73 (1920). Aneit y um: no. 837 (see p. 128). Liparis condylobulbon Reichenbach f. in Hamb. Gartenz. Xvi. 34 (1862), Eromanga and Banks Group: nos. 321 and 455 (see p. 129). Dendrobium (§ Aporum) sp. Efate:no. 190 (see p. 131). Eria Kajewskii Ames in Jour. Arnold Arb. xii. 135 (1932). Aneit yum: no. 820 (see p. 135). Ceratostylis subulata Blume, Bijdr. 306 (1825). Efate:no. 234 (see p. 138). | Earina Brousmichei Kriinzlin in Lecomte Not. Syst. 1v. 136 (1928). Aneit yum: no. 844 (see p. 138). Appendicula reflexa Blume, Bijdr. 301 (1825). is) _Aneityum and Ero manga: nos. 915 and 326 (see p- . _, Spathoglottis Petri Reichenbach f. in Gard. Chron. n. ser. vii Me aan. ae Eromanga : no. 306 (see p. 140). 110 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Saccolabium Kajewskii Ames in Jour. Arnold Arb. x1. 141 (1932), — Efateand Banks Grou p: nos. 205 and 448 (see p. 141). ZINGIBERACEAE a Zingiber zerumbet Roscoe apud Smith, Exot. Bot. 1. 105, t. 112 (1805); in Trans. Linn. Soc. vimt. 348 (1807). i. Aneityum: Anelgauhat Bay, common in rain-forest at 18) m., no. 850, March 2, 1929 (plant up to 1.5 m. high; flowers white)— _ Also New Caledonia, Australia (Queensland), Fiji, Samoa, Society, Marquesas, Mariana and Bismarck Islands, New Guinea, Malaysia — and Hawaii. Alpinia Blumei K. Schumann in Bot. Jahrb. xxv1r. 282 (1900)? — Banks Group: Vanua Lava, common in rain-forest at se level, no. 416, July 5, 1928 (tall plant about 3 m. high, along water : courses; flowers bright pink). a The characters agree well with the description of this Javanest species except that the inflorescence attains 14 cm., but the flowers — are lacking and it is impossible to know the shape of the bractlet. In any case the plant belongs in the genus Guillainia. MARANTACEAE Donax arundastrum Loureiro, Fl. Cochin. 11. (1790). a ; an ks Group: Vanua Lava, common in swampy Mir” Hote at sea level, no. 420, July 6, 1928 (plant up to 2 m. high with pei radiating from one node; flowers white to cream-colored; ~? em. in diam.).—Also in the Philippines and Asia. : | i L . _ AMARYLLIDACEAE ae : hit pedunculatum R. Brown, Prodr. Fl. Nov. Holl. 7 Aneityum: Anelgauha Pek | : at Bay, common along sea shores 9 805, bias 21, tied (large plant up to 1.5 m. high, growing — 7 : merous, long; flowers white).—Also *®" Waledonia, Australia (Queensland, New South Wales, South Si tralia), : co se f ci gnd Malaysad and Cook Islands, New Guinea, Mon a, i wy | - TACCACEAR Hie a Z are ay Gen. Char. 70, t. 35 (1776). | ne no. 345, hake 5 1998 ti Bay, common in rain-forest at err the islands: 4 (plant up to 80 cm. high growing wild OF cif been a aR T SS Yi ic: 18 manufactured from the bulbs and | | donia, Loya " revenue for the mission stations).—Also New » Loyalty Islands, Australia (Queensland, South A 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES lll Fiji, Tonga, Samoa, Society, Marquesas, Ellice, Marshall, Caroline, Mariana, Solomon and Bismarck Islands, New Guinea, Malaysia and Hawaii.—Vernacular name “ Yovoli.” DIOSCOREACEAE Dioscorea bulbifera Linnaeus, Sp. Pl. 1033 (1753). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 896, March 11, 1929 (wild yam growing over the small trees; not edible).—Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland), Fiji, Tonga, Samoa, Soci- ety, Marquesas, Caroline and Mariana Islands, New Guinea, Moluccas, Malaysia and Hawaii. Dioscorea nummularia Lamarck, Encyc. Méth. ut. 231 (1789). ? Eromanga: Dillon Bay, common in native gardens, rain- forest at sea level, no. 250, May 15, 1928 (climbing on sticks placed by natives who take great care of its cultivation; this red yam is the chief food of natives).—Also Australia (Prince of Wales Island), Fiji, Society, Caroline and Bismarck Islands, New Guinea, Moluc- cas, Philippines.— Vernacular name “Loop.” It is strange that the collector indicates D. nummularia as the edible species and D. bulbifera as the wild species, and one may ask whether the labels have not been interchanged. LILIACEAE Smilax vitiensis A. De Candolle in De Candolle, Monogr. Phaner. 1. 204 (1878). Banks Group: Vanua Lava, common in rain-forest at 500 M., no. 483, July 16, 1928 (vine growing over rain-forest trees; fruit Purple to black when ripe).—Also Fiji and Bismarck Islands. Smilax sp. Aneityum: Anelgauhat Bay, common in rain-forest at 210 M., no, 819, Feb. 28, 1929 (vine growing over rain-forest trees; fruit 1.3 em. long, 1.5 cm. in diam., black when ripe). ees eae cymosum A. Cunningham in Bot. Mag. t. 3131 2). Aneityum: Anelgauhat Bay, common in rain-forest at 24 level, no. 832, Feb. 28, 1929 (vine growing over rain-forest trees; i when ripe).—Also Loyalty Islands, Australia (Queen Neer 112 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Dianella ensifolia [A. P. DeCandolle in] Redouté, Lil. 1. 1, t.1 (1802). Eromanga: Dillon Bay, common in poor red soil, open country at 400 m., no. 346, June 1, 1928 (plant about 1 m. high; flowers pale blue).—Also New Caledonia, Australia (Queensland, North Australia), Fiji, ‘Tonga, Society, Caroline and Mariana Is- lands, New Guinea, Malaysia and Hawaii.—Vernacular name “Did- and-Did.” Dracaena sp. Aneityum: Anelgauhat Bay, common in rain-forest at sea level, rio. 881, March 6, 1929 (up to 6 m. high; fruit 2.5 cm. long, 3 em. in diam.; black when ripe; leaves used for making grass ski Wall. of the Philippines and Asia. A " i Ke Fic. 1—CHAMBEYRENL aoe eee rope a of frulte (Pe YRENIA sp.: 1. Fruit. 2. Longitudinal section of CyPHOPHOENIX sp.: 3. Fruit. 4. Longitudinal : nal section of fruit. rts). | The leaves of the plant resemble exactly those of D. aurantiaca — HP. 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 118 PALMAE Chambeyronia sp. aff. C. macrocarpa Vieill. Fig. 1: 1,2 Banks Group: Vanua Lava, rare in rain-forest at sea level, no. 428, July 6, 1928 (Palm up to 15 m. high; fruit edible). Cyphophoenix sp. nov. ? Fig. 1: 3,4 Banks Group: Vanua Lava, common in rain-forest at 100 m., no, 465, July 11, 1928 (tall palm 15 m. high; fronds about 2 m. long). The fruit resembles completely that of C. elegans H. Wendl. of New Caledonia, the only species of the genus, but differs in its smaller size (1.8 cm. X 0.4 cm.), its turbinate shape and its smooth pericarp. Kajewskia aneityensis Guillaumin, gen. nov. et sp.nov. Fig. 2 Palma 20 m. alta, foliis ultra 4 m. longis pinnatisectis, segmentis lanceolatis (circa 60 cm. X 8 cm.) acuminatis in axilla subtusque ad costae basin brunneo-filamentoso-squamatis. Spadix valde Ee. 2.—Kasewsk1a Staminate flower. 2. ANEITYENSIS Guillaumin: 1. 5 Con cc section of staminate flower. 3. Stamen. 4, Pathe te flower. 6. Longitudinal section of pistillate flower. ‘amosus, ramis angulosis brunneo-squamatis, floriferis circa 15 cm. : bongs valde sinuosis, floribus ¢' et @ intermixtis singulis vel 2-nis — paete 2, & ellipsoideis ° mm. X 7 mm.) vertice rotunda- 114 JOURNAL OF THE ARNOLD ARBORETUM [rot xi of tis symmetricis, sepalis rotundatis ad 4 mm. longis imbricatis, — petalis ellipticis apicem versus attenuatis crassis valvatis, stamini- — bus ©, antheris linearibus apice recurvis undulatis 4 mm. longis, — loculis parallelis basi leviter divaricatis, filamentis duplo brevioribus connectivo nigro continuis, pistillodio basin versus incrassato im | collum tenuem attenuato stamina superante, 2 globosis (8 mm. b: diam.), sepalis laxis valde imbricatis ad 8 mm. longis ovato-rotun- d datis, petalis inclusis valde appresse cucullato intricatis 7 mm. — longis ovato-rotundatis, staminodiis 0, ovario oblonge turbinato : tertia suprema parte conice stigmatoso, ovulo 1 parietali ad loculi | basin inserto. ' Aneityum: Anelgauhat Bay, not common in rain-forest at — sea level, no. 784, Feb. 19, 1929 (large palm to 18 m. high; fronds up i to 3.5 m. long; stamens white, anthers cream-colored). a Near Actinokentia, a genus represented only by A. divaricatt — Dammer (= A. Schlechteri Dammer) of New Caledonia, but differs — in its much larger flowers resembling those of Arenga, not regularly arranged in three’s (2 lateral and 1 central flower) and in the ab- sence of the staminodes in the pistillate flower. 68 PANDANACEAE Determined by U. Marte. : u —— Cominsii Hemsley in Hooker’s Icon. Pl. xxvu. t. 2604 Banks Group: Vanua Lava, common in rain-forest at 100 m., no. 471, July 12, 1928 (up to 4 m. high, much smaller than af coastal relative; fruit red when ripe, on a cylindrical spike).— found in Solomon Islands. Freycinetia tannaensis Martelli in Jour. Arnold Arb. xi 1930). Plate 8 Tanna: on Mt. Tokosh Meru, very common in rain-forest, | 800 m., no. 163, March 15, 1928 (climbing up the trunks of tree ARACEAE ee Epipremnum pinnatum (L.) Engler in Engler, Pflanzenr. rv. 8 p. 60 (Arac.) (1908). Macnee i Aneityu m : Anelgauhat Bay, common in rain-forest at 210 m., no. 830, Feb. 28, 1929 (climbing up the trunks of trees: fr el ere long).—Already found on Tanna; also in New © al ana a Australia (Queensland), Fiji, Tonga, Marshall 3° marck Islands, New Guinea, Moluccas and Malaysia. Jour. ARNoLD Ars., Voi. XIII. | PLATE 43 INAENSIS Martelli go Bh ARES TEIN EG ALL LLNS Photograph of type in the Herbarium Martelli ape he 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 115 CYPERACEAE! Pycreus polystachyus Beauvais, Fl. Owar. 11. 48, t. 86 (1807). Eromanga: Dillon Bay, common in cleared rain-forest at sea level, no. 277, May 19, 1928 (small sedge 30-40 cm. high).—Also New Caledonia, Australia (Queensland, New South Wales), Samoa, Cook and Bismarck Islands, New Guinea, Malaysia.—Vernacular name “Overi-car.”’ Kyllingia monocephala Rottboell, Descr. Ic. 13, t. 4, fig. 4 (1773). Eromanga: Dillon Bay, commonly growing in cleared rain- forest at sea level, no. 276, May 19, 1928 (sedge 25-40 em. high; macerated with leaves of ‘“‘Denyung”’ (Cane grass) in cold water and drunk for spleen trouble; macerated with leaves of “‘Naivoss,”” “Tomirirri” and ‘‘Nesivinesip” (shrub) in cold water and drunk by women in state of pregnancy for good health).—Already found on Aneityum, waste place (MacGillivray, no. 53, December 1858 {unpublished locality]) and on Efate; also New Caledonia, Aus- tralia (Queensland, New South Wales), Norfolk Island, Fiji, Tonga, Samoa, Cook, Society, Caroline, Mariana, Santa Cruz, Solomon and Bismarck Islands, New Guinea and Malaysia—Vernacular name “Polell,’ Fimbristylis communis Kunth, Enum. Pl. 1. 234(1837). Eromanga: Dillon Bay, common in open grass country at 300 m., no. 318, May 28, 1928 (sedge about 50-75 cm. high; plaited together by women for grass dresses).—Already found on Aneityum, swamp (MacGillivray, no. 24, Oct. 1858 [unpublished locality] ).— Also New Caledonia, Australia (Queensland, New South Wales, North Australia), Tonga, Cook, Society, Mariana and Bismarck Islands, New Guinea and Malaysia.—Vernacular name “Farll.” Lepidosperma sp. cf. L. elatius Labill. of Australia (Victoria, Tasmania), Tanna: Mt. Tokosh Meru, common in rain-forest at 1000 m., no, 155, March 15, 1928 (sedge on very top of the mountains). Cladium Milnei C. B. Clarke in Kew Bull. Add. ser. 8,p.46 (1008). : Eroma nga: Dillon Bay, common in poor red soil, bracken “ountry, no. 327, May 29, 1928 (about 1 m. high; seed bright papery! ~~“Mlready found on Aneityum.2—Vernacular name Did-and- : | To the ts ci * “ye lantes ues [des Nouvelles Hé ) $ ssc am i, ee ring cma eS k . * or iz ie le > . ora ke seem Maclillivray, n. 7), Sept. 1879; Ryncho- Nouvelles H 116 JOURNAL OF THE ARNOLD ARBORETUM (vou. xm GRAMINEAE Determined by A. 8. Hircucock Isachne distichophylla Munro in Jour. Bot. vit. 178 (1869), nom, | nud.; Hillebrand, Fl. Hawaiian Isl. 504 (1888). Tanna: Mt. Tokosh Meru, common in rain-forest at 1000 m., | no. 147, March 15, 1928 (small grass right on the top of Tokosh — y Meru).—Also in Hawaii. Brachiaria subquadriparia (Trin.) Hitchcock in Lingnan Sci. Jour. | vir. 214 (1931). Eromanga: Dillon Bay, common in cleared rain-forest at sea level, no. 278, May 19, 1928 (short grass 30-40 cm. high; one of the commonest grasses in the islands, growing among the cocoaiuts _ after the scrub has been felled; the natives say it is introduced).— Also Australia (Queensland, South Australia), Mariana and Bis- marck Islands.—Vernacular name ‘‘ Noth-lor-si.” Brachiaria ambigua (Trin.) A. Camus in Lecomte, Fl. 6ér Indo-Chine, vir. 433 (1922). | | Eromanga: Dillon Bay, common in clearings in rain forest at sea level, no. 279, May 19, 1928 (grass about 60 cm. high; intro- duced).—Already found in Aneityum; also New Caledonia, Fy Samoa, Cook, Mariana and Solomon Islands. | Mus. Honolulu, visi. 177 (1922) | Eromanga: Dillon Bay, common in rain-forest clearings # sea level, no. 280, May 19, 1928 (grass 0.60-1 m. high; cattle are a le phen: of it).—Also in South America.—Vernacular name "N_ e-Tme. | therisma microbachne (Presl) Hitchcock in Mem. Bisho? Eromanga: Dillon Bay, common in rain-forest clearité pe Pit no. 281, May 19, 1928 (grass about 80 em. high in the es ) Pir, found on Aneityum, Epi and Espiritu Santo;® i aad ds ledonia, Loyalty, Fiji, Society, Gambier and Marques? | and Hawaii.—Vernacular name “‘Nooh-nempeli.” Oplismenus compositus (L.) Beauvois, Essai Agrost. 54 (l i! ' ety: . a Cenchrus calyculatus Cavanilles, Ic. v. 39. t. 463 (1799). Eromanga: Dillon Bay. common in rain-forest at ees ee pees 1928 (tall grass about 1.25 m. high; eet Calokinis ee found on Espiritu Santo; also a Musiatis Inude wat a) Cook, Society, Gambier, Caroling * and Hawaii.—Vernacular name “ Ularhal. 1932] . GUILLAUMIN, FLORA OF THE NEW HEBRIDES 117 Coix lacryma-jobi Linnaeus, Sp. Pl. 972 (1753). Tanna: Lenakel, common in rain-forest clearings at 200 m., no. 118, March 6, 1928 (weed in native gardens).—Already found on Tanna; also New Caledonia, Queensland, Fiji, Tonga, Samoa, Society, Marshall, Solomon, Bismarck Islands, New Guinea and Malaysia. Miscanthus sinensis Andersson in Oefv. Svensk. Vetensk.-Akad. Stockh. 1855, p. 166. Tanna: Lenakel, common in rain-forest soil at 100 m., no. 39, Feb, 22, 1928 (tall cane 4-6 m. high; used by natives for arrows and roofs of houses).—Already found in Aneityum; also New Caledonia, Fiji, Tonga, Cook, Society Island and Malaysia. Centotheca latifolia Trinius, Fund. Agrost. 141 (1820). Eromanga: Dillon Bay, common in semi-cleared rain-forest at sea level, no. 338, May 29, 1928 (chewed by natives and applied to burns).—Already found in Aneityum and Tanna; also New Caledonia, Australia (Queensland, North Australia), Fiji, Tonga, Samoa, Cook, Society, Marquesas, Caroline, Mariana, Bismarck and Admiralty Island, New Guinea and Malaysia.—Vernacular name “ Now-now.”’ , CONIFERAE Podocarpus imbricatus Blume, Enum. Pl. Javae, 89 (1827). Podocarpus cupressinum R. Brown apud Mirbel in Mém. Mus. Hist. Nat. Paris, xm. 75 (Geogr. Conif.) (1825), nomen.—R. Brown apud Bennett, Pl. Jav. Rar. 25, t. 10 (1838-52). Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 849, Aug. 2, 1929 (up to 12 m. high; seed about I cm. long, lcm. in diam. resting on a fleshy growth 0.5 cm. long).—Already found on Aneityum; also Fiji Islands, Malaysia and Philippines. Agathis obtusa (Lindl.) Masters in Jour. Roy. Hort. Soc. Lond. xvi. 197 (1899). Aneityum: Anelgauhat Bay, common in rain-forest at 150 M., no. 706, Feb. 5, 1929 (large tree up to 40 m. high on well drained slopes where the soil is intensely volcanic and «FEE PERTES typical of the country where the Kauri is found on the islands; this specimen is remarkable for the small size of the cones and veal sien! sunken places on their surface; the best Se - building j i - common in rain- uilding in the southern hemisphere); Cs ataaeed); *mmon in rain-forest at 100 m., no. 707, found on Aneityum and Eromango.—Determined by . #- 118 JOURNAL OF THE ARNOLD ARBORETUM [vou xm CYCADACEAE Cycas circinalis Linnaeus, Sp. Pl. 1188 (1753). Eromanga: Dillon Bay, common in rain-forest and open | country at 300 m., no. 400, June 8, 1928 (Cycad up to 10 m., high; fronds up to 2 m. long; fruit yellow, borne in fours).—Already found in the New Hebrides; also New Caledonia, Loyalty Islands, Aus- tralia (North Australia), Fiji, Tonga, Caroline, Mariana, Solomon, Bismarck and Admiralty Islands, New Guinea, Malaysia.—Ver- nacular name “‘No-mall.”’ ‘ This is perhaps the same species as that named C. neo-caledonica, — by Linden (without description). | PTERIDOPHYTA DrETeRMINED By E. B. CopeLanpD HYMENOPHYLLACEAE Trichomanes Boryanum Kunze, Farnkrauter, 237 (1847). Aneityum: Anelgauhat Bay, common in rain-forest a 150 m., no. 883, March 7, 1929 (about 15 cm. high). Banks | Grou p: Vanua Lava, common in rain-forest at 400 m, 2 | rain-forest where the rain-fall reaches 250 inches per annum, 20 461, July 10, 1998, ce icho Bauerianum Endlicher, Prodr. Fl. Norf. 17 (1888). ; Aneityum: Anelgauhat Bay, common in gullies in forest at 600 m., no. 864, March 5, 1929 (growing in shade at high { altitudes). Tanna: Mt. Tokosh Meru, common in rain-forest | at 800 m., no. 164, March 15, 1998. tat Trichomanes meifolium Bo | Spec Eee 509 (1810). ry apud Willdenow, spec | A neityum : Anelgauhat Bay, common in rain-forest mou ‘ ‘ais at 750 m., no. 867, March 5, 1929. a Banks Group: Vanua Lava, common on mountains a erates yd at —_ ape the rain-fall reaches about 250 inche : ‘per annum, no. 454, July 12, 1998. Babe The specimen which is sterile may belong to T. aphlebioid® : Christ.» . epee ‘ apriresier: formosum Brackenridge in Bot. Wilkes ue 7p EXD. XVL 268, t. 82, fig. $ (1854). ‘/- 50 me Anelgauhat Bay, common in rain-forest = M+» Nos. 868 and 871*, March 5, 1929 (small fern oF" ” upon rain-forest trees), 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 119 CYATHEACEAE Balantium, an B. dubium? Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 885, March 9, 1929 (large tree fern; trunk up to 1.5 m. high; fronds up to 2.8 m. long). Balantium stramineum (Labill.) Diels in Engler & Prantl, Nat. Pflanzenfam. I.—4, p. 119 (1899). Aneityum: Anelgauhat Bay, common in rain-forest at 210 m., no. 859, March 4, 1929 (fronds up to 2.8 m. long; main stem growing close to the ground). Cyathea laciniata Copeland in Univ. Calif. Publ. Bot. xu. 889 (1931). Aneityum: Anelgauhat Bay, common in rain-forest at 600 m., no. 876, March 5, 1929 (small tree fern up to 3.1 m. high; fronds up to 1.5 m. long). - Cyathea leucolepis Mettenius in Ann. Mus. Bot. Lugd.-Bat. 1. 56 (1863). Aneityum: Anelgauhat Bay, common in rain-forest at 210 m., no. 861, March 4, 1929 (tree fern up to 6 m. high; fronds up to 3 m. long). Cyathea medullaris (Forst.) Swartz in Jour. Bot. Schrad. 1800, pt. 11. 94 (1801), Aneityum: Anelgauhat Bay, common in rain-forest up to 210 m., no. 857, March 4, 1929 (tree fern up to 9 m. high; fronds up to 3 m. long). Tanna: Lenakel, common in rain-forest at 150 m., no. 54, Feb. 24, 1928 (tree fern up to 8 m. high; fronds 3-3.5 m. long), POLYPODIACEAE Dryopteris rubrinervis (Mett.) Christensen, Ind. Fil. 289 (1905). Tanna: Lenakel, common in rain-forest soil at 200 m., no. 61, Feb. 24, 1998 (large fern in gullies; fronds about 2 m. long). _ ‘robably identical with D. urophylla Christens. Dryopteris urophyila (Wall.) Christensen, Ind. Fil. 299 (1905). Aneityum: Anelgauhat Bay, common in rain-forest at 210 m., no. 858, March 4, 1929 (large fern; trunk small at base up to about 0.3 m. high; fronds up to 2.1 m. long). Dryopteris sp. : ; ; Aneit yum: Anelgauhat Bay, common lm rain-forest @ 210 ™., no. 860, March 4, 1929 (main stem not more than 0.3 m. high; fronds up to 2.1 m. long). + ats 120 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Dryopteris sp Tanna: Lenakel, common in heavy rain-forest at sea level, no. 5, Feb. 20, 1929 (fern growing in clumps, 0.6—1 m. high). Dryopteris sp. . Tanna: Lenakel, common in heavy rain-forest at sea level, : no. 5, Feb. 20, 1928 (fern growing in clumps, 0.6—1 m. high). } Polystichum aristatum (Forst.) Presl, Tent. Pterid. 83 (1863). Aneityum: Anelgauhat Bay, common in dense shade in rain-forest at 300 m., no. 847, March 2, 1929 (fronds up to 1 m. long). Campium lonchophorum (Kunze) Copeland, comb. nov. Acrostichum lonchophorum Kunze, Farnkrauter, pl. 2 (1840). Aneityum: Anelgauhat Bay, common in rain-forest at 450 m., no. 893, March 9, 1929 (fronds up to 0.60 m. long). Dipteris conjugata Reinwardt in Syll. Pl. Ratisb. 1. 3 (18%). Aneityum: Anelgauhat Bay, common on banks of rail- forest streams at 150 m., no. 916, March 17, 1929. Eromanga: Dillon Bay, common in rain-forest at 400 m., no. 325, May o 1928 patie, in clumps, about 60 cm. high) .Vernacninn name “Nenn-parr” (under no. 325). | Pasp ciliata Klotzsch apud Kuhn in Linnaea, XXXVL 128 Aneityum: Anelgauhat Bay, common at 750 m., 00. sh March 5, 1929 (fern climbing up trees; stem dark brown). 3 ( inn umata pusilla (Mett.) Carruthers in Seemann, Fl. Vit- $85, | Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 855, March 4, 1 t a outside the moss on trees), 929 (small climbing fern jus pearing Davallia solida Swartz in Jour. Bot. Schrad. 1800, pt. 1 87 ast) | Aneityum: : Anelgauhat Bay, common in rain-forest cei | | level, no. 835, March 1929 (fern with an underground begs st sometimes growing on the trunks of trees). Microlepia speluncae (L.) Moore, Ind. Fil. xcu. (1857)- oe ; . anna: Lenakel, common in heavy rain-forest S0 soil at © i vel, no. 6, Feb. 20, 1998 (80 cm. to 1.25 m. high). - Tapeinidium tenue P. Bae Mins, Bell cis bp aoa) Copeland in Bernice ey pete iy Anneliens:! at Bay, common in rain-forest 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 121 150 m., no. 894, March 9, 1929 (fern growing on ground; fronds up to 0.6 m. long). Dennstaedtia samoensis (Brack.) Moore, Ind. Fil. 307 (1861). Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 884, March 9, 1929 (tree fern with small trunk about 0.45 m. high; fronds up to 3 m. long). Lindsaya decomposita Willdenow, Sp. Pl. v. 425 (1810). Banks Group: Vanua Lava, common in rain-forest at 700 m., no. 458, July 10, 1928. Athyrium accedens (BI.) Milde in Bot. Zeit. 1870, p. 353. Tanna: Lenakel, common in rain-forest at 150 m., no. 133, March 8, 1928 (about 1 m. high). Athyrium sp. Aneityum: Anelgauhat Bay, common in rain-forest at 600 m., no. 877, March 5, 1929 (tree fern about 2.5 m. high; fronds up to 2 m. long). Athyrium sp. ‘anna: Lenakel, common in rain-forest soil at 200 m., no. 55, Feb. 24, 1928 (large fern; fronds about 2-2.5 m. long). : Asplenium affine Swartz in Jour. Bot. Schrad. 1800, pt. 1. 56 (1801), Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 888, March 9, 1929 (fern growing on trees or on the ground; fronds 1 m. long). Asplenium amboinense Willdenow, Sp. Pl. v. 303 (1810). Aneityum: Anelgauhat Bay, common in rain-forest at 210 M., no. 838, March 2, 1929 (small-leaved fern growing on § trees in dense shade). Asplenium brachycarpum (Mett.) Kuhn in Linnaea, XXXVI- 104 (1869). __Aneityum: Anelgauhat Bay, common in rain-forest at _ 600 m., no. 878, March 5, 1929 (small tree fern up to 1.5 m. high; tunk 45 cm. high; fronds up to 1 m. long). | jfgplenium laserpitiifotium Lamarck, Eneycl. Méth. u. 310 1786), Aneityum: Anelgauhat Bay, common in rain-forest . 300 m., no. 848, March 2, 1929 (fronds up to 1.5 m. long). Tan eae nakel, common in rain-forest at 100 m., no. 71, March 1,1 _ out 1m. high). Te 122 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm | Asplenium nidus Linnaeus, Sp. Pl. 1079 (1753). Banks Group: Vanua Lava, common in rain-forest at sea level, no. 444, July 12, 1928 (growing on trees, common throughout the group). Asplenium sp. Aneityum: Anelgauhat Bay, common in rain-forest at — $00 m., no. 887, March 9, 1929 (growing sometimes on trees, — sometimes on rocks; fronds up to 0.6 m. long). | Blechnum gibbum (Labill.) Mettenius in Ann. Sci. Nat. sér. 4 Xv. 68 (1861). Aneityum: Anelgauhat Bay, common in rain tome at 210 m., no. 856, March 4, 1929 (small tree fern up to 1.5 m. high, bearing distinct sterile and fertile fronds; fronds up to 1.3 m. long). ” Tanna: Ikiti, common along rain-forest streams at 100 my no. 171, March 16, 1998 (small tree fern up to 1.5 m. high; growing in clumps with 10 or more stems in one clump). Blechnum opacum (Baker) Mettenius in Ann. Sci. Nat. sé. s ‘ Xv. 69 (1861). pte eityum: Anelgauhat Bay, common in rain-forest at 600 m., no. 879, March 5, 1929 (small fern bearing distinct sterile and fertile fronds). : Blechnum Patersoni (R. Br.) Mettenius, Fil. Hort. Bot. ti 64, t. 4, figs. 4-10 (1856). Tanna: Mt. Tokosh Meru, common in rain-forest at 500 no. 156, March 15, 1928 (about 1 m. high). q Sphenomeris chusana (L.) Copeland in Bernice P. Bishop ws if Bull. urx. 69 (1999). \ : ae c eity a2 a Anelgauhat Bay, common in ee +f | no. March 9, 19 i : si to 0.65 m. - lone) 29 (growing on the groun (1882), peri Mettenius apud Kuhn, Chaetopt = Aneityum: Anel ee gauhat Bay, common in iio 750 m., no. 870, March 5, 1999 (small fern up to 15 cm. high). ay Nat hirsuta (Poir.) Mettenius in Abhandl. Sencieth ] wrt, iia iy 25, n. 17 (1859), : Anei rege yum: t Bay, common in the opel gi signee! ve wa level, no. "928 March 17, 1929. 1982] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 123 Adiantum aneitense Carruthers in Seemann, FI. Vit. 346 (1873). Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 889, March 9, 1929 (fronds up to 0.65 m. long). Adiantum diaphanum Blume, Enum. PI. Jav. 215 (1828). Tanna: Lenakel, common on sides of gullies in rain-forest at 200 m., no. 56, Feb. 24, 1928 (very small fern). Pteris comans Forster, Fl. Ins. Austr. Prodr. 79 (1786). Aneityum: Anelgauhat Bay, common on banks of creeks in rain-forest at sea-level, no. 719, Feb. 9, 1929 (fronds up to 3 m. long). Pteris ensiformis Burmann, FI. Ind. 230 (1768). Aneityum: Anelgauhat Bay, common in rain-forest at 150 m., no. 904, March 11, 1929 (up to 0.6m. high). Pteris tripartita Swartz in Jour. Bot. Schrad. 1800, pt. um. 67 (1801). Aneityum: Anelgauhat Bay, common along streams in ‘ane raha at 150 m., no. 892, March 9, 1929 (fronds up to 2.8 m. ng). Vittaria lineata (L.) Smith in Mem. Acad. Sci. Turin, v. 421, t. 9, fig. 5 (1798). Aneityum: Anelgauhat Bay, common in rain-forest rf 150 m., no. 934, March 17, 1929 (up to 0.3 m. high; growing on bark of trees). | Hymenolepis mucronata Fée, Gen. Fil. 82 (1850-52). Aneityum: Anelgauhat Bay, common in rain-forest at 210 m., no. 890, March 9, 1929 (climbing upon trees). Polypodium blechnoides (Grev.) Hooker, Sp. Fil. rv. 180 (1862). Aneityum: Anelgauhat Bay, common in rain-forest near the summit of the mountain at 450 m., no. 886, March 9, 1929 (small fern). Polypodium scolopendria Burmann, FI. Ind. 232 (1768). Eromanga: Dillon Bay, common in rain-forest at 500 m., nO. 380, June 7, 1928 (small fern).— Vernacular name Nemkar Wworamel,”’ Selliguea feeioides Copeland in Bernice P. Bishop Mus. Bull. UX. 94 (1929). A Aneityum: Anelgauhat Bay, common m rain-fo: * 'M., no. 875, March 5, 1929 (climbing upon trees). Tanna: 124 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm “Mt. Tokosh Meru, common in rain-forest at 800 m., no. 165, March 15, 1928 (climbing along the trunks of trees). Cyclophorus acrostichoides (Forst.) Presl, Epimel. Bot. 13) (1849). | Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 714, Feb. 9, 1929 (growing on tree trunks and climbing by means of runners; common throughout the islands). Tanna: Lenakel, common in rain-forest at 200 m., no. 101, March 5, 19% (parasitic fern). Drynaria rigidula (Sw.) Beddome, Ferns Br. Ind. t. 314 (1869). Aneityum: Anelgauhat Bay, common in rain-forest at se level, no. 834, March 1, 1929 (growing on trees; stem shielded by _ a brown heart-shaped growth). GLEICHENIACEAE Gleichenia Brackenridgei Fournier in Ann. Sci. Nat. sé 5 XVII. 269 (1873). Aneityum: Anelgauhat Bay, common in rain-forest at 750 m., no. 873, March 5, 1929 (large fern up to 1.8 m. highs # a much smaller form i is common on the waste-lands). Gleichenia linearis (Burm.) Clarke in Trans. Linn. Soc. set. 1. (Bot.) 428 (1880). | Tanna: Mt. Tokosh Meru, common in rain-forest: at 800 2 m., no. 161, , March 15, 1928. ie Sleichenia €mia oceanica Kuhn, Verh. Zool.-Bot. Ges. Wien, x 583 pe 3 Tanna: Mt. Tokosh M : b at 800 nae 1 60, h 15, } eTu, common in rain. fores . SCHIZAEACEAE (L.) Smith in Mem. Acad. Sci. rus * Schizaea 422, t. 9, fig. yay 1 Mater ss Anelgauhat Bay, common in rain-forest 1 : | Bider no. 918, March 17, 1929 ena plant up to 0.60 m- beh no. 286, Meo Dillon Bay, common in rain-forest at as : ‘In poor LAA 1928 (plant with single stem 40 cm.); 4 pet (small plan racken country at 400 m., no. 355, June > de 286). t 20 em. high).—Vernacular name “Nempati a | i no. 28) A remeriahes be to the collector’s note under no. 286 this plant > 1a worn n when a war was on in old 3 1932] GUILLAUMIN, FLORA OF THE NEW HEBRIDES 125 days it was a sign that the wearer was a non-combatant and was respected as such; the name “‘Nempari”’ means peace-maker. Lygodium reticulatum Schkuhr, Kryptog. Gew. 1. 139, t. 139 (1809). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 882, March 7, 1929 (climbing as high as 9 m. into trees). Eromanga: Dillon Bay, common in poor red soil, bracken country, at 300 m., no. 303, May 24, 1928 (climbing fern with a twining stem often 3-4 m. long; used by natives as a twine for house-building).—Vernacular name “Nor-rep”’ (under no. 303). OSMUNDACEAE Leptopteris Wilkesiana (Brack.) Christ, Farnkr. d. Erde, 334 (1897). Aneityum: Anelgauhat Bay, common in rain-forest at 750 m., no. 869, March 5, 1929 (small tree fern up to 1.8 m. high; fronds up to 1 m. long). OPHIOGLOSSACEAE Ophioglossum pendulum Linnaeus, Sp. Pl. ed. 2, m. 1518 (1763). Aneityum: Anelgauhat Bay, common in rain-forest at 210 m., no. 825, Feb. 28, 1929 (parasitic fern hanging down from a large Asplenium). Tanna: Lenakel, not common in rain- at 100 m., no. 139, March 8, 1928 (parasite on a large Banyan ig). LYCOPODIACEAE Lycopodium cernuum Linnaeus, Spec. Pl. 1103 (1753). Aneityum: Anelgauhat Bay, common in shade of rain- forest at 150 m., no. 854, March 4, 1929 (climbing Lycopod). Tanna: Mt. Tokosh Meru, common on shad in ran forest at 900 m., no. 158, March 15, 1928. Efate : Undine Bay, common in shade of bracken on high open hills at 250 m., no. 224, April 27, 1998. | Lycopodium oceanianum Herter in Bot. Jahrb. xuut. beibl. 98, P. 52 (1909). . ey Efate: Undine Bay, Mt. McDonald, common in rain-forest at 600 m., no. 239, April 29, 1928 (parasite on trees). —Vernacular fame “Tassel Fern.” Lycopodium Phlegmaria Linnaeus, Sp. Pl. 1101 (1753). _ Aneityum: Anelgauhat Bay, common in raietorest 3 210 m., no. 824, Feb. 28, 1929 (growing upon trees). — 126 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm Lycopodium squarrosum Forster, Fl. Ins. Austr. Prodr. 86 (1786). Aneityum: Anelgauhat Bay, common in rain-forest at 300 m., no. 843, March 2, 1929 (spore cases light yellow before they burst, easily discernible at end of stem). PSILOTACEAE Psilotum complanatum Swartz, Syn. Fil. 188, 414 (1806). Aneity um: Anelgauhat Bay, common in rain-forest at 150 m., no. 919, March 17, 1929 (parasite, up to | m. high; growing from holes and in forks of trees; sporangia cream-white, minute). Eromanga: Dillon Bay, common in rain-forest at 200 m., no. 377, June 7, 1928 (parasite hanging down from trees).—Ver- nacular name “‘Ami-ghokul” (under no. 377). . Psilotum nudum (L.) Grisebach in Abh. Ges. Wiss. Gottingen, vit. 278 (1857). Tanna: Lenakel, rare in rain-forest at 150 m., no. 134, March 8, 1928 (parasite growing in crevices of roots of Banyan trees). ADDITIONS Vol. x11. 230 ° Heritiera litoralis Dryand.—Add: Already found on Efate. Vol. xu. 260. Add under SamypDAcEsE: Casearia Melistaurum Sprengel, Syst. 1. 354 (1825). Aneityum: Anelgauhat Bay, common in rain-forests up to 500 m., no. 981 (coll. J. P. Wilson), Sept. 1929 (small tree — trunk about 45 cm. diam.; flowers green, fruit clusters at base . leaves).—Also New Caledonia.— Vernacular name “ Nidu-pualau- Vol. xu. 23. Add before CorpIA Myxa Contin. subcordate, Tamarck, ‘Tabl. Encycl. Méth. + ime 1899 (1791). Efat e: Fila Island, Vila, common on sea coast, 0, 184) April 13, 1928 (tree about 8 m. high; natives say the fruit is eaten): —Already found on Efate; also New Caledonia, Loyalty Islands, Australia (Queensland), Society and Marquesas Islands, New Guinea and Malaysia. MUSEUM NATIONAL D’HISTOIRE NATURELLE Paris. : ee tee ee Pe Le ee ae ge ee be dee ere ot eee Meee er. ee bell 1982] AMES, FLORA OF THE NEW HEBRIDES 127 CONTRIBUTION TO THE FLORA OF THE NEW HEBRIDES AND SANTA CRUZ ISLANDS ORCHIDS COLLECTED BY S. F. KAJEWSKI IN 1928 AND 1929 Oakes AMES A. GuiiiauMtn, in his enumeration of the species that comprise the flora of the New Hebrides, included less than forty represent- atives of the Orchidaceae. With the exception of Corymborchis, Bulbophyllum, and Habenaria which is represented in the tropics of both hemispheres and, if broadly interpreted, extends into the _ temperate zones, all of the genera enumerated by Guillaumin are paleotropical. Collections recently made by S. F. Kajewski in 1928 and 1929 for the Arnold Arboretum have increased substantially the orchid flora of the New Hebrides and the adjacent islands of the Santa Cruz Group. Several of the genera found by Kajewski are additions to the region and ten of the species are new. Unfortunately some of Kajewski’s material was collected without flowers and admits only of tentative or suggestive identification. _Two courses are open to the student of a flora which is but little known and which shows a close relationship to the floras of distant regions: he may disregard slight differences between species in hand and species from other localities, and recast pub- lished descriptions so that they will include his material, or he _ May throw emphasis on slight differences and, in anticipation of More intensive exploration and a wider knowledge of the flora, interpret specific characters in a very narrow sense. In the follow- ing treatment I have adopted the latter course, being prompted to do so by the conviction that an erroneous extension of range as -& genus or species is more detrimental to scientific progress ‘tan the increase of synonyms. Corymborchis veratrifolia (Reinw.) Blume, Fl. Jay. nov. ser. t e 105, t. 43, fig. 1 (1858). _ Tanna: Mt. Tokosh Meru, terrestrial, in rain-forest at 200 m»_ 10. 146, March 14, 1928 (pretty, white-flowered ground orchid, } m- h; sweetly scented).—Already found on Espiritu Santo (Camp- The flowers are somewhat smaller than in typical material. A ry smaller flowered plant with narrower leaves, collected by Dr- Morrison on Efate, August 20, 1896, appears to belong here. 5 eo Kew.). It is probable that the specimens examined — 128 JOURNAL OF THE ARNOLD ARBORETUM [vou xm resent a new species, but until more material is available it would be unwise to separate Kajewski’s specimens from C. veratrifolia. C. veratrifolia has already been reported as a native of the New Hebrides. It is a widely distributed species ranging from India to New Guinea. Malaxis neo-ebudica Ames, sp. nov. | Herba terrestris. Caules elongati, foliosi. Folia ovato-lanceolata, acuminata, in petiolum sulcatum contracta. Pedunculus elongatus. Bracteae inflorescentiae anguste lineares. Racemus cylindraceus, | elongatus, multiflorus, floribus flavidis. Pedicellus cum ovario . quam bractea subtendens paulo brevior. Sepala lateralia elliptica, | uninervia, obtusa, in sicco pellucida. Sepalum dorsale anguste — ellipticum, valde obtusum, uninervium. Petala lineari-oblonga, — obtusa, uninervia. Labellum trilobatum, lobi laterales post — columnam in auriculas elongatas producti, lobus medius plus minusve triangulus, haud retusus, utrinque paucidentatus. : Stem up to 12 em. long, ascending, bearing about 10 membra- — naceous leaves. Leaves variable, the lowermost ones much the | smallest, uppermost leaves about 13 cm. long, up to 3 cm. wide, — contracted at base into broad sheathing petioles, lanceolate, acuminate-acute. Peduncle slender, conspicuously angulate when | dry, together with the raceme about 14 cm. long. Raceme slender — about 10 cm. long, many-flowered. Bracts of the inflorescence — elongated, linear-lanceolate, about 7 mm. long, exceeding Ut — flowers. Pedicellate ovary about 5 mm. long. Flowers yellow. : Lateral sepals 2 mm. long, about 1 mm. wide, elliptical, obtus® : l-nerved, spreading. Upper sepal similar, but slightly narrower. — Petals about 2 mm. ling: te linear-oblong, obtuse, 1-nerved. at lum 3-lobed, including the lateral lobes 3.5 mm. long. Lateral | lobes auriculate, about 2 mm. long, narrowly triangular, acute % 4 obtuse, twice or thrice longer than the column. Middle lobe 14 _ mm. long from tip to point of attachment, triangular, terminating | in a bluntly triangular lobule on either side of which see’ or : 2-3 irregular teeth ? Tanna: Lenakel, small plant growing in shade of are tes flowers yellow; alt. 100 m. in rain-forest. No. 137 (type im Ames no. 36366), March 8, 1928. The nearest ally of this species appears to be M. agg) cn i. O. Ktze. from which it differs in the structure of the lip. fe (Reichb. f.) O. Ktze. is also a closely allied species, bat vegetatively from M. neo-ebudica. Malaxis xanthochila hese Ames & Schweinfurth i in A Orch. v1. % hanes | .. i: 1932] AMES, FLORA OF THE NEW HEBRIDES 129 Microstylis xanthochilaSchlechter in K.Schumann & Lauterbach, Nachtr. Fl. Deutsch. Schutzgeb. 102 (1905). Aneityum: Anelgauhat Bay, common in rain-forest at sea level, no. 837, March 1, 1929 (small plant up to 1 ft. high; flowers yellow). —Also New Guinea. Two specimens which are referable to this species, one collected by Dr. R. Morrison on Aneityum at Anelgauhat in June 1896, the other collected by L. Cheeseman at South West Bay, Malekula in January 1930, are preserved in the Kew Herbarium. Although the basal auricles of the lip in the Malekula plant are longer than is typical, it agrees in essential characters with M. zxanthochila. Malaxis lunata (Schltr.), comb. nov. (Microstylis lunata Schlechter in Fedde Rep. Spec. Nov. IX, 162 [1911]) differs from M. xantho- chila in the basal auricles of the labellum and in the spatulate petals. The type specimen of M. lunata was collected by Dr. Morrison on Aneityum near Anelgauhat. Liparis condylobulbon Reichenbach f. in Hamb. Gartenz, xvmI" 34 (1862). Liparis confusa J. J. Smith, Fl. Buitenz. v1. (Orch. Jav.) 275 (1905). Eroman ga: Dillon Bay, common in rain-forest at 400 m., no. 321, May 29, 1928 (growing on trees thickly shaded from the sun). Banks Grou p: Vanua Lava, common in rain-forest at 350 M., no. 455, July 10, 1928 (orchid growing in dense shade on small tree in gully; perianth organs small, outer ones pale green, inner ones brown).— Vernacular name “ Ute-melme-pohl ” (under no. 32 1). In the Kew Herbarium there are five specimens representing this species, obtained in Aneityum in 1896 by Dr. R. Morrison. This is one of the most widely distributed species of the genus, very common in the Philippines, Sumatra, Java, Celebes, eo and New Guinea. woelogyne rata Lindley in Jour. Hort. Soc. 1v. 221 (1849). Sa nta Cruz Group: Vanikoro, common in rain-forest at ea level, no. 679, December 4, 1928 (beautiful, large-leaved orchid on trees in Vanikoro; leaves dark green; flower yellow- =m, orange center; most beautiful of Vanikoro orchids). ; The Specimen collected was in bud, and not wholly satisfactory 2 fir identification, but there is sufficient evidence to warrant referring eile the widespread C. asperata, a species which ranges from the Mala Peninsula to Sumatra, Borneo, the Philippine Islands and ie: 7 y lamellata Rolfe in Kew Bull. 1895, 36. med eain-forest at : Mhadee Cruz Group: Vanikoro, common in rain-forest 3 Bei: el, no. 687, December 5, 1928 (large-leaved orchid common in 4 * Scrub on large trees; flower light yellow-green). — 130 JOURNAL OF THE ARNOLD ARBORETUM [vom Pholidota imbricata Lindley in Hook. Exot. Fl. 11. t. 138 (1825). Santa Cruz Group: Vanikoro, common in rain-forest at sea level, no. 681, December 4, 1928 (orchid common on trees close to salt water creek; fruit yellow, brown when ripe). Eromanga: Dillon Bay, common orchid on trees in gullies in rain-forest at 300 m., no. 390, June 8, 1928 (capsules opened when ripe and the seeds — put on the natives’ faces as a substitute for the face powder of i commerce). ~ : The material, on which the identification of no. 681 rests, hb devoid of flowers. It is highly probable that this material is conspecific with P. grandis Kraenzl. apud Guillaumin in Bull. Soc. Bot. France, uxxvi. 301 (1929), non Ridl. in Jour. 5tr. Br. Roy. | Asiat. Soc. xix. 32 (1907). Kraenzlin described P. grandis with — the aid of fruiting specimens to which only vestiges of the flowers i adhered. In view of the occurence of what appears to be P. im | bricata on Eromanga, and taking into consideration the extensive — range of this species, from India to Burma, Assam, Siam, Malay — Peninsula, China, Sumatra, Java, the Philippine Islands, Celebes _ Borneo and New Guinea, I suspect that P. grandis Kraenzl. 8 — referable to it.— Vernacular name “ Utnimpoll” (under no. 390). Pseuderia vanikorensis Ames, sp. nov. Caules elongati, vaginis foliorum omnino obtecti, plus minusve — flexuosi. Folia disticha, anguste lanceolata, usque ad apicem — attenuata, acuta. Vaginae foliorum tubulatae, arcte adpressae: Racemi pauciflori, laterales. Sepala lateralia falcata, valde car nosa, leviter carinata. Sepalum dorsale spathulatum, concavils culum, plus minusve rigidum, prope apicem carinatum. etal 4 oblongo-ligulata, apicem versus paulo latiora, obtusa, uninervia ° . Labellum ellipticum, subacutum, dense papillosum; callus linearis in disco decurrens. Columna carnosa, arcuata, in pedem breve™ producta, labello articulata. a Stems 5-6 dm. long, slender, about 6 mm. in diameter, cones” ' by the sheathing bases of the leaves, more or less flexuose. Leave numerous, distichous, 13-14 cm. long, about 2.5 cm. wide, al em. apart, narrowly lanceolate, tapering gradually to an acute UP — at the base and contracted into a short sulcate peu” Petiole articulate with a tubular sheath. Racemes produced ir the sepals shorter than the dorsal sepal, 8 mm. long, about 2.5-3 “~ q 1932] AMES, FLORA OF THE NEW HEBRIDES 131 Dorsal sepal 1 cm. long, 2 mm. wide above the middle, concave- spatulate, obtuse, conspicuously carinate near the tip on the outer surface, arcuate. Petals about 7.5 mm. long, 1.5 mm. wide, oblong- ligulate, obtuse, slightly broader near the tip than at the base, I-nerved, slightly curved, fleshy. Labellum about 5 mm. long, hardy 4 mm. broad at the middle, elliptic, subacute, densely papillose, margin ciliolate. Disc with a fleshy keel extending from the base of the labellum almost to the apex on the inner sur- face; keel suleate at the base. Column 5 mm. long, arcuate, produced at base into a short foot with which the labellum is articulated. Santa Cruz Group: Vanikoro, in rain-forest at 150 m., no. 669 (type in Herb. Ames, no. 37190), December 3, 1928 (large climb- ing orchid hanging from rain-forest trees; leaves light green and glossy; flowers purple splashed with yellow). In general facies this species closely resembles Pseuderia frutex Schltr. but differs in the form of the labellum. Dendrobium (§ Aporum) sp. Efate: Fila Island, Vila, common on rocky beach, no. 190, April 14, 1928 (small orchid growing on large rocks close to salt water). The specimens are in fruit and indeterminable. Vegetatively the plants resemble closely Dendrobium ventricosum Kraenzl. from the Philippines. _ Dendrobium Goldfinchii F. v. Mueller in Wing’s South. Sci. Ree. Wan, 1883), Santa Cruz Grou p: Vanikoro, common at sea level, no. * 636, November 20, 1928 (growing on rain-forest trees).—Already Cry q on Aurora (Im Thurn, no. 341).—Also New Guinea and Admiralty Islands. Dendrobium Kajewskii Ames, sp. nov. : ey : Caules elongati, foliosi. Folia coriacea, elliptica vel eunraee- > apice inaequaliter bilobata. Pedunculi validi, elongat.. _ “Semi multiflori. Sepala lateralia anguste triangularia, undulata, : yeutum formantia, Sepalum dorsale simile. Petala ligulata. ; Tabellum, trilobatum; lobi laterales rotundati, verrucosl; lobus _-medius ovatus; discus tricarinatus. Columna sections. . . Roots elongated, about 2 mm. in diameter, whitish. Stems . about 2.5 dm. or more tall, jointed, the upper internodes swollen; nodes 3-$.5 em. apart. Leaves coriaceous, 5-11 em. long, 2.5-$ = wide, elliptic to elliptic-oblong, unequally bilobed at the —? Contr: at base into closely appressed complanate sheaths, 132 JOURNAL OF THE ARNOLD ARBORETUM [voL. xim 3 only the uppermost leaves persisting at anthesis. Peduncles including the raceme elongated, up to 4 dm. long, 3 mm. or more in diameter, rigid, obliquely ascending, produced opposite the leaves, paucibracteate below the racemes. Bracts tubular, closely — appressed, scarious. Racemes many-flowered, about 2 dm. long. Bracts of the raceme scale-like, much shorter than the pedicels. Pedicels slender, ascending, with the ovary 2.5 cm. long. Flowers — large and showy. Lateral sepals 2.2 cm. long, about 5 mm. wide, narrowly triangular, undulate margined, acute, forming with the elongated foot ofthecolumn a subacute mentum. Upper sepal about 2 cm. long, 7 mm. wide, oblong-lanceolate, acute, 5-nerved, undulate — margined. Petals ligulate, about 2.8 cm. long, 4 mm. wide, undulate margined. Labellum 2.5 cm. long, about 1.5 cm. wide, 3-lobed; lateral lobes rounded in front, rugulose or verrucose on the inner — surface, about 1.5 cm. long; midlobe hardly 1 cm. long, 6 mm. wide, ovate, acute, undulate margined. Disc provided with three thick- ened keels. Column including the foot 1.5 cm. long, free portion 5 mm. long. Santa Cruz Group: Vanikoro, common in rain forest, no: 638a (type in Herb. Ames, no. 37111) and 638b, November 20, 19% (common orchid in the scrub on mangroves). Under no. 638, Kajewski collected a series of specimens which differed from one another in floral characters and appeared to be separable. To these specimens he assigned the numbers 638, 6359 and 638b. Number 638 is referable to Dendrobium macranthum A. Rich. Numbers 638a and 638b are conspecific differing from one another only in the color of the flowers and in slight differences the form of the midlobe of the labellum. In 638a the prevailing color is yellow with purple lines, the labellum being greenish-Y° the labellum is bright yellow-green with purple lines and dots. . Picepiet Kajewskii differs from D. macranthum in the ea sepals and the very different labellum. From the. CM D. conanthum Schitr., it differs chiefly in lacking supe meri carinae on the disc of the labellum and in having the late! verrucose or rugulose on the inner surface. ‘3 Dendrobium macranthum A. Richard, Sert. Astrol. 15; t- 6058 Santa Cruz Grou p: Vanikoro, common in rain- orest, o 638, November 20, 1928 (common orchid both in the scrub “ & mangroves). This species was originally found on Vanikof flowers are somewhat smaller than °° 1933] AMES, FLORA OF THE NEW HEBRIDES 133 Dendrobium (§ Dendrocoryne) macrophyllum A. Richard, Sert. Astrol. 22, t. 9 (1834). Santa Cruz Group: Vanikoro, common in rain-forest at sea level, no. 637, November 20, 1928 (a common orchid; flower yellow with purple spots, very pretty).—Also New Guinea, Java and the Philippine Islands. Dendrobium separatum Ames, sp. nov. Caules elongati, validi, foliosi. Folia disticha, oblongo-lanceo- lata, apice inaequaliter bilobata, acuta, in sicco chartacea. Racemi laterales, plus minusve decemflori. Flores congesti, aurantiaci. Sepala lateralia oblonga, subacuta, mentum elongatum formantia. Sepalum dorsale oblongum, prope apicem angustatum, subacutum vel obtusum. Petala margine minute erosa, elliptico-lanceolata, obtusa vel subacuta, concaviuscula, uninervia. Labellum elonga- tum, simplex, infra medium leviter constrictum, superne in laminam oblongo-ellipticam productum, lamella transversa ornatum; lamella denticulata. Columna generis. Stems up to 1 m. long, stout, leafy, concealed by the sheathing bases of the leaves, yellow after the fall of the leaves, deeply sulcate in dried specimens. Leaves spreading, 1.5-2 dm. long, 1.9-2.5 cm. wide, oblong-lanceolate, tapering to an acute unequally bilobed tip; internodes about 2.5 cm. long. Peduncles lateral on defoliated stems, including the raceme about 2 cm. long, produced at the nodes, rather fleshy, ascending or at right angles with the stem. mes about 1.5 cm. long, about 2 cm. in diameter, bearing ten or more orange flowers. Bracts of the raceme about 3 mm. long, narrowly lanceolate, acute, rigid. Flowers 1.5 cm. long, _ Heshy. Lateral sepals 1.4 cm. long including the mentum, 2.5 mm. wide, oblong, subacute, strongly concave, fleshy. Upper sepal 5 mm. long, oblong, narrowed to a rounded or subacute apex, strongly foneave. Mentum about 9.5 mm. long, forming a blunt spur Which is open in front. Petals 5 mm. long, about 2 mm. wide, eliptic-lanceolate, obtuse, with an erose margin, 1-nerved. Label- hm 13 em. long, slightly constricted below the middle, simple, 3 _™m. wide above the middle, dilated upward forming an oblong- tic smooth lamina with the apical margin slightly inrolled, trav- _ “ted, where it is constricted, by a fleshy retrorse plate which Pe Snag denticulate. Column including the foot 11 mm. long, 4 iv. Sen ta Cruz Group: Vanikoro, in rain-forest on giant | reac at $50 m., no, 503 (type in Herb. Ames 37154), September Re » 1928 (straggling orchid up to 1 m. long, hanging in strings; flowers : ‘range; found only on old stems). 134 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm This species has the general facies of Dendrobium triviale Kraenzl. but differs in the structure of the labellum and in having the transverse plate or lamella on the disc distinctly erose on the upper edge. The transverse lamella which serves as a dividing wall — between the basal and apical parts of the labellum suggests the specific name of the plant. Dendrobium vanikorense Ames, sp. nov. Planta D. acuminatissimo habitu omnino congruens. Caules | pergraciles, elongati. Folia disticha, linearia, acuta. Flores singuli, e nodis superioribus exorientes. Sepala lateralia anguste triangularia, acuta, mentum obtusum formantia. Petala linearia, usque ad apicem attenuata. Labellum anguste ellipticum, margine — denticulato vel breviter fimbriato; callus angustus prope i stat. Columna generis. Stems up to 1 m. long, about 1.5 mm. in diameter, yellowish, — concealed by the tubular sheaths with which the leaves are a — ticulated. Leaves distichous, 8-11 cm. long, about 3 mm. wide, tapering gradually to an acute tip, linear-triangular, about 12 mm. — apart. Inflorescence breaking through the leaf-sheaths opposite — the leaves, 1-flowered, alternating on the upper part of the stem. Peduncle 6 mm. long, subtended by a closely appressed complanait _ bract about 5 mm. long. Lateral sepals 12-15 mm. long, about — 3 mm. wide above the mentum, narrowly triangular, acute, forming — an obtuse mentum 4 mm. long. Upper sepal 14-16 mm. long similar to the laterals, subcaudate at the tip. Petals 12-15 m™ _ long, about 1 mm. wide near the base, linear, tapering mary to the acute caudate tip, 3-nerved. Labellum simple, 11 mm long, 4-5 mm. wide, elliptical with a shortly fringed or denticulaté margin, upper surface thickly studded with simple or sever toothed, complanate processes, and provided near the base WI" o elongated central callus 4 mm. long. Column fleshy, stoul including the foot 6 mm. long. 3 Santa Cruz Grote p: Vanikoro, common in rain-forest a = m., no. 663 (type in Herb. Ames, no. 37191), November : ee up to 1 m. long, hanging down from the rain-10 ¥ “es; Mower creamy yellow tinged with purple); poor red soil ¥ rain-forest at 50 m., no. 65 : erage ondad z “ cif 657, November 25, 1928 @ a0 ey 1G 8rowing on small trees; flowers whitish); a sterile specime! which is vegetatively similar to no. 663. y - general habit this species resembles D. acuminatissin® : : fly in having a simple lip- ye | be sel bl eh te sa (es RNS Vena 1932] AMES, FLORA OF THE NEW HEBRIDES 135 Eria Kajewskii Ames, sp. nov. Radices fibratae, pubescentes. Pseudobulbi cauliformes, clavati, bracteis vaginantibus omnino obtecti. Folia prope apicem pseu- dobulbi conferta, oblonga, utrinque attenuata, apice inaequaliter bilobata. Pedunculi e parte superiore pseudobulbi exorientes, foliis multo breviores, glabri. Bracteae inflorescentiae in sicco brunneae, ellipticae. Pedicellus cum ovario gracilis, elongatus. Sepala lateralia anguste triangularia, acuta, mentum formantia. Sepalum dorsale simile. Petala oblongo-lanceolata, acuta. Label- lum ovatum, margine irregulariter dentatum; discus glaber, prope basim bicallosus. Columna generis. Roots fibrous, elongated, branching, finely pubescent. Pseudo- bulbs stemlike, 14-17 cm. long, clavate, when dry 7 mm. in dia- meter near the base, up to 11 mm. in diameter near the summit, concealed by brownish tubular closely appressed sheaths, the upper portion bearing about 5 more or less crowded leaves. Leaves 11.5- 17 em. long, 1.5-2 cm. wide, oblong, narrowed toward the ends, unequally bilobed at the apex. Peduncles arising from the upper part of the pseudobulb, interspersed among the leaves, about 9 cm. long, floriferous almost to the base. Flowers about 15 in a ose raceme. Floral bracts elliptic, brownish when dry, about -- 5mm. long. Pedicel with the ovary about 1 cm. long, slender. Lateral sepals 11 mm. long, about 3 mm. wide at the base, narrowly _ tiangular, acute, forming with the foot of the column a short - obtuse mentum, 3-nerved; mentum 2 mm. long. Upper sepal ‘ap $e eS eS 11 mm. long, about 2 mm. wide, oblong, acute, 3-nerved. Petals +10 mm. long, 2.5 mm. wide near the base, oblong-lanceolate, acute, 3-nerved, the outer nerves branched. Labellum 5 mm. long, about 2.5 mm. wide, rather fleshy, smooth, ovate, acute or Subacute, margin with several blunt teeth on either side above the middle, 3-nerved with the outer nerves branching, bicallose near the base; calli abbreviated, one on each outer nerve. Column meluding the foot 5 mm. long. ; | Reityum: Anelgauhat Bay, common in rain-forest at - = m., no. 820 (type in Herb. Ames, no. 37850) (small orchid Ze Srowing on trees; flowers inconspicuous). : Ae Fria Kajewskii is closely allied to E. retroflera Lindl. which it _ ‘Tesembles closely in habit, but it is clearly distinguishable from lah, Species and from all close allies in having the margin of the bellum bluntly dentate above the middle. En vanikorensis Ames, sp. nov. . ae ss _, Caules elongati, erecti, vaginis foliorum omnino ees Folia £4 oblonga, utrinque attenuata, quam racemus ongiora, Sa aa nh 136 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm dense fusco-villosa, in petiolum vaginantem contracta. Racemi pauciflori, rubro-villosi. Bracteae inflorescentiae dense rufo- — villosae, anguste ovatae, rigidae. Flores tomento rufo vestiti. éralia triangulari-lanceolata, valde incrassata, intus et extus dense rufo-villosa, mentum obtusum formantia. Sepalum dorsale triangulari-lanceolatum. Petala oblonga, obtusa, extus dense rufo-villosa, leviter concava. Labellum prope apicem con- — spicue constrictum lobum transverse ellipticum vel reniformem _ formans, per lobum medium callo prominenti ornatum, infeme | usque ad basim attenuatum; discus intus villosus, per medium 4-nervius. Columna generis. Stems up to 1 m. long, concealed by the cylindrical foliar sheaths — with which the leaves are articulated; sheaths 1.5-3 em. long, — densely covered with brownish, more or less matted hairs. Leaves distichous, up to 12.5 em. long, 1.7—2.5 cm. wide, oblong-lanceolate, _ coriaceous, densely pubescent above, covered beneath with red- dish appressed hairs. Racemes 5 or more, produced on the upper part of the stem, situated opposite the leaves, about 3.5 cm. long with a cupuliform sheath at the base. Rachis of the raceme reddish villose. Bracts subtending the pedicels about 1 em. long, rigid, covered with reddish hairs. Ovary sessile, up to 2 cm. long, covered with reddish hairs. Flowers about 1 cm. long. erat sepals 1 em. long, about 5 mm. wide near the base, triangular lanceolate, acute, covered inside and out with bristle-like reddish : hairs, fleshy, rigid, forming a prominent obtuse mentum. Upper sepal about 1 em. long, similar to but narrower than the laterals : Petals 8 mm. long, 1.5 mm. wide, linear-oblong, rounded at the tp, 4 sparsely pubescent inside, conspicuously hairy on the outside, i mid nerve prominent. Labellum about 1 cm. long, 3 mm. wile constricted 2 mm. below the apex forming a terminal reniform® transversely elliptical lobe; lamina oblong, contracted toward the base, ornamented inside and out with long closely appress . or slightly spreading reddish hairs; disc 4-nerved, ecarinate, with vs bere: fleshy “us extending along the middle of the term? : - Column characteristic of the genus. Bebe ey m., no. 512 (type in Herb. Ames no. 37110), September be Allied to E. iodantha Schitr. from which it differs in the struc" of the lip. = wae ef vanikorense Ames, sp. nov. | waka epiphytica. Radices dense pubescentes. Bb o > nn arete vaginatum, vaginis brunneis °P 1932] AMES, FLORA OF THE NEW HEBRIDES 137 Pseudobulbi obpyriformes, in sicco valde sulcati, monophylli, statu juvenili vaginis vestiti. Folium variabile, 3.5-7.5 em. longum, 1-1.7 cm. latum, valde coriaceum, oblongum vel oblanceolato- oblongum, apice tridentatum. Pedunculi uniflori ut videtur, singuli vel bini, folium aequantes vel breviores, plus minusve 4 cm. longi, paucibracteati, bracteis acuminatis. Flos 1 em. longus, aurantiacus. Sepala cupulam formantia, ovata usque ad medium cohaerentia. Petala lineari-lanceolata, trinervia. Labellum prope medium valde calcarato-saccatum, ovato-acuminatum. Columna valida, labello brevior. Rhizome rather stout, elongated, when young concealed by brownish tubular bracts, those subtending the pseudobulbs broadly ovate, chartaceous. Roots fibrous, elongated, branching, densely pubescent. Pseudobulbs 1—1.5 cm. long, obpyriform, longitudinally suleate, monophyllous. Leaf extraordinarily variable, coriaceous, 3.7-7.5 cm. long, 1-1.7 cm. wide, oblong or oblong-oblanceolate, tapering gradually to a blunt bilobed apex, the midnerve extended into an apicule that almost equals the lobes in length. Peduncles or geminate, shorter than or equalling the leaf, including the pedicellate ovary 2.5-4 cm. long, one-flowered, slender, only one peduncle floriferous at a time. Pedicel and ovary 1-1.5 em. long, subtended by a scale-like acuminate bract. Flower urceolate, Tem. long, orange with the tip of the sepals yellow. Sepals similar, ovate-lanceolate, about 4 mm. wide at the point of union, free Portion about 4 mm. long, united for more than half their length, lateral ones forming a protuberant blunt mentum. Petals free, about 9.5 mm. long, hardly 2 mm. wide, linear-lanceolate, acute, Snerved. Labellum about 9 mm. long, parallel with the column, erect, with a rounded blunt sac near the middle; lamina ovate, cuminate-acute from a broad base. Column 6 mm. long. Santa Cruz Group: Vanikoro Island, common on large trees at 50 m., no. 641 (type in Herb. Ames no. 37194), November 22, 1928. Also from the New Hebrides, Eromanga, Peak south of Bay, Dr. R. Morrison, August 5, 1896 (Herb. Kew). N.E. Traitor’s Head, summit of old crater growing on rotten trunk, 2,400 feet alt., L. Cheeseman, no. 71 (Herb. Kew). Aneityum, Peak of Ithumu, Dr. Morrison, June 30, 1896 (Herb. Kew). The genus Mediocalcar has not been reported heretofore from the New Hebrides or from the Santa Cruz Group. M. epigeytet way ally of M. bulbophylloides J. J. Sm. differing from it * ™ ’ Separated pseudobulbs, larger leaves and in the yeeyee | the flowers. Mediocalcar is essentially New Guinean with paid _ RW outlying species. M. ponapense Schitr., from the Caroline 138 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Islands, one of the few species reported from beyond New Guinea, differs from M. vanikorense in being much more robust, with — larger pseudobulbs and dissimilar leaves. | Ceratostylis kaniensis Schlechter in Fedde Rep. Spec. Nov. Bei- : hefte 1. 245 (1912). ; Santa Cruz Group: Vanikoro, common in rain-forest at 150 m., no. 575, November 6, 1928 (a plant found on a large Kaun tree).—Also New Guinea. Ceratostylis subulata Blume, Bijdr. 306 (1825). . Efate: Undine Bay, common in rain-forest at 500 m., no. 234 April 28, 1928 (growing on bark of tree).—Also Java, Sumatra, Malay Peninsula, Assam. \ The material studied lacks flowers and leaves. In general aspect 7 the plant resembles (’. kaniensis Schltr. eed asta Brousmichei Kraenzlin in Lecomte, Notulae Syst. Iv. 136 Aneityum: Anelgauhat Bay, common in rain-forest at 1000 ; feet, no. 844, March 2, 1929 (orchid growing on rain-forest trees; flowers white) —Also New Caledonia. ey Kajewski’s specimen differs from the type in being taller with | much longer leaves. A flowerless plant from Efate, Undine Bay . 7 ca ae ay: ‘se ay. collected by Dr. Morrison in August 1896 appears to belong here (Herb. Kew). : Appendicula reflexa Blume, Bijdr. $01 (1825). Oe _ Aneityum:Anelgauhat Bay, common in poor red soil countt) at 500 feet, no. 915, March 17, 1929 (ground orchid 2 ft. highs flower white and green). Eromanga: Dillon Bay, common” poor red soil, bracken country, no. 326, May 29, 1928 (orchid go" mg on ground in shade of bracken and stunted brush; flower white). Pay Sumatra, Borneo, Celebes, and the Malay Peninsul. 4g ernacular name “ Orvum-nge-nompull” (under 326). iP vanikorensis Ames, sp. nov. didi gracilis. Caules vaginis foliorum omnino obtect back om: lanceolato-linearia, acuminata, apice rotundato- eget: Pa Pepe lectentia, Inflorescentiae omnino laterales. -F. ‘ o re _— aria, trinervia, mentum obtusum formant ea a. — segeete: ellipticum. Petala oblongo- : age” : Stems slender, 1.5 dm. long in : be: - Jong in the fragments that were probably attaining a height of 25 or 30 em.,entirely conceal eae | ee * 1932] AMES, FLORA OF THE NEW HEBRIDES 139 sheathing tubular bases of the leaves, flexuose. Leaves distichous, 3.5-5.5 em. long, 8-15 mm. wide, about 1 cm. apart, narrowly lanceolate, bluntly bilobed at the tip with the midnerve projecting between the lobes, obliquely ascending. Inflorescences opposite the leaves, breaking through the tubular sheaths with which the leaves are articulated, about 1 cm. long, several-flowered. Bracts of the inflorescence about 3 mm. long, rigid, lanceolate. Lateral sepals about 3 mm. long including the blunt mentum, about 2 mm. wide at the base, strongly concave, broadly triangular, ending in a much thickened tip, 3-nerved. Upper sepal about 2.5 mm. long, approximately 1 mm. wide, narrowly elliptical, obtuse, 3-nerved, thickened at the apex, strongly concave. Petals 2 mm. long, 1 mm. wide, oblong-lanceolate, rounded at the apex, usually I-nerved. La- bellum 3 mm. long, 2 mm. wide, simple, lightly saccate at the broad base, in front of the sac expanded into a more or less elliptical obtuse lamina with an apical callus; sac with a large callus in the center; callus transversely elliptical with the ends continued into pe keels. Column including the foot 2 mm. long, rostellum ifid. Santa Cruz Group: Vanikoro, in rain-forest at 300 m., no. 594 (type in Herb. Ames, no. 37193), November 11, 1928 (orchid growing on rocks in a gully with water running around them; flow- ets greenish white). ae In general aspect this plant resembles A. anceps BI. but is dis- similar from it in the structure of the lip. A. reflera BI. is also a close ally but different from A. vanikorensis chiefly in the shape of the petals. Calanthe Vaupeliana Kraenzlin in Notizbl. Bot. Gart. Berlin, ‘Santa Cruz Group: Vanikoro, common in rain-forest i 800 m., no. 599, November 11, 1928 (a ground orchid a meter high, ound growing at high altitudes; flower beautiful, white, length of Three specimens from the New Hebrides, collected on Aneityu ” _ in June 1896 by Dr. Morrison, are preserved in the Kew Herbarium. Of th i : uhat is described as havin. ese specimens from Anelga piecagrek ‘ucturally the plants collected by Kajewski x a : by Morrison on Aneityum are similar and agree 17 a Jum is acters with C. Vaupeliana. The apical lobe of the la #1 Lit is Variable. In the type it is oblong; in Morrison’s specimen’ © ‘“tongly dilated from the base. Kajewski’s specimen, with regard 140 JOURNAL OF THE ARNOLD ARBORETUM vonxm | to the lip, is almost intermediate between Vaupel’s Samoan Plants | and Morrison’s plants from Aneityum. 7 pathoglottis Petri Reichenbach f. in Gard. Chron. n. ser. vill | one pened | Eromanga: Dillon Bay, common in red soil, bracken country 1 at 300 m., no. 306, May 24, 1928 (ground orchid 1.5 m. high with pretty pale puce flowers).—Vernacular name “ Lar-ver-ae.” In the Kew Herbarium there are three specimens from the ee Hebrides, that are referable to S. Petri. Two of these were collected by Dr. R. Morrison on Aneityum near. Anelgauhat in 1896. The third one was collected i in March 1930 on Eromanga by L. Cheese | man (no. 14). Spathoglottis Vieillardii Reichenbach f. in Linnaea, x11. 85 (1877). Santa Cruz Group: Vanikoro, common in rain-forest at sea level, no. 563, November 6, 1928 (a ground orchid common throughout the New Hebrides; color of the flowers varies from white) to purple; diameter 4 cm.; inside of flower yellow) —Already ported from Tanna.—Also ew Caledonia. Bulbophyllum sp. Santa Cruz Group: Vanikoro, common in rain forest 150 m., no. 680, December 3, 1928 (an orchid growing 02 ee close to the sea; flowers dark red, inconspicuous petals; fruit 5 a | long, 2.4 cm. in diameter). | This species is in advanced fruit. It is probably referable ¥ B. praealtum Kraenzl. from Samoa, but the flowers are too oll to be serviceable in attempts at identification. In gen ‘4 the plant resembles closely Macrolepis longiscapa A- Rich oe Pgs be identical with it. Macrolepis longiscapa was ound on Vanikoro. The genus Macrolepis is referable to ‘Bulle phyllum, but Richard’s species has never been transferred ° pagent. B. longiscapum Rolfe from Fiji would * m t the description to belong in the same alliance with MC pebna te Lip ic Cruz Group: Mesikoro , common in rain-f fe m., no. 586, November 8, 1928 (leebing sens hae up* that are more or less lean spots). “Ales New ¢ 20964 hemes yellowish green 1982] AMES, FLORA OF THE NEW HEBRIDES 141 Vandopsis Quaifei (Rolfe) Schltr. frown Espiritu Santo is a closely allied species and may prove to be conspecific with V. Warocqueana. In the Kew Herbarium there is a specimen from Aneityum that is surely referable to V. Warocqueana, but it is in a fruiting condition and hardly satisfactory for critical study. This specimen was collected by Dr. R. Morrison on June 20, 1896 on the coast east of Anelgauhat. Saccolabium Kajewskii Ames, sp. nov. Caules vaginis foliorum obtecti. Folia ligulata, coriacea, utrin- que attenuata, apice inaequaliter et obtuse bilobulata, basi in petiolum brevem conduplicatum contracta. Pedunculus elongatus, infra racemum paucibracteatus, _bracteis infundibuliformibus. Racemus cylindraceus vel cylindraceo-secundus. Sepala anguste ovata vel elliptica. Petala elliptica, acuta vel subobtusa, trinervia. Labellum longe calcaratum; lamina infra apicem suborbicularis, antice in apicem complanatum contracta. Stem about 5 cm. or more long, rigid, completely concealed by the sheaths to which the leaves are articulated. Leaves dis- tichous, 14-21 cm. long, up to 2 em. wide, oblong, ligulate, un- equally and bluntly bilobed at the apex. Peduncle with the raceme up to 19 em. long, with about 3 infundibuliform bracts. Raceme up to9 cm. long, densely many-flowered. Bracts of the inflorescence seale-like, very much shorter than the pedicels, 1.5 mm. long. Pedicel with the ovary 9 mm. long. Flowers including the spur of the lip 1.3 em. long. Lateral sepals about 5 mm. long, 2 mm. ven ovate or elliptical, obtuse, concave. Upper sepal similar. Pe about 5 mm. long, 2 mm. wide, elliptical, acute, 3-nerved. Labellum including the spur 1 cm. long, in front dilated into an orbicular m: ina which terminates in a complanate tip; lamina with an intramarginal keel; spur 7 mm. long, inflated below the middle. Column 1.5 mm. long. _ Efate: Undine Bay, common in rain-forest along sea shore, no. 205 (type in Herb. Ames no. 36375), April 4, 1928 (color of flowers — . ranging from creamy white to white with green, some purpi> others pink). Banks Group: Vanua Lava, common along se sali ‘no. 448, July 9, 1928 (found growing close to the sea on frees WN” trunks are well shaded). _Two specimens referable to this species are in the Kew Herba- - pov in August 1896 ae. : Sate, UndineBay, 12 A 1] They were collected on Efate differs from the type by Dr. R. Morrison. One of these specimens oS in having broader leaves. _ RNOLD ARBORETUM, Gio Sokesinke nena 5 BE A AS ire at) Ras = AT + 142 JOURNAL OF THE ARNOLD ARBORETUM [vou xm A NEW GENUS OF THE ORCHIDACEAE FROM THE NEW HEBRIDES Oakes AMES Amone the orchids collected in the New Hebrides by Dr. R Morrison in 1896 and by L. Cheeseman in 1930, there is a strikingly distinct species from Eromanga Island. I have been unable to refer it to any genus heretofore described. In general facies it suggests some of the species of the African genus Acrolophia, but is clearly differentiated by the very dissimilar means by which the pollen masses are attached to the viscid disc of the rostellum, by the absence of a well developed spur at the base of the labellua and by the aspect of the inflorescence. - For the opportunity of studying the material on which the fob lowing description is based I am indebted to the Director of # Royal Botanic Gardens, Kew. an an ilus Ames, gen. nov. a Divbic: Acrotonae; Tribus: Kerosphaerinae; Subtribus: Poly pone ane Sepala lateralia patentia, basi columnae pedi | a liberum. Petala sepalis similia. Labellum pias nba nines inarticulatum, basi contracta incumbens, erectum, trilobatum, in saccum vel calcar abbreviatum _ minusve globosum productum, antice in laminam plus mil beer. roductum, prope basim laminae bicallosum; lobi later#l pee YG labelli conspicue papillosus vel pilosus. Columm ae in pedem brevem producta. Anthera terminalis, glandula tates globosa, stipitibus duobus distinctis affixd, : Vernatio ze aes 1; terrestris. Folia equitantia, numet’ Acrolophiarum, ue leativa. Species una adhuc noe cb i=} Ames, sp. nov. Caules abbreviati, basibus f oi #9 # ees plus minusve equitantia, oblique trilobatum, toto ise _ oblongo-lanceola tary hasi medium inaequaliter ares | et tum; lobi lat 1982] AMES, A NEW GENUS OF THE ORCHIDACEAE 143 : obtusi; lobus medius anguste ovatus vel obovatus, obtusus, prope basim bicallosus callis papillosis, prope apicem crista pilosa ornatus. Columna elongata, incrassata, in pedem brevem producta. Pol- linia duo, globosa. Roots very stout, when dry about 5 mm. in diameter. Stem much abbreviated, about 1.5 cm. long, concealed by the equitant leaves. Leaves 1.5—4 cm. long, + 8 mm. wide, duplicative, per- sistent, narrowly triangular, acute, about 6, ascending with the apices curving inward toward the base of the elongated peduncle, conspicuously many-nerved when dry. Peduncle wand-like, slender, about 2 mm. in diameter when dry, about 5 dm. long, provided with about 5 tubular acute bracts below the raceme. Bracts 1-1.5 cm. long. Raceme about 10-flowered, 4-10.5 cm. _ long, about 4 cm. in diameter. Flowers about 5 mm. apart, white _ transfused with mauve. Bracts of the raceme 4-5 mm. long, narrowly lanceolate, acute, spreading, much shorter than the _ Pedicels, brownish when dry. Pedicels about 1 cm. long, very _ slender, obliquely ascending. Ovary about 3 mm. long, arcuate. lateral sepals 12-15 mm. long, 2.5 mm. wide, oblong-lanceolate, or subacute, membranaceous, forming a blunt mentum, d-nerved with the outer nerves less conspicuous than the three middle ones. Upper sepal similar, about 12 mm. long. Petals about 12 mm. long, 3-3.5 mm. wide, similar to the sepals. Label- 20. 37776) ; tableland, Morrison, s. n., July 16 & 22, 1986; ae plateau, Morrison, s. n., July 16, 1896; without exact location, Morrison, s. n., July 17, 1896; tableland of voleanic ash, Pak. Tass, and also among bracken, and at edge of belts bs : ude 500-700 feet, August 4, 1930, L. Cheeseman, no- 59. CO sO, See a tee Fe 144 JOURNAL OF THE ARNOLD ARBORETUM [vou. xi - It is clear that this species belongs to the subtribe Polystachyeae. Vegetatively it resembles Acrolophia tristis Schltr. & Bolus and A. lamellata Schitr. & Bolus, but is separable from them in the structure of the pollinarium, in the much shorter leaves, and through ~ the absence of a conspicuous spur at the base of the labellum. In the specimens studied the leaves are closely appressed to the base of the slender peduncle and are nearly parallel with it. Whether — or not the leaves elongate as the plant matures is a question that © cannot now be answered. ARNOLD ARBORETUM, Harvarp UNIVERSITY. 2 Sayer ann 1932] KOBUSKI, THE CHINESE SPECIES OF JASMINUM 145 SYNOPSIS OF THE CHINESE SPECIES OF JASMINUM CLARENCE E, Kosuskti JASMINUM is essentially a tropical or subtropical genus and is found in eastern and southern Asia, Malaysia, Africa, Australia _and even tropical America. There are probably over three hundred species recognized in the genus at the present time. In the her- barium of the Arnold Arboretum are represented species from all these regions. However, the region best represented is China. Of the few species not found in this herbarium, photographs of the type specimens were taken by Alfred Rehder on a recent visit to European botanical institutions from which he brought back over 500 photographs of types of Chinese ligneous plants. This gave us a nearly complete representation of the Chinese species of Jasminum and made possible the present study. Since the time of Linnaeus there has been made only one attempt toward a complete representation of the genus; this was published in 1844 by A. P. De Candolle (Prodr. vir. 300-316). Of the Chinese species three enumerations were published, namely by Hemsley, Léveillé and Chung. W. B. Hemsley (Enumeration of all the Plants known from China proper, Formosa, Hainan, the Corea, the Luchu Archipel- ago and the Island of Hongkong) in Jour. Linn. Soc. xxvi. 78-82 (1889) recognized fifteen species. H. Léveillé in handling his “Jasmina Sinensia” in Fedde, Rep. Nov. Spec. xu. 149-150 (1914) mentions thirty-three species in his key while H. H. Chung Publishing a “Catalogue of Trees and Shrubs of China” in Mem. Sei. Soe. China, 1. no. 1 (1924) lists 27 species and two varieties. __ Asalready mentioned in this paper there are over three hundred Species. Early in the study we thought that most of these species __ Were distributed in southeastern Asia and Malaysia and that here y Probably was the main center of distribution of the genus. How- _ €ver, in careful listing of all species of the genus as recog by Index Kewensis” we discovered that while there were approxl- _ _Mately one hundred species of Jasminum found in China (especially _ South-west) and the Himalayan region of India there were about _ “le same number of species found in Malaysia and about one a hundred twenty species in Africa. Besides these numbers there - 48a sprinkling of a few species in Australia, Polynesia and tropical Des ica. All together this makes the genus pan-tropical with : - Possible center of distribution found in the area including eT the ne as and China. : eet or) | 146 JOURNAL OF THE ARNOLD ARBORETUM [voL. XUI These numbers have been taken from species recognized by the Index Kewensis. Incidentally, of the fifty-six species belonging within the scope of this study approximately one-half turned out tobe synonyms. No new species was found among the unidentified — material either at the Arnold Arboretum or the material borrowed — from other institutions. This may intimate that since Jas- minum is composed mostly of such interesting and decoratively beautiful shrubs that the majority of species were found early by plant collectors and described at that time. It is evident that later botanists encountering some of these already described new species for the first time, lacking both material with which to compare them and literature, also have described several of the outstandingly distinct species a second and even a third time. At present, when the material (types or photographs of types) was brought together for the first time it was positively amazing to find the species duplications which had taken place. One cannot be positive but it seems quite logical to assume that the same situation may be found in a study of Jasminum from any other distinct locality. This would decrease the number of recognized species considerably. The genus is botanically separated quite easily into four distinct — sections; the sections being based on leaf characters. These sections were all introduced by DeCandolle in his Prodromus. — The Alternifolia group with alternately arranged leaves is perhaps the most outstandingly different. It contains five species and varieties. The opposite-leaved sections are three, namely: Unt foliolata (18 species); Trifoliolata (9 species and varieties); Pin- natifolia (5 species and varieties). The delimiting characters of these groups can easily be ascertained from their significant names. In addition to the material in the herbarium of the Arnold Arboretum, photographs of types were secured and loans of material obtained from the herbaria of the New York Botanic Garden, the Philippine Bureau of Science, Gray Herbarium of Harvard Uni versity, Royal Botanic Garden at Edinburgh, British Museu of Natural History and Royal Botanic Gardens at Kew. Opportunity taken at this time by the author to express his gratitude to all members of these institutions mentioned above who aided ™ photographic work or selection of the specimens loaned. "thanks are due Mr. Alfred Rehder, Curator of the Herbarium, f° the suggestion of the problem and for the ever kind and helpful interest which has been shown me at all times during the progr of this work. The abbreviations of the herbaria used in this paper are 35 follows: - a4 1932] KOBUSKI, THE CHINESE SPECIES OF JASMINUM 147 (AA) = Arnold Arboretum of Harvard University (B) = British Museum of Natural History (Br) = University of Breslau (C) = University of California (E) = Edinburgh, Royal Botanic Garden (G) = Gray Herbarium of Harvard University (K) = Kew, Royal Botanic Gardens (NY) = New York Botanical Garden (P) = Philippine Bureau of Science Key To THE SERIES A. Leaves alternately arranged.............-....-++- 1. ALTERNIFOLIA AA. Leaves opposite in arrangement. Leaves com Si Aaivee trifdliolate ....: L.-ciile-s cs tee 2. TRIFOLIOLATA CC. Leaves five-foliolate or more................++- 3. reenter rls bg eke eee 4, UNIFOLIOLATA image 1. ALTERNIFOLIA DC. eee ale ee OH ele we we Ee a 66 6 a Oe ele lee ee ee ©. Sal tower — ate cm. across. yx lobes and pedicels glabrous. Be Pin ites 3a. J. heterophyllum var. glabricorymbosum alyx lobes and pedicels alloc. ish ophyllum var. subhumile J. heter BB. Leaves only ternate or pinnate; leaflets 1.5-3.5 cm. long; inflorescence 3-8 flowered 4. J. humile . 28S 8 6 8 Oe eb. oe SNS Ce Ee oe 6 Orem @ ORES FS Om 1. Jasminum floridum Bunge in Mém. Div. Sav. Acad. Sci. bet Pétersb. 1. 116 (Enum. Pl. China Bor. 42) (1833). —DeCandolle, . Vill. 318 (1844).—Nicholson, Ill. Dict. Gard. 1. 207 (1877).— uel in Ann. Mus. Ludg.-Bat. 1. 263 (1866); Prol. Fl. Jap. 151, 859 (1866-67) ._—Franchet & Savatier, Enum. Pl. Jap. 1. 314 (1875). 2 : 7 tlooker £. in Bot. Mag. crx. t. 6719 (1883).—Hemsley in Jour. . Soc. xxv. 78 (1889).—Diels in Bot. Jahrb. xxxx. 534 (1900). , Ill. Handb. Laubholzk. 1. 839, figs. 527h-i, 528a Ee Bibacids (911) ecco in Fedde, Rep. Spec. Nov. xm. 149 (1914) a der in Sargent, Pl. Wilson. mu. 614 (1916); Man. Cult. Tree ‘ pee hg i aeadl .—Chung in Mem. Sci. Soc. China, I. 70. 1 | Argyt Level i in Mem. Acad. Ci. Art. Barcelona, x1. n0- a 7 2, 17 (1916) .—Syn. nov, 148 JOURNAL OF THE ARNOLD ARBORETUM [vou. Xm Chihli: Peking, C. A. Skatchkov (1848-1857) (G). Kiang- su: Le kien, Ch. d’Argy, May (1846-66) (AA, type of J. Argyt); Nanking, L. F. Tsu, no. 629, May 19, 1921 (C). Western Hupeh: north and south of Ichang, alt. 300—700 m., E. H. Wilson, no. 789 in part, June and December 1907 (bush 1 m. tall; flowers yellow; fruit black) (AA, G); open, grassy slopes, Ichang, W. Y. Chun, no. 3468, July 20, 1922 (bush 1 m. high; flowers yellow) (AA); thickets, Hsing-shan hsien, alt. 600 m., E. H. Wilson, no. 789 in part, June 1907 (AA); moist shady cliff, Gian Gia-kou, W. Y. Chun (Huang Tsung), no. 3572, July 27, 1922 (shrub 1-2 m.; leaves dull green above; flowers yellow) (AA); without precise locality, E. H. Wilson (Veitch Exped. no. 83) April 1900 (AA, NY); without precise locality, A. Henry, nos. 2700 (NY, G) and 6288 (AA, NY). Kansu: near Kua tsa, F. N. Meyer, no. 181%, Nov. 5, 1914 (C). Western Szechuan: Nanch’uan, Fu pei tsui, C. Bock and A. von Rosthorn, no. 1151, Oct. 1891 (photo. and fragment of holotype of J. floridum var. spinescens) (AA); Nanch’uan, Taho-kou, (. Bock and A. von Rosthorn, no. 163, July 1891 (AA); side of river, Nanchuan hsien, alt. 1525-1830 m., W. P. Fang, no. 788, May 15, 1928 (shrub 1 m.; flowers yellow) (AA, NY); side of brook, Sungpan hsien, W. P. Fang, no. 4239, August 17, 1928 (shrub 2-3 m.; flowers yellow) (AA, NY). In comparing J. Argyi Léveillé with material of J. floridum Bunge, no differences warranting specific delimitation could be found, nor does J. floridum var. spinescens Diels seem to be different enough | from typical J. floridum to be maintained as a distinct variety. 2. Jasminum Giraldii Diels in Bot. Jahrb. xxrx. 534 (1901).— Pampanini in Nuoy. Giorn. Bot. Ital. n. ser. xvu1. 689 (1910).— Léveillé in Fedde, Rep. Spec. Nov. xm. 149 (1914).—Bailey, Standard Cycl. Hort. 1719 (1915).—Rehder in Sargent, Pl. Wilsoa- ut. 614 (1916); Man. Cult. Trees Shrubs, 765 (1927).—Chung @ Mem. Sci. Soc. China, 1. no. 1, 216 (1924). ay asminum tsinli: Lingelsheim in Fedde, Rep. Spec. Nov. Belh- at foot of Lao y huo, G. Giraldi, no. 1519 (photograph of syntype 7) Herb. Mus. Berlin) (AA); same locality, June 4, 1897, G. Giraat (AA); northwest of Hangcheng hsien, W. Purdom, no. 368, date lacking (1 m. high; flowers yellow) (AA); Tai-pei-shan, W. Purdom nos. 907 and 908, in 1910 (AA); at foot of mts. at Schian on south Hsian fu, Tsin ling schan, alt. 500 m., W. Limprichb, Ee ee lee ee 1932] KOBUSKI, THE CHINESE SPECIES OF JASMINUM 149 2689 (photograph and fragment of type of J. tsinglingense from Br. in AA). Hupeh : dry rocky places, Fang hsien, alt. 1000 m., E. H. Wilson, no. 598, May 23, and Nov. 1907 (abundant; 1-2 m.; flowers deep yellow; fruit black) (AA, G, NYC), Lingelsheim’s species Jasminum tsinglingense is undoubtedly synonymous with J. Giraldii Diels. dines i 3a. Jasminum heterophyllum Roxb. var. glabricorymbosum W. W. Smith in Notes Roy. Bot. Gard. Edinburgh, x11. 209 (1920). Yunnan :on ledges of limestone cliffs, on Langhong—Hoching divide, lat. 26° 16’ N., alt. 2440 m., G. Forrest, no. 9990, May 1913 (shrub 2-3 m.; flowers deep golden yellow, fragrant) (AA); in open situations by streams, Yungpe mountains, lat. 26° 45’ N., alt. 2740 m., G. Forrest, nos. 11037 (NY), 11177 (C), and 11472, Sept.—Oct. 1913 (erect shrub 2-2.5 m.; fruit black) (AA); Mengtze Woods, alt. 1475-1525 m., A. Henry, nos. 9107, 9107a and 91078 (slender shrubs 2-2.5 m.; flowers yellow; fruit black) (AA, NY); forests, Ta Song pin, Simeon Tén, no. 99, March 26, 1916 (tree 3 m.; flowers yellow) (AA); eastern slopes of Likiang Snow Range, Yangtze watershed, Prefectural District of Likiang, J. F. Rock, hos. 3879, 3955, May-Oct. 1922 (shrub 2-2.5 m.; flowers orange- yellow) (AA, N Y); mountains south of Likiang Sungkwe Hochin Range, J. F. Rock, no. 8292 in 1928 (shrub 1.5-2 m.; flowers yellow) (AA); western slope of Likiang Snow Range, Yangtze watershed, - F. Rock, no. 8589 April 1923 (shrub $ m.; drooping yellow flowers) (AA); banks of Djiper Ka Ku stream, near Ngaza, drainage basin of the Yangtze, west of Likiang, J. F. Rock, no. 10602, in 1923 (shrub 1 m.) (AA); Shweli River drainage basin and environs of Tengyueh, J. F. Rock, no. 8033, February 1923 (shrub 1,2-2 m.; flowers citron-yellow) (AA); headwaters of the Red River or from Maokai to Tatsang, alt. 1675 m. J. F. Rock, no. 4 3026, April 8, 1922 (flowers yellow) (AA). wee | _“" 8b. Jasminum heterophyllum Roxb. var. subhumile (W. W. 150 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm its densely villous pubescence found on the calyx and pedicels, The pubescence on J. heterophyllum is of a puberulent nature. It is interesting to find these two extreme variations, namely the present villous variety along with the glabrous variety mentioned — before both growing in China, while the typical species has been found only in the Himalayan region. 4. Jasminum humile Linnaeus, Spec. Pl. 1. 7 (1753).—Aiton, — Hort. Kew. 1. 9 (1789).—Ker in Bot. Reg. v. t. 350 (1819).—St. Hilaire, Pl. France, 1m. t. 110 (1808); Traité Arbriss. & Arbustes, 1. t. 90 (1825).—DeCandolle, Prodr. vr. 313 (1844).—Clarke in Hooker f., Fl. Brit. Ind. m1. 602 (1822).—Nicholson, Illustr. Diet. Gard. 1. 207 (1887).—Dippel, Handb. Laubholzk. 1. 146, fig 91 (1889).—Bailey, Cyclop. Amer. Hort. 1. 843, fig. 1191 (1900); Stand. Cyclop. Hort. m1. 1719, fig. 209 (1915).—Léveillé in Fedde, Rep. Spec. Nov. xm. 149 (1914).—Brandis, Ind. Trees, 452 (1906). _ —Schneider, Ill. Handb. Laubholzk. m. 840 (1911).—Rehder in Sargent, Pl. Wilson. 1. 615 (1916); Man. Cult. Trees Shrubs, 765 (1927).—Chung in Mem. Sci. Soc. China, 1. no. 1, 216 (1924).— Boynton in Addisonia, x11. 55, pl. 412 (1927). Jasminum chrysanthemum Roxburgh, Cat. Hort. Beng. 3 (1814), nomel; FI. Ind. 1. 98 (1820). Jasminum revolutum Sims in Bot. Mag. in Bot. Reg. m1. t. 178 (1817); v1. notes 2 (1820). Cab. x. t. 966 (1824).— De Candolle waceum Wallich, Cat. No. 2888 (1829), nomen nugn 7 Gear Gen. Syst. rv. 63 (1838).—DeCandolle, Prodr. vil. 2 | 2 en ta aes Leeilé var. siderophyllum Léveillé, Cat. Pl. 1 | : Kansu : Lower Tebbu country : slopes along 60% of Chulungapu, alt. 1980 m., J. F. Rock, no. 15035, Sept-O% 1926 (shrub 1-1.7 m.; flowers yellow; fruit black) (AA); ee’ slopes with Oaks in Nyibaku, alt. 1900 m., J. F. Rock, no- ye Sept. 9, 1996 (shrub 1-15 m.; flowers rich yellow) (AA); ou” . of forests along stream in Wantsang valley, alt. 2135-2200 B | J. F. Rock, no. 14658, Aug. $1, 1996 (shrub 1-1.5 m.) (AA). We?) ern Szechuan: descent of Hsao-chin-lo, Monkong © 1932] KOBUSKI, THE CHINESE SPECIES OF JASMINUM 151 alt. 2300-3000 m., E. H. Wilson, no. 2809, June 1908 (bush 1-2 m, tall; flowers yellow) (AA); dry arid places, Maochou, alt. 1800- 1600 m., E. Z. Wilson, no. 2811, May 24, 1908 (bush 0.3-2 m. tall; flowers yellow) (AA); in dumetis mont., inter Kalapo et Linku, alt. 3000 m., C. K. Schneider, no. 1302, May 17, 1914 (AA); inter Huale et Mo lien, alt. 2600 m., C. K. Schneider, no. 4091, May 25, 1914 (AA); Teng-hsiang-ying, alt. 2100 m., Harry Smith, no. 1890, May 20, 1992 (in prato fruticoso-herboso) (AA); inter Wen-ch’uan-hsien et Hsin-p’u-kuan, alt. 1300 m., Harry Smith, no. 2463, June 28, 1992 (flor. sulphurei) (AA); without exact locality, alt. 3000-3300 m., E. H. Wilson (Veitch Exped. no. 4077) June 1904; alt. 2300 m., E. H. Wilson (Veitch Exped. no. 4078) August 1903 (AA). Yunnan: high plateau between Talifu and Likiang to the foot of the Likiang Snow Range, J. F. Rock, no. 3293, May 6-11, 1999 (shrub; flowers yellow) (AA); 1917 (shrub 1-2 m.; flowers deep golden yellow) (AA); vallons de Suen-oui, alt. 2400 m., E. E. Maire (AA, no. 451), July (arbrisseau Pine forests on the eastern flank of the Tali Range, lat. 25° 40’ N., - alt. 2440-3050 m., G. Forrest, no. 4659, June-August 1906 (shrub ~* m™., flowers orange-yellow, fragrant) (AA); in calcareous mountains, region of Yunnan fu, alt. 2200 m., O. Schoch, a 36 (AA); in hedges near Hoching, alt. 2500 m., C. K. Schneider, S71 Sept. 26, 1014 (AA); in thickets near streanis at foot of mountain near Lichiang, alt. 3000 m., C. K. Schneider, no. 3222, utes: inodores) (AA). Southeast Tibet: open scrub by ‘Stfeams on the Salween-Kiu-chiang divide, Tsarong, lat. 28° &'N., long. 98° 15’ E., alt. 2135-2440 m., G. Forrest, no. 18927, September 1919 (shrub 1-2 m.) (AA); amongst rocks on dry stony ‘Slopes in side valleys on the Salween-Kiu-chiang divide, Tsarong, lat. ag° 40''N., long. 98° 15’ E., alt. 2440 m., G. Forrest, no. 19251, /“Piember 1919 (shrub 0.5 m.) (AA). 152 JOURNAL OF THE ARNOLD ARBORETUM [voL. xnt In examining the Indian along wath the Chinese material, one, at first, is impressed with the seeming differences and inclined to recognize the species J. revolutum Sims and J. Wallichianum Lindley as distinct from J. humile L. This separation would be based on the larger, more numerous and rather obtuse leaflets of — the Indian material as compared with the small, acute, and lesser — numbered leaflets of the Chinese species. The number of flowers — in the inflorescence of the Himalayan material is greater (usually | 8-10) as contrasted with the usual 3—5-flowered inflorescence of the Chinese specimens. On careful study of the massed material, however, one finds a sufficiently distinct graduation or variation — in all the characters mentioned making it difficult to recognize anything other than a single, very variable species, J. humiale. Series 2. TRIFOLIOLATA DC. A. Calyx lobes foliaceous. B. Leaves persistent, coriaceous, present at time of iver BB. Leaves deciduous; flowers appearing before leaves. C. Plants erect or scandent; ; simply branched. D. Leaves uniforml © CS ag Mice ees J. nudiflorum — y gree DD. —- dnivitar or sue enw .6a. J. wir aureum cately ram oeew eer ee eee eee seer ee eeerereee eee ‘D. eke tly quite vestigial; terminal leaflets same size of only larger than lateral . i - > bd : ° aoe. Dracram 1 | overs and only sister chromatids are paired as shown in diagt3™ \ When the homologues are separated at metaphase the crossover * 1932] ; SAX, MECHANISM OF CROSSING OVER 185 the chiasma will result in two non-crossover chromatids and two | crossover chromatids. Every chiasma represents a crossover, but according to Maeda and Belling every point of contact or node at diplotene is not necessarily a chiasma, but may be an overlap or temporary fusion of the paired chromatids. The assumptions upon which this theory is based have been criticized by McClung (1927). The more recent arguments of Belling and Darlington will be con- sidered later in this paper. In 1930 the writer presented an hypothesis to account for the | mechanism of crossing over which seems to have considerable cyto- logical support and is in accord with the genetic requirements. This theory is based on the usual assumption that chiasmata are formed by the alternate opening out of sister and non-sister chro- — matids at diplotene. A crossover is caused by a break in the two | crossed chromatids at a chiasma between diplotene and late diakine- ; sis. (Diagram 2.) : It is assumed that sister chromatids are always associated at the _ Spindle fiber attachments during the first meiotic civision, which is _ Maccord with both the cytological and genetic evidence. The second meiotic division separates the sister chromatids and is an equational division for regions of the chromosome between the Spindle fiber and the first crossover. | The paired chromatids are assumed to be associated, gene by gene, throughout their length, so that a crossover usually occurs between ey same: two consecutive genes in each chromatid as shown in m3. The spindle fiber is to the right and is terminal. 22 ——)+———-. a ) eee = -3-- GO i i = =O — = “°° 7% iar e---o--©--+--- | Diagram 3 oe In rare cases unequal crossing over occurs (Sturtevant we ee ~ ean 1931) which can be accounted for if the genes in the iad 186 JOURNAL OF THE ARNOLD ARBORETUM [voL. xa of the chiasma are not closely paired. In such a case crossing over — need not occur between the same two consecutive genes in each — chromatid, but may result in gene duplication and deficiency as ) shown in diagram 4. hs ' 4 8 bs 12s 2 OLS t Sage reese paler re ees D1aGrRam 4 Gene duplication and deficiency occur in case of the bar genein — the X chromosome of Drosophila only when there is crossing over : between forked and fused, so Sturtevant concluded that crossing over does not occur between sister chromatids in this region of the — X chromosome. The order of the two duplicated genes bar and — infrabar may be BB’ or B’B. In diagram 4 the order of these genes — is BB’ but if the chromatids cross and break above these two genes — the order will be B’B. All of these genetic observations can readily — be explained on the writer’s hypothesis. Crossing over between sister chromatids, according to this theory, — cannot occur before the first regular genetically detectable crossoveh and at other points can occur only when a half twist occurs in one pair of sister chromatids ‘accompanied by an opening out of non-sis u ter pairs of chromatids on both sides of the half twist. (Diagram5.) Diacram 5 ea The first crossover from the spindle fiber end of the chromosome would invariably occur between non-sister chromatids, but in some : cases subsequent crossovers might occur between sister ge Thus there would be little chance for a sister thread crossove? ee oo the bar locus in the X chromosome of Drosophila becavs apd is Only i bout 12 units from the fiber attachment point. hes . hypothesis is again in accord with the genetic observation 193) crossing over does not occur between sister chromatids at the et locus (Sturtevant 1928). yale | Factors which suppress crossing over in Drosophila (Gowe? ae | 1928) also inhibit chromosome pairing. Crossover re@ fost known to be caused by inverted segments (Sturtevant 1926) t#™* 1932] SAX, MECHANISM OF CROSSING OVER 187 locations (Dobzhansky 1931) and may be caused by deficiencies. Any factor which inhibits chromosome pairing would of course de- crease or eliminate crossing over between homologous chromosomes on either theory of crossing over. JANSSEN’S “PARTIAL CHIASMATYPY” HYPOTHESIS According to Darlington (1930) “partial chiasmatypy,”’ which demands crossing over between two of the four chromatids at every chiasma, is the only possible working hypothesis for the correlation of the cytological and genetical observations.” Darlington’s argu- ment is based on the types of chromosome association in triploids and tetraploids. The drawings and diagrams in his 1930 paper “are the result of a special search for the missing configuration in this critical material.” In Tulipa and Hyacinthus triploid chromosomes Were occasionally found to be associated as shown in diagram 6. A third chromosome intercalated between two chiasmata of a bivalent by a single chiasma is assumed to be critical evidence that this chiasma is caused by a previous crossover between two chromatids. Ifa third chromosome in triploids is invariably intercalated between two chiasmata in the other two chromosomes by a single chiasma there might be some justification for Darlington’s conclusion, but since the frequency of such types is not known, the evidence pre- sented is not critical. ; These types of trivalents can also be explained on the assumption that chiasmata are caused by alternate opening out of sister and non-sister chromatids and that one chiasma disappears before dia- Kinesis due to breaks in the chromatids as shown in diagram 7. DiaGRaAMs 6 AND 7 According to this theory of chiasma formation chromosome Cc must “orm two chiasmata with chromosome B. But if one of these chias-- : mata frequently breaks, as would be expected accordin g to the “nter’s theory of crossing over, then at diakinesis chromosome . will often be associated with B only by a single chiasma. Darling- : ‘on’s tetraploid configurations and Oecnothera figure 8 chrom eeont _ arlington 1931 B) can be interpreted in the same way, and canno ig eee og |S iar 188 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm be considered as critical evidence in favor of Janssen’s partial chias- — matypy theory of crossing over. | If all three homologous chromosomes in triploids can pair simul- taneously at the same locus, as Belling has described in Hyacinthus, — then Darlington’s “critical” figures could be obtained without any — crossing over. . In a later paper Darlington (1931la) attempts to correlate the — frequency of chiasma formation with the percentage of crossing over in Primula sinensis. The SBGL chromosome, which is assumed to _ be one of the longest chromosomes of Primula ‘sinensis, was esti- mated from crossover data, to be 111.6 genetic units long, but this _ may be incorrect since Haldane’s (1919) correction of map distance is not valid (Morgan, Bridges and Sturtevant 1925). According to Darlington’s theory of crossing over each chiasma will result in 50 per cent crossing over and the length of a chromo- some in genetic units can be calculated as 50 X the number of chiasmata. Darlington finds an average of 3.5 chiasmata per bi- valent which he considers may be an under-estimate for the longest _ chromosomes. Due to an error, Darlington calculated the crossover length, based on number of chiasmata, as 58.3 to 116.7 units, which he says is “in strict agreement with expectation.” But the correct — length based on chiasma frequency, should be 3.5 X 50 or 175 units as an average, or to range from 116.7 to 233.4 units, as Darlingtom later discovered. Obviously the length of the longest chromosome — as measured by crossing over is only about half as long as the length expected, if each chiasma represents a crossover. ae On the writer’s theory of crossing over, chromosome SBGL should be somewhat more than 100 units long which is “in strict agreement : with expectation”! It is probable however, that neither the genetic - ame! the cytological data are adequate for any serious calculation of the crossover length of any of the chromosomes in Primula. In peek to explain the absence of crossing over in the Drosophila ar! On the alternative theory of crossing over, it is assumed ie # : , chiasmata are formed as they are in the Phe — perhaps less frequently, and that chromosome &""~ 1932] SAX, MECHANISM OF CROSSING OVER 189 ment from diplotene to metaphase is sufficiently gradual so that the chromatids can adapt themselves to the changes without breaking. A gradual development would permit the chromatids to contract as the chromosomes contract so that no coiling of the chromonemata would occur. The chiasmata would be free to terminalize without breaks in the chromatids, as seems to be the case in the Orthopteran chromosomes. Since terminalization is easily effected the meiotic metaphase in the male should be of short duration as compared with the same stage in the female, as seems to bethe case (Huettner 1930). The reverse would be the case according to Darlington’s theory. Crossing over in the female is most frequent towards the ends of the second and third chromosomes of Drosophila melanogaster so that only two subterminal chiasmata would usually need to be pulled apart, but in the male the two reciprocal chiasmata near the spindle fiber would have to be terminalized for practically the entire length of the chromosomes. Darlington does not attempt to explain why breaks occur in the chromatids’at pachytene, why the breaks in the crossover chroma- tids almost invariably occur at the same level, why chromatids should recombine in new association after they break, why only two chromatids cross over at any one locus, or why one crossover inter- feres with the occurrence of a second one in the same region. : According to Darlington, at least one chiasma is essential for pairing of homologous chromosomes and every chiasma represents a crossover. Therefore, crossing over must be universal in all normal species and no normal chromosome ‘can be less than 50 genetic units long. No explanation has been presented to account for the behavior of the fourth chromosome of Drosophila melano-— gaster. This chromosome is about as regular in conjugation and disjunction as the X chromosome, but no crossing over occurs in the fourth chromosome. There are also other obvious objections to — Darlington or eo of chromosome pairing (O’ Mara, in press). BELLING’S THEORY OF. CROSSING OVER Belin (1981 a) has recently presented an hypothesis to account the mechanism of crossing over. The homologous chromosomes me as single chromatids. Half twists occur in the paired homolo- gues at early pachytene before the secondary split has begun. the secondary split occurs the new chromomeres must form new connecting fibers and at each twist they take the shortest route : ~ connecting adjacent chromomeres. Thus crossing over would Ghee oy between the two new chromatids formed at late pachy- i Te b jously invalid because in Drosophila RRs SR 190 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm crossing over is found in more than 50 per cent of the emerging X chromosomes, and nearly 75 per cent of the third chromosomes have one or more crossovers (Redfield 1930). ! More recently Belling’s (1931 b) modified his hypothesis. This new theory seems to be plausible since it accounts for crossing over between any two chromatids, and provides a mechanism to explain translocations and inversions in somatic chromosomes. If Janssen'’s partial chiasmatypy theory of crossing over is correct Belling’s hypothesis of the mechanism involved would seem to be the only logical explanation. Unfortunately there are a number of serious objections to this theory. | According to Belling a half twist between single chromatids occurs - at pachytene. When the new chromomeres are produced at the secondary split the connecting fiber between genes may remail with the old gene (genes and chromomeres are assumed to be sy? onymous) according to the laws of chance. The new connecting fibers then unite the free genes by the shortest path (Diagram §). As shown in the diagram 8, the new connecting fibers may result in @ crossover between a’ and b’ or between a’ and b. If the necting fibers remain with the old genes or pass to the new ones at random there will be random crossing over between any two ast) four chromatids. 7 | Random assortment of connecting fibers would also meat that - both somatic and meiotic chromosomes crossing over betwee? sistet threads would be very frequent. In a chromosome containing * hundred genes 50 sister crossovers would be expected. In i? matic chromosomes this sister crossing over would result ait twisting of the two chromatids at late prophase and at ” a if the new connecting fibers which unite old and new genes per . random on either side of the old connecting fiber. But there #* only a limited number of twists or half twists in the somatic eae eae ees figure 1) and between paired chromatids at me Belling’s theory is also based on the assumption that the ore gous chromosomes first pair as single chromatids at meiosis: * Work of Kaufmann (1926) and Sharp (1929) seems to show tbat 1932] SAX, MECHANISM OF CROSSING OVER 191 somatic chromosomes in certain plant species are already split into two chromatids at the preceding anaphase. According to Kauf- mann (1931) the chromosomes at the telophase of the last pre- meiotic division show the parallel chromonemata. In Paratettiz, the chromosomes are split longitudinally when they enter the spermatid (Robertson, 1931 a), and Robertson (1931 b), also shows that the homologous chromosomes in the Tettigidae are divided into sister chromatids when they begin pairing at meiosis. Similar observations have previously been made by Robertson (1916) and McClung (1928). These observations cannot be recon- ciled either with Belling’s theory of crossing over, or with Darling- ton’s (1931) theory of meiosis. One other point in comparing the two theories has been presented by Belling (1931 b) in connection with the behavior of unequal homologues. One pair of unequal homologues were found in Aloe purpurascens. If chiasmata are formed by the alternate opening out of sister and non-sister pairs of chromatids, in some cases a short arm should be associated with a long one at diplotene or dia- kinesis, but no such association was found. On Belling’s hypothesis no pairing of long and short arms would be expected since only sister chromatids are paired at diplotene. If the chiasma is to the distal side of the fiber constriction, as shown in diagram 9b, then a short and a long arm should be associated on the writer's hypothesis, unless a crossover had occurred. On Belling’s hypothesis such a chiasma (9a) should invariably result in a distribution of a long and a short chromosome to each pole at the first meiotic division, but Belling observes that “in some cases they are observed to separate mto short plus long and short plus long chromatids.” * ~ ~~ woe" - bi »* op | rs oo 2c “——S otet a - cA ges i: od ce 4 4 4 + 0% ¢ & = Pa s ae - o. a ra If, in most cases, the segregation at anaphase is into two short and a long homologues, as Belling’s statement would imply, then the sma must be to the right of the fiber attachment point (9c) or the homologues are terminally associated without chiasma forma- ‘on. On the writer’s hypothesis a chiasma to the right of the fiber sf i m alee cs a . mt * z. ‘ 4 i . i but never thes tien ak 2 1 192 JOURNAL OF THE ARNOLD ARBORETUM [vou xm vations are inadequate for any critical test of the method of chiasma formation. REDUCTION IN NUMBER OF CHIASMATA The writer’s theory of crossing over postulates breaks im the chiasmata so that a decrease in the number of chiasmata would he expected between early diplotene and metaphase if crossing overt occurs. At early diplotene, as the homologous chromonemata open oil, the nodes and internodes are often very numerous. This condition is indicated in chromosome number 3 in figure 2, and is clearly shown in Lathyrus (Maeda, 1930, Fig. 17 8) and in Zea (Cooper and | Brink 1931, Fig. 1). These nodes may be chiasmata, most of whieh subsequently meet as the homologous chromonemata open ot more completely at the internodes. There is also a possibility that some of these nodes which disappear are due to breaks in chiasma at this stage of meiosis. Since the individual chromatids cannot be identified many of these early nodes may be only temporary adhe sions as Belling has suggested. . In Tulipa Newton (1926) pictures about 30 chiasmata in 11 diplo- tene chromosomes (Fig. 25). At early diakinesis the number of chiasmata are especially clear, and show a total of about 27 chiss | mata for the 12 chromosomes (Fig. 31) while at late di i og | Newton found only 15 chiasmata (Fig. 30). Thus the avem® | number of nodes or chiasmata per bivalent chromosome is reduced | from almost 3 at diplotene to only 1.25 at late diakinesis. In Lilium longiflorum, Belling (1928) found a decrease in ee of nodes between diplotene and late diakinesis of 43 per cent an : suggeststhat more nodes would have been found if counts could pe have been made at an earlier stage. Belling states that “the nodes { which disappear between diplotene and late diakinesis do not a” to be all or mainly twists.” ele In . later paper Belling (1981) concludes that in the same sper of Lilium the number of nodes is reduced from 42.5 at diplote the about 30 at late diakinesis and metaphase, and suggests aor _ nodes which disappear are half-twists or temporary adhesions ©” pe ae - D arlington (1931 a) finds an average of 3.5 chiasmata at dij , - Primula sinensis but only 1.89 chiasmata at metaphase: bivalents at metaphase are associated only by terminal 2 r This decrease in number of chiasmata is attributed to terminal WAS SEEDY VS eS t gale pees) es havefouror . hace é€ * ~ 1982] SAX, MECHANISM OF CROSSING OVER 193 In Rosa blanda, Erlanson (1931) obtains an average of 1.94 chias- mata per bivalent at early diakinesis, but only 1.31 at metaphase. Doubtless a greater number of chiasmata would have been found at diplotene. In Matthiola Philp and Huskins (1931) found that the average chiasma frequency was 2.26 per chromosome at early diakinesis and only 1.54 at metaphase. The counts of chiasma frequency at different stages of meiosis are summarized in Table 1. The estimated average crossover length of the chromosomes is calculated from the number of chiasmata which disappear between diplotene and metaphase on the basis of the writer’s hypothesis. Since Belling considers that the chiasmata found at metaphase constitute all or most of the true chiasmata formed, the crossover length is also calculated on this basis. Dar- lington considers the nodes found at diplotene in Primula as chias- mata and attributes the loss of chiasmata to terminalization, so crossover lengths are also calculated on diplotene and early dia- kinesis counts. TABLE 1 Ave. no. xta. per bivalent Estimated Ave. c.o. length Diplotene or early Sige diakinesis Metaphase No. lost Ss. B. af a ee eee 1.3 1.5 15 65 140 Lilium... 35 25 10 50 195 175 Secale... 2... 85 18 17 85 90 175 Primula ..... . 3.5 1.9 1.6 80 95 175 pa Ce 1.9 1.3 6 30 65 95 Matthiola.... 2.3 1.5 8 40 15 115 Callisia... . 2 1.5 2.7 135 15 210 Counts of chiasma frequency at early diakinesis do not represent the number formed at early diplotene so that in most of the genera _ listed above, the number of chiasmata lost should be increased con- siderably and consequently the crossover length would be increased on the writer’s and on Darlington’s hypotheses. Little is known Concerning the crossover lengths of the chromosomes of the species listed above, but the calculated lengths are comparable to those found in Drosophila and Zea. Two of the autosomes in. Drosoph _Melanogaster are somewhat more than 100 units long and in Zea the 5 chromosomes containing the most mutations range from about “ : _ t068 units long (Lindstrom 1931). The chiasma frequency in Zea— chromosomes is about 1.5 at metaphase (Randolph), but data on - Carlier Stages are very meagre although Fisk (1927, Fig. $2) pictures _ two bivalents with about 5 nodes each. The genetic and vies “al work on Zea should soon provide adequate data for cor | _ “eMparison of chromosome behavior and crossover frequency. 194 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Darlington assumes that the reduction in number of chiasmata— is due to terminalization. But as Belling (1931 b) points out, the | coiling of the chromonemata would prevent any appreciable move- ment of the chiasmata after diakinesis. No coiled chromonemata — have been described by Darlington, probably due to inadequate : fixation or staining for showing this structure, rather than the ab | sence of coiled chromonemata in the species studied. Coiled chro monemata have been described in many species and recently Shinke — (1930) has described such chromosome structure in about 25 differ ent genera. The coiling of the chromonemata probably begins at late diplotene so that no extensive movement of the chiasmata cal occur, in most cases, between diplotene and metaphase. If its” assumed that the chiasmata are terminalized one might expect that — all of them would frequently terminalize before metaphase all produce univalent chromosomes. In Primula, and in the Solanacet, the bivalents at metaphase are usually associated only at the two ends forming the typical ring shaped bivalent. Darlington assuilis _ that an average of about 1.5 chiasmata pass off the ends of the Primula bivalents. But why should the remaining two chiasmalt always stop at the ends of the bivalent? If each chiasma represents a crossover then the M chromosome Vicia faba must be more than 400 crossover units long. Mae (1930) finds an average of 8.1 chiasmata in this chromosome, at the number may be as high as 13 in some cases. On any theory 4 crossing over a twist must occur in the chromatids either at the chiasma or at an internode for every crossover. If one oF _ crossovers occur there should often be some interlocking of hom gues at anaphase as seems to be the case in Lilium and Laity’ (Maeda 1930, Sax 1930). With 8 crossovers one might expect © siderable difficulty in separating homologues at the first meiotle division. If the chiasmata are not the result of crossovers thet : would of course be no difficulty in the division of homolog _ (McClung 1927). ees ie NON-DISJUNCTION AND CROSSING OVER. : Non-disjunction in Drosophila melanogaster was first described Brudges in 1916. Primary non-disjunction is caused by the P duction of ae >. Bay and “no-X”” eggs which when combi 1 - pocmeal sperm produce male and female exceptions. These exceptions occur with a frequency of about 1 in 2000. Male est fe tions are usually produced from 4 to 8 times as frequently ® _ male exceptions (Bridges 1916, Safir 1920, Mavor 1924, Andes” aes : aoe excess of male exceptions, which are produced = 1932] SAX, MECHANISM OF CROSSING OVER 195 “no-X”’ eggs, is attributed tothe failure ofthe XXs to pass to either pole at the first reduction division. In normal stocks of Drosophila little if any crossing over takes place between the two X chromosomes which pass to the same egg. (Bridges 1916.) In high non-disjunction lines, crossing over is greatly reduced in both the normal and in the exceptional progeny (Morgan et al. 1925). In Anderson’s (1929) high non-disjunction stock crossing over between scute and forked was reduced from 62 per cent to 20.9 per cent, and in the X Xs which produced exceptional females the crossing over was only 7.3 per cent. Crossing over was almost eliminated in the region of vermillion, 40 units from the left end of the X, but increased towards the ends and was almost normal at the left end. Dr. Anderson informs me that the reduction in crossing over was caused by a translocation involving the X and the third chromosome. Dobzhansky (1932) also finds that translo- cations reduce crossing over and that non-disjunction is positively correlated with the length of the autosome attached to the Y. Due to the attraction of different chromosome segments chromo- some pairing is often weak and crossing over is reduced. In Anderson’s (1931) primary exceptions produced by X-rays the total crossing over between scute and forked was about 60 per cent of the normal, but crossing over was found in all regions, and in two of the regions studied crossing over was almost normal. About 14 per cent of the XXs were homozygous for the forked locus where less than 5 per cent would be expected. This excess of homozygo- sis at forked is attributed to non-disjunction at the second matura- tion division, the only logical explanation of the results obtained. In view of the great irregularity of crossing over in different regions ‘ompared with the control it seems possible that there was 6 per cent of crossing over to the right of forked. At any rate crossing over between the XXs which pass to the same egg, is not confined ‘ the distal ends of the chromosomes and some crossing over occurs to the right of forked. ‘ ; non-disjunction is due to a failure of XXs to separate at the - first maturation division it is rather difficult to account for the de- “tease in crossing over in high non-disjunction stock if chi “ Pivegiy: crossovers. According to Darlington (1931) — vm the only bonds between homologous chromosomes and f a fons pout ate formed the chromosomes would not be expected to rm bivalents at meiosis. Failure of the XXs to disjom would then © attributed to more than the usual number of chiasmata, but an — of chiasmata would also mean an excess of crossing seas either Belling’s or Darlington’s interpretation. On the writer's 196 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm interpretation the decreased amount of crossing over in high non disjunction lines could be attributed to few breaks at chiasmata 9 that the homologues would be united by more than the usual nun ber of chiasmata and could not separate so readily. There is good evidence, however, that non-disjunction is reallya | failure of chromosome pairing at the later stages of the first meiolie divisions. The decreased crossing over in high non-disjunctiot | strains of Drosophila, and in the XXs which pass to the same eggit normal stock, can be attributed to non-conjugation or to pairingit: only restricted regions of the chromosome. If only a few chiasmata are formed, due to incomplete pairing, then crossing over would be | reduced on either theory of crossing over. The premature sepale tion of homologues could be attributed to either early terminalis | tion of these few chiasmata, or to breaks which would result in cross-overs. On Darlington’s and Belling’s theory, all of the chias- | mata must be prematurely terminalized. The failure of chias@ formation, or breaks the few chismata formed, would produce unpaired homologues at the first maturation division. If ‘thee univalents pass at random to either pole, then half of the should contain no X or 2 X chromosomes, and half should co tain the usual 1X. The univalents often fail to reach the polesit species hybrids and in haploids, and are often lost in the cytoplasm. If then one of the two X chromosomes should be lost there will bea excess of no-X eggs which produce the male exceptions. Su! behavior of the univalents would account for the excess of male | compared with female exceptions. This explanation of non junction seems to be more plausible than the assumption that » two XXs are so intimately paired that they fail to divide an¢ both pass to the same pole or fail to reach either pole. 2 4 If non-disjunction is due to a failure of the X chromosomé . separate it is difficult to account for the decrease in crossing in high non-disjunction lines of Drosophila on the theories of ye over proposed by Belling and by Darlington. On the other ™ | if non-disjunction is due to non-conjugation at metaphase, ! a probable, then all chiasmata must be broken or pre urely minalized. If chiasmata are crossovers then all of the bo. kg somes which show crossing over must have separated prem ) by complete terminalization and elimination of all chiasmal® “ie non-disjunction stock of Anderson’s about 7 pe the crossovers between the two XXs which pass to ee ee —s occur to the right of forked, and in the X-ray material © vit b oohiys AEE be may have occurred to the right of rete Js an additional crossovet © Pen 1932] SAX, MECHANISM OF CROSSING OVER 197 to the left of forked. If crossovers produce chiasmata then in these chromosomes there are usually two chiasmata which must be terminalized to produce “‘non-disjunction”’ and one of the chiasmata is to the right of forked so that it would have to be prematurely terminalized for practically the entire crossover length of the chro- mosome. If such terminalization is possible it would seem that non- disjunction should be frequent, but primary exceptions occur with a frequency of about one in 2000, in normal stocks of Drosophila, and in only 2.5 per cent of the progeny from high non-disjunc- tion lines. The cytological evidence also indicates that extensive terminalization or movement of chiasmata is improbable (Belling 1931 b). If Painter’s (1931) cytological map of the X chromosome is cor- tect there is some possibility that premature terminalization could occur because the region from scute to forked constitutes only about a third of the cytological length of the X chromosome. Stern (1931) working with the same stock finds that the region from scute to forked constitutes about one-half of the X chromosome and his figures clearly support this interpretation. It would seem improb- able that two chiasmata could be prematurely terminalized, one for more than half the length of the X chromosome. Even in species where the h log h are always iated by only terminal chiasmata at metaphase there is little or no tendency for premature separation of the chromosomes. On the writer’s theory of crossing over it is not difficult to ex- plain “non-disjunction” even where crossing over occurs near the spindle fiber end of the X chromosome. If few chiasmata are formed at diplotene and all of them break in certain bivalents, then these chromosomes will be loosely associated so that precocious disjunc- tion before metaphase would be possible. There are, however, certain types of double crossovers that are difficult to explain unless some terminalization of chiasmata occurs. hh Anderson’s high non-disjunction data there is 1 and in the X-ray ote... aab sa data there are 4 double crossovers of the type sha” babe tieee ciprocal-equational double crossovers can only occur, on the Seer hypothesis, if there is an unbroken chiasma between ee chromatids between the first and second genetically detectable ‘“ossovers, or a twist in the sister chromatids between the second = third crossovers. In Anderson’s (1931) table 3, one of these *elprocal-equational crossovers is the result of a second and _ Sover since the forked locus is homozygous, but the other three Oe crossovers apparently involve a first and second eeeeees 198 JOURNAL OF THE ARNOLD ARBORETUM [vou xm | because forked is heterozygous in these cases. Under such condi tions it would be necessary to assume that a chiasma between sister chromatids must be prematurely terminalized to produce “non disjunction.” Since these chiasmata would be to the left of forked | and in two cases to the left of cut, it is not impossible that prema — ture terminalization might occur. It is also possible that thes three reciprocal-equational crossovers are produced by second and third crossovers, and that the first crossover was a reciprocal to the right of forked, or that these exceptional crossovers are producel by the non-disjunction at the second maturation division. aj The writer’s hypothesis offers a simpler and more plausible er planation of “non-disjunction”’ than Janssen’s partial chiasmatypy hypothesis, regardless of which interpretation of “ non-disjunction 1s correct. | RANDOM CROSSING OVER BETWEEN THE FOUR CHROMATIDS. Crossing over is not limited to two of the four chromatids, be- cause more than 50 per cent of the X and third chromosomes | Drosophila which emerge from the reduction division, have on¢% more crossovers. In 62 units of the X chromosome 54 per cent of the emerging chromosomes show at least one crossover (Andersot | and Rhoades, 1981), and in the third chromosome 72.9 per cent | the chromosomes are crossovers (Redfield, 1930). If crossing ove | occurs only between the new chromatids as Belling (1931 a) has suggested, then not more than 50 per cent of the emerging CHM” | somes should be cross overs, regardless of the map length of ad ‘ chromosome. a More recently Belling (1931 b) has assumed that “the old « 48 necting fiber is indifferent as to which chromiole it will remain 4 — On this assumption crossing over would be at random betweet the ; four chromatids. Bee: Belling’s hypothesis can be modified to comply with aha cytological and genetic data. If the old connecting fibers ust! remain with the old genes at the time the new chromatids - formed, then crossing over will usually occur betwee? the te new chromatids as Belling (1931 a) has assumed. In some © the old connecting fiber will unite a new and an old gene 9° © crossovers will occur between sister chromatids. Such crossover if sufficiently numerous, would result in random crossing over e™ | all crossovers at chiasmata were between the two new chrom on Crossing over between sister chromatids would not be depen®™ a q chromosome pairing and would be expected to occur wih frequency in all regions of the chromosome. Poe 1932] SAX, MECHANISM OF CROSSING OVER 199 On the writer’s hypothesis random crossing over among the four ehromatids can occur only if half-twists are sufficiently numerous in the paired sister chromatids. If the chromosomes are already split into sister chromatids at the time of meiotic pairing it would seem improbable that the same two non-sister chromatids would be paired throughout their length. If the secondary split occurs after pairing it would also be improbable that the sister chromatids would lie parallel throughout their length. One would expect some twists in the paired sister chromatids in either case. Such twists are found in both somatic and meiotic chromosomes. In normal stocks of Drosophila and in cases of ““non-disjunction,” the genetic assortment of chromatids will be at random for the first crossover, because of their free assortment at the second maturation division. In attached-X stock, however, the random distribution at both first and second crossovers can be tested. Random crossing over can occur in attached X Xs, on the writer’s theory, only if twists in paired sister chromatids are sufficiently humerous. Such half-twists could occur between the point of spin- dle fiber attachment and the first chiasma, or between the second and third chiasmata. (Diagram 10.) They could not occur between the first and second chiasmata without the formation of an addi- tional chiasma between sister chromatids, but such chiasmata would be expected only in rare cases. co. = ge Dracram 10 Only two types of first crossovers need be considered; (0) me twists m either pair of sister chromatids between the attachment point ane the fir crossover chiasma, and (1 a) one half twist in only ope 8 of sister chromatids, Aa. The other two types, (1b) a half vguiey chromatids Bb, or (la—1b) in both pairs of sister chromatids, need _ Rot be considered since they are reciprocal to the first two types. _ etween the first and second crossovers four types of chromatid iation must be considered if crossing over is at random. et Pet Ne (0) no twists in either pair of sister chromatids, (1 a) one 2S twist in chromatids Aa, (1b) one half twist in chromatids Bb, 200 JOURNAL OF THE ARNOLD ARBORETUM {vou xu 4 and (la-Ib) a half twist in each of the paired sister chromatids. | If such half-twists are frequent there might be several of them be- : tween the point of attachment and the first crossover, or between | crossovers, but if these twists occur at random they will produce the 7 same types and proportions of crossovers as expected on the assump- | tion that the above four types occur in equal proportions. ae | If no half twists occur as we would expect types (O) and (la-Ib) | to be formed in equal proportions, but no types la or 1b would b expected. At a given chiasma crossing over could occur between A and b chromatids, or between B and a. If the first crossover between A and b the second crossover may occur between A andb or between aand B. With no half twists in pairs of sister chromatids — only equational crossovers could occur in attached XXs. If chiasmata are produced by previous crossovers, the modifica- | tion of Belling’s hypothesis would seem to be the most plausible — explanation of the mechanism involved. If sufficient crossovels occur between sister chromatids, the detectable crossovers will be | at random between the four chromatids. The association of chro : matids shown in diagram 11 will give the same random crossing. | over as direct crossing over between A and B or b, and B with © either A or a. ays ind . to ist ie ' . re aoe chromatid Bb, or (la-1b) an equal number of sister crossovers | each of the pairs of sister chromatids. ay the : On either theory we have 8 types of chromatid association an ‘ attached X bivalents, which produce 16 classes of crossove™ | results of random crossing over at two points in the bivas t shed 3S ulin a . : a Cee e re shown in table 2. 202 JOURNAL OF THE ARNOLD ARBORETUM [vou xm In the second column are listed the types of first crossovers which — have been considered on the writer’s hypothesis and on the modified | hypothesis of Belling. The genetic constitution of the attached | XXs which pass to the egg cell are indicated under crossover types. — Difference between chromatids A and a, or B and b, are not indi cated since they cannot be detected by genetic tests. The first detectable crossovers from the attachment point are shown to | occur in the ratio of 2 : 1 if crossing over is at random. ‘Three types of second crossovers occur :—(1) equationals homozygous to the left _ ab eet (= rap (2) equational crossovers homozygous to the right — aa e or =) or (3) reciprocal crossovers (E or 2 . With random | crossing over these types should appear in the ratio of 2 : 8:1. At the left of the first crossover chiasma one-half of the attached Xchr- | mosomes should be homozygous, and one-fourth should be homo — zygous recessives. At the left of the second crossover chiasma one — fourth of the attached XXs should be homozygous and one-eighth | recessives. | If crossing over is at random the first crossovers from the spindle fiber end in attached XX should be equationals and reciprocals the proportion of 2:1 (Table 2). Anderson (1925) found equé | tional and reciprocals in his attached X data, in the proportion ‘ 4 29.7 to 15.6. Although there is a slight excess of equationals the : ration is very near random expectation. If equational and recip — cal first crossovers occur in the ratio of 2.1 then the percentage : homozygous recessives should be half of the crossover distance be : tween the spindle fiber and the first crossover. For forked, which ‘ is about 10 units from the spindle fiber, the percentage of recessV i in attached XXs was found to be 5.2 by Anderson, 4.9 by Rhoades ; and 5.1 by Sturtevant. The genetic evidence indicates that the ; chromatids are assorted at random at the first crossover- Random assortment of chromatids would be expected if there® | an average of 0.5 or more half-twists or sister crossovers DE” the attachment point and the first crossover. Even with relative! q few half-twists or sister chromatid crossovers in the X chromoso™ random crossing over might be expected at the first cross0ve | chiasma because the region between the spindle fiber and ie a crossover would usually include more than half of the length of HF | chromosome (Stern 1931). roy The randomness of chromatid association at the second crossove : oe be determined from the proportions of types of past | overs in attached X and XXs from “‘non-disjunction.” The™ 1932] SAX, MECHANISM OF CROSSING OVER 208 disjunction types of crossovers can be derived from table 2 by com- bining each two crossover combinations into the four possible com- binations expected if non-disjunction is due to non-conjugation at metaphase. The classes of crossover combinations will be doubled but the proportion of types of second crossovers will remain the same as shown in table 2. The available data on types of second crossovers are presented in TABLE 3. Types of 2nd crossovers. (Cf. Table 2) ) (3) Anderson 1925 Table 6............... 1 (.6) 7 (8.0) 1 (.6) Anderson 1929 Table 24-25........... 1 (.05) 6( .25) 1 (.09) Anderson 1981 Table 3............... 3 3 1 The attached X and high non-disjunction data show a lower pro- portion of type 1 second crossovers than would be expected if cross- ing over is at random, but the numbers are too small to be of much value. In the X-ray non-disjunction data there is an excess of type 1 second crossovers caused by non-disjunction at the second division. The third test of randomness of crossing over can be made by comparing the amount of homozygosis in attached X chromosomes with the amount expected on random assortment of chromatids. The amount of homozygosis at the left end of attached X, or XXs from non-disjunction, is dependent on the number and types of crossovers. The amount of homozygosis at the left of first and second crossover chiasmata is shown in table 2. Fifty per cent of the two emerging X chromosomes should be homozygous to the left of the first crossover and twenty-five per cent at the left of the second, It is first necessary to calculate the number of chiasmata which produce crossovers. As Belling (1931 b) has pointed out, the rela- hons between crossover chiasmata and crossing over is as follows if ‘rossing over is at random. TABLE 4 es . * 1 chramatias oe 0 aay 2 3 ’ Pei i. 8 Ee a : 0 = 16 Number of te a 8 ‘OssOver S@= 4 8 chiasmata Cer. 6 6 2 1 Aa } 4 6 4 | __The data from Anderson and Rhoades (1931) table 1 have been : ised the calculation of the frequency of chiasmata which produce 204 JOURNAL OF THE ARNOLD ARBORETUM crossovers in the bivalent X chromosomes. Crossover recombina | tions are shown for 26,908 X-chromosomes of which 46 per cent showed no crossing oyer in the 62 units between forked and scute, — 46 per cent were single crossovers, 7.6 per cent were double cros- overs, 0.2 per cent were triple crossovers and .01 per cent were : quadruple crossovers. 1] crossing over is at random chiasma vik Bet can be chai | from crossover frequency as shown in table 5 i i] TABLE 5 2a | Calculation of chiasma frequency based on crossover frequency i the X chromosome of Drosophila Number of crossovers per chromosome 416% 46% 7.6% 0.2% 1% ath Xtal, ol 04 06 ae | | 45.00 45.96 7.54 16 | | Sed Bik. ihe ee 48 48 8218 | 45.83 45.48 7.06 ioe —2d Kita... 2... 7.06 14.12 = 982 ee 38.77 31.36 eee —Ist Xta......... 841 x2 = 62.7 cage fi None be Total chiasmata in bivalents. ta. 8.4% : 62.7% al sy Only one chromatid in 16 produced by quadruple chias ae _ Produce crossovers) will be a quadruple crossover, 50 the perce oo of quadruple crossovers must be multiplied by 16 to obtain number of quadruple chiasmata. Similar calculations are used 1 phveng chiasmata frequency from single, double, and triple With the alts frequency of single, double, and triple es | i Sed 8 possible bor determine e the percentage of homozygosis e ) o. note a 1932] SAX, MECHANISM OF CROSSING OVER 205 gosis at the left of the first crossover chiasma is 50, the second 25, (Table 2) and the third 37.5, if crossing over is at random. For homozygous recessives these percentages are 25, 12.5, and 18.75 respectively. The following table shows the percentage of homozy- gous recessives expected 62 units from the spindle fiber attachment in XX chromosomes. Chiasmata = _ Single Double Triple 62.7% 28 .2% 1.38% % homo. recessives = 15.67% 3.52% A% = 19.4% We would expect then, if crossing over is at random, to find 19.4 per cent of homozygous recessives in attached XXs at a point 62 units from the right end of the chromosome. The percentage of homozygosis found is significantly lower than expected. Sturtevant (1931) found 17.1 per cent homozygosis for scute, which is about 72 units from the right end of the X chromosome. This analysis was based on approximately 25,000 flies. Rhoades (1931) found about 18.6 per cent homozygosis for scute, but for ruby, which is about 64 units from the spindle fiber attachment, the percentage of homo- zygosis was found to be 17.7. Counts were made on about 42,000 flies. At a point 62 units from the spindle fiber the percentage of homozygosis recessives would be about 17.4 where 19.4 per cent would be expected on the basis of random crossing over. Anderson’s attached X data shows a similar discrepancy between the percentage of homozygosis found and the percentage expected on random crossing over. Chromosomes homozygous for the forked locus are assumed to have crossed over to the right of forked, and such chromatids are classed as crossovers in this region. Only bead of the chromatids, equational at forked, is a crossover to the right of forked, but both are included to make up for the reciprocal crossovers to the right of forked which cannot be detected from the data, Considering these crossovers to the right of forked, there are 59.5 per cent of the chromatids with no crossovers, 37.2 per cent with one, and 3.3 per cent with two. In terms of crossover chias- Mata, 25.6 per cent of the bivalent chromosomes have no chiasmata between the spindle fiber and end, 61.2 per cent have one chiasma, | and 13.2 per cent have double chiasmata which break. If crossing over is at random the percentage of homozygosis for cut should be 16.5 per cent in the XX chromosomes listed in Anderson’s table 6. The percentage of homozygosis actually found was 15.5, although for tan, still further to the right, the value was 16.1. ‘he percentage of homozygous recessives at the left end of at- : X chromosomes is lower than expected if crossmg Over 4 _ t random. But, as Anderson and others have suggested, the 206 JOURNAL OF THE ARNOLD ARBORETUM [vou xm lower viability of the homozygous recessive segregates would re. duce the percentage of these classes so that the true value might well approach the percentage expected on random assortment, | If there is any significant deficiency of homozygosis in attached X chromosomes, it would indicate that second cross-overs are not entirely at random. On the writer’s hypothesis such a deficiency | could be attributed to few half twists between the first and second cross over. For instance, if the average frequency of half twists were 0.25, the percentage of homozygous recessives 62 units from the fiber would be only 17.6 instead of 19.4 expected on random assortment. Belling’s recent theory should always give random assortment for all cross overs. The modification of Belling’s the | ory suggested by the writer would also fail to account for any de — ficiency of homozygous recessives. If crossing over is invariably at random between the four chromatids, Belling’s theory would seem to be the most valid interpretation of the mechanism d crossing over. If, however, crossing over is not entirely at random, é the writer’s hypothesis seems to offer the only solution. Although — there is a deficiency of homozygous segregates from attached Xx | chromosomes, the nature of the genetic evidence does not justify any final conclusion concerning random assortment of the chrom tids at the second cross over. re) CHROMOSOME PAIRING AND CROSSING OVER. ae When crossing over is eliminated in the Drosophila female there also a loose association of homologous chromosomes. (Gowe? | 1922, 1928.) This behavior is undoubtedly analagous to the case : asynapsis in Zea (Beadle 1930). A decrease in crossing ovet may : also be caused by inversions and translocations. Dr. Andersol® — forms me that his high non-disjunction line was caused by an ~ sion. In this case the decrease in crossing over was also associatet — with an increase in “non-disjunction” or failure of chromosomt is ee ee pairing. Dobzhansky (1931, 1932) has found that crossing “ : decreased and non-disjunction increased in flies heterozyg0 aes translocations. This behavior is attributed to conflicting attrae | a oe either theory of crossing over the reduction or elimination & ing over could be attributed to a differential rate of chrom | Some pairing compared with ch tid a es At pachyte P ” four chromatids are associated, at diplotene only two chromati® can be associated and at telophase the two chromatids are only at the spindle fiber iring | layed then the sister chromatids might be at a stage commonly | 1932] SAX, MECHANISM OF CROSSING OVER 207 found at diplotene, before the homologous chromosomes are paired, so that no chiasmata could be formed. In the cases of inversion no pairing of homologous genes would occur between inverted and normal chromosome segments and no crossovers could be produced in such regions. In heterozygous translocations chromosome pairing is delayed so that few chiasmata can be formed. The genetic and cytological evidence shows that crossing over is most frequent at the distal end of the X chromosome (Painter, 1931) and at the distal ends of the third chromosome (Dobzhansky, 1930). This localization of crossing over would seem to indicate that chromosome pairing in Drosophila begins at the distal ends of the chromosomes and proceeds towards the spindle fiber. Crossing over would occur at the ends of the chromosomes because the chro- matids are not sufficiently differentiated to prevent chiasma forma- tion, but towards the spindle fiber the paired sister chromatids be- come so united that chiasma formation is no longer possible. Since chiasma formation is associated with crossing over, on either theory, there would be few crossovers in the spindle fiber region and frequent crossovers at the distal ends of the chromosomes. Crossing over is increased in the X chromosome and in the auto- somes of Drosophila by changes in temperature and by X-rays. ‘ough 1917; Stern 1926; Muller 1925, 1926.) The increase in crossing over occurs primarily in the region of the spindle fiber at- tachment, but not in regions where crossing over is frequent in untreated flies. This behavior could be attributed to an accelera- tion of chromosome pairing so that in the region of the fiber attach- ment the chromosomes would be paired before the sister chromatids had sufficiently developed to prevent chiasma formation. Thus ‘rossing over would be increased in the spindle fiber region of the osome. ifferences in types of chromosome association at meiosis might also be attributed to differences in the region where pairing begins. Pairing begins at the spindle fiber and proceeds slowly toward the _ ends, the chiasmata would be localized in the region of the fiber, $x the case in F ritillaria (Newton and Darlington, 1930). If pairing * Completed before sister chromatids are sufficiently developed, "n chiasmata will not be localized, but will be more or less eat wily distributed along the bivalent chromosome as is the case n uss Lilium and Vicia. If pairing begins at the ends and pro- wus Slowly toward the middle, or if the sister chromatids develop Pein, then the chiasmata will be terminal, as found in the Solana- Graubard (1932) has recently presented evidence, based on cross- 208 JOURNAL OF THE ARNOLD ARBORETUM [vou xm | ing over in homologous chromosomes heterozygous for an inversion, ; which seems to indicate that pairing begins at the spindle fiberin the second chromosome of Drosophila. SUMMARY. A study of chromosome behavior at different stages of meioss in a number of species of plants, together with an analysis of the genetic evidence in Drosophila, has provided some critical evident | in regard to the cytological mechanism of crossing over. Janssen’s partial chiasmatypy hypothesis has been supported by Darlington, Belling and Maeda, but it is shown that theres no critical evidence in favor of this theory. Belling has offerd the only explanation of the mechanism of crossing over in conme- tion with Janssen’s hypothesis, but this explanation is not in accord with certain cytological and genetic evidence. Neither Belling’ ; theory of crossing over nor Darlington’s theory of meiosis can be reconciled with the cytological work of Robertson, Kaufmann ant | 4 arp. | According to the writer’s hypothesis crossing over is caused by breaks in two of the chromatids at a chiasma so that crossing 0" should be correlated with a reduction in the number of chiasmalé between the diplotene and diakinesis stages of meiosis. In Callisa repens there is a considerable reduction in number of chiasmata Pt bivalent between the diplotene and metaphase stages. The work other cytologists shows that the numbers of nodes or chiasmata reduced during the prophases of meiosis in T ulipa, Lilium, Prime, Rosa and Maitthiola. MER with crossing over is attributed to the penne asmata all of which are broken before diakinesis, resulting gon The ratio of different ty : -s and the types of second crossovers aB ~ centage of homozygosis in attached X chromosomes in Drosophilt deficiency of homozygosis. It is impossible to account f 102] SAX, MECHANISM OF CROSSING OVER 209 genetic results on Belling’s hypothesis or on any logical modifica- tion of his hypothesis. tion of chiasmata and crossovers is attributed to the type of chromosome pairing and to the relation between chromosome pairing and chromatid development. LITERATURE CITED ALEXANDER, J. ge yea Chemistry. Vol. 11, p. 1029. The Chemical Ca atalogue Co. Ni ork. ANDERSON, BG. hin 4). Mich and the Bibs 0 25 of non-disjunction in Drosop 523-5 ogaster. aa i indukt. Abstamm. u. Vererb. 51: a hae Sh - (1931). The constitution of primary exceptions obtained after A-ray treatment of Drosophila. (Genetics, 16: 386-396. REO , E. G. anpj Marcus M. Ruoapes (1931). The distribution of fess s ‘the X chromosome of Drosophila. (Papers Mich. Acad. | Baan . 2 W. W. (1030), Genetical and cytological studies of perenne Amira Be mays. (Cornell Univ. Agr. Exp. Sta. Mem. 129: 3-2: fe at, K too Die cytologischen Grundlagen der Ver hehe. rf 412. r Ee eee: ~ 928). Patio, of ecnce during maturation divisions aga jum and other plants. (Univ. Calif. Pub. Bot. 1 ” Bot <7 ; sao) meres of liliaceous plants. Univ ‘Cali. Pub. 16: 311-338 Uitins in flowering plants. (Univ. Calif. Pub. Bot. oy Non-disjunction as proof of the chromosome theory . grypeaiae 52, 107-163.) ERSON (1925). Crossing over in the X chromo- of Drosophila melanogaster. (Genetics, 10: {oe é R. M. Ousnrcur (1926). ‘The multiple “X-ple” and netics, 11: 4 eR. A. Brink (1931). Rat pews ical bag g: for Serduy gs bey between n non-homolog wipes spe biide (Proce. 334-338, ) 930). A cytological demonstration of crossing over. Lond, ser. B. 107: 50- [eiosis in diploid and tetraploid Primula sinensis. ). alae (Biol. Rev. 6: 221-264.) ( ages Translocations involving the third and the fourth of Droso; phila melanogaster. (Genetics, 15: 347-399.) pot rhgee of the second chromosome of Drosophila’ RAth ign uf Riese ad 671 observed in translocations co than Ne Nat. 65 65: 214-232.) is on je likeiceaeng bom Syn wag I. Fad peest # 3 Chromosome organization (Cytologia, 210 JOURNAL OF THE ARNOLD ARBORETUM wa: J Fixry, M. A. (1930). Phenomena of heterotypic ie in the p lia mother ary of a tetraploid form of Rumex scutatus var. typicus. ak . 50: 387-419.) re Fisk, (1927), ). The chromosome of Zea Mays. (Amer. Jour. Bot. use Gowen, a W. pe geld hee chromosome non-disjunction and the ger 11-212 . Genetic non-disjunctional forms in Drosophila. (Am Nat. 65: 193-213 .) a Gowen, M. Aer ag ge a Cop omnes nee Dr ‘at. 56: eager Ses (1932). Inversion in Drosophila melanogaster. Conti, 173 Haxpang, J. B. 8. (1919). The combination of mryrigs tidy es, and the ea lation of distances between the loci of linked factors. Goak Genet. I supra’ 4. F. (1930). Recent criticisms concerning Meiosis in Dre ogaster. (Science, 71: 241-243.) rey JANSSEN, Pr. et (1924). La chiasmatypie dans les insects. (La Cet eR cates B. P. (1926). Chromosome structure and its relation to the ie e cycle. I. Somatic mitoses in Tradescantia pilosa. epi oni Bot. 13: 59-80.) yy iter, ee Chromosome structure in its relation to the eycle. II. Podophyllum peltatum. (Am. Jour. Bot. 13: te nT . (1931). nemata in somatic and meiotic mitoses. wo Linpetnow, E a (1931). The genetics of Maize. (Bull Torr. Bat. Ch Manon T. (1980 a). The meiotic divisions in pollen mother cll of 7 minii e a one cells of Vicia faba L. L. (Mem Il. Sci, Ky egret Univ. 8: 125-180) Mavor, J. W. (19 924). age tel iy Peegeetaccion of aon auneees by X-rays. ie McCune, C. E. (1928). Differential chromosomes of Mecostethus i ad (Zeitschr. Zell. u. microse. Anat. 7: 756-778.) Bid. mere te! ‘The chiasmatype theory of Jannsen’s. (Quart, Rev. Bt Moncax, L. V., (1925). bla melanogaster wf oraan, L. V., (1925). Polyploidy in attached X’ chromosomes eee Ie 10: ias-1"3) ae ———.. (1981). Proof that Bar changes to n Bar by unequal er 511 over. Be Nat. Acad. Sci. 17: 20-272.) : fF efor ens Jour. Linn, § Sve | arses SAX, MECHANISM OF CROSSING OVER 211 _ Newron, W. c. F., & C. D, Daruincron (1930). Fritillaria 3 ormation and distribution. (Jour. Gen. 22: 1-14.) O'Mara, J. Siceaeses pairing in Yucca flaccida. (Cytologia, III. 66-76.) i hromosome of Droso- i ogaster. Layne 74: 647-648.) Pur, J. &C. L. .. Hsia aq thcby laxeghp ten cay veekayy 7 Br. ap anid in relation to the of double flowers. (Jour. PLov ne H. (1917). The effect of temperature on crossing over in Droso- Jour Zool. 24: 147-209.) Ranpourn, L. F. (1928). rete rac gh in Zea Mays L. (Cornell ni Univ. _ ta. Mem. 117 Reprie.p, 1930). Crossing over in the third chromosome of trip- loids of Drosophi melanogaster. (Genetics, 15: 205-252.) aygerven bes araces M. ‘por: The frequencies of ceigig factors in X females phila melanogaster 16; 375- 385) sega W. R. B. (1916). Chromosome studies. I. (Jour. Morph. 27: —. ai a). A split in osomes about to enter the spermatid (Paratettix texanus). (Goosen 16: Ho: 340-852) 8 ew b). Chromosome studies. IT. Tettigidae — ~evrtht to the presynapsis ny Othoar. Pitorsh. and . Si: candi 920). Genetic and cytological examinations of the or i Sacre Sasa is Dacieaiihe sadudaapeter. ( » 5: (Zeitschr. indukt. Abstam. u. Vererb. 41: 259-284.) Sane, 1. W, (1929). Structure of large somatic chromosomes. (Bot. Gaz. Sons, N (1936). On the spiral structure structure of chromosomes in some higher aa "CC (Mem. gy etn ceader a biatged 5: 239-245.) nthe ae: Srmx,C 1926). effect of temperature proning over in 12: 1931 -genetische destosdibbashes be als Beweise fir eis 10 } latin Uae (Biol. Zentralbl. $1: : Sn The effects of maagny creams eet nt Ge ber x locus in Drosophila. (Genetics, 10: 117-1 ) A527 age Sg’ A crossover reducer in Drosophila due to ee eo 02), a section of the third chromosome. (Biol. Zentralbl. 46: on. (1G. of the so-called mutation at the bar locus endtety ‘attached-X lines of Drosophila melanogaster, havior of heterozygous attached-X’s. (Pub. ) : and chromosome organization in Feat Stenobthrus) bas vuber and Trimerotropis suffusa. (Jour i, E. ee eet eo 1232. The . p. ; MacMillan Co, Sew , 212 JOURNAL OF THE ARNOLD ARBORETUM Yamana, G. & Y. Suvoto (1925). On the behavior of the nue cleolus | somatic mitosis of higher plants, with microchemical notes. (Bot. Tokyo, 39: 205-219.) 1, ZiRkKE, C. (1931). Brey of the root tip and cambium of Pinus strc (Cytologia, 105. . _ CYTOLOGICAL LamoRarony, ARNOLD ARBORETUM, Harvarp UNIVERSITY. DESCRIPTION OF PLATE 44. _ Meiosis in the pollen mother cells of Callisia repens. Magnification X¢ Figure 1. fons ese stage. No continuous spireme observed at any sta meiosis. Figure 2. lita pachyrtene The six chromosomes can be observed clearly at this stage. The nucleolus is always attached to one of the rete bivalents 7 Figure 3. eet nena showing the chromosomes forming the intern Figure 4. Diplotene ata | owing the number of nodes, most of which ate bably chiasmat a. Figure 5. Early “diakinss showing the reduction in number of nodes or paca with the diplotene stage. i Figures 6 and 7. 7. Dakine he nucleolus disappears between aia meta: Figure 8. Metaphase of the first meiotic division. The chromosomes shoW u the coiled chromonemata. Figure 9. Late anaphase showing spindle fiber rae no points. Tw m tachment while 10% uk ee fi achments. ext Figure 10. The nucleolus seems to be attached to the the same cart De all cases. This chromosome at late diplotene 1s Typical diplotene stages are shown 1n t! Osis Figure 11. The chromosome at the left is from a p. m. c. at early C8 Rages —. show the number ‘aa el | pachytene to a dakineeis and the reduction in the 2 ~ nodes or chiasmata between diplotene and cian e Jour. ARNOLD ARB. VOL, x1II. PLatTE 44 _ _MEcHAntsM oF Crossine OvER 1 Ae . 1982] ANDERSON AND DIEHL, TRADESCANTIA PROBLEM 218 CONTRIBUTIONS TO THE TRADESCANTIA PROBLEM EpGar ANDERSON AND D. G. Dien With plate 45 and three text figures CONTENTS Introduction Tradescantias of the St. Louis Region Tradescantia reflera Tradescantia bracteata Literature Cited INTRODUCTION Tue Genus Tradescantia is as taxonomically difficult as it is cy- tologically superb. On the one hand its fugitive flower parts render herbarium specimens a poor record of the living plant; on the other those very phenomena (fragmentation, polyploidy, ring-formation) which attract the cytologist, produce intricate and puzzling rela- _ tionships between species. It would remain a difficult genus to _ Monograph if its floral characteristics were as well preserved in | herbarium specimens as are those of the genus Aster for instance. me Monographing Tradescantia from herbarium specimens alone would b ed be almost like working with a collection of Compositae _ ftom which the flowerheads had been removed. Yet though the genus is an extraordinarily difficult one from the _ thodox taxonomic viewpoint, it has much to recommend it for % cyto-genetic studies. It grows easily from seeds or transplants and “0 withstand all sorts of mistreatment. It has a long-blooming _ Petiod and artificial pollinations are not difficult to make. The | _ “Tomosomes are enormous (they can sometimes be counted umicler : low Power) and are readily studied in smear preparations. Both ited Ploid and diploid species and varieties occur in nature, and White, blue, and pink forms are already available for genetical analysis, It is by all odds one of the most promising native Aer | Sn genera for cyto-genetic investigations. Yet fundamental bppehe 7 hs as to the evolutionary importance of cytogenetic OG : will not be possible until they can be viewed against a backgrou of sound pees mae Mop tt : ; Dev en nae bene syeenep ees researc. Tr adescantia ts who are studying 214 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm would record, in addition to cytological details, the exact source of their material, and the nature of the pubescence on the leaves, on the sepals, and on the ovary, it would be possible to apply their findings to other problems besides those of chromosome structure. TRADESCANTIA IN THE ST. LOUIS REGION The following paper reports a preliminary cytological and geneti- cal analysis of the Tradescantias of the Saint Louis region. It isa ‘report of progress” in an attempt to describe these species as they occur in that region, to evaluate the forces which have produced them, — and to measure the evolutionary processes which are taking place within them at the present time. Figure chrom, he . : eral Habit of 2 of r pions (ef el r, " beostows (ualddle), ned Tr fsa ea cone : 1932] ANDERSON AND DIEHL, TRADESCANTIA PROBLEM 215 There are at least three species of Tradescantia in eastern Missouri. There may be more; there cannot possibly be fewer. If there eventually prove to be ten or twenty species they will even then fall into three groups. These three groups are separated from each other morphologically, ecologically, and cytologically, We are identifying them provisionally as Tradescantia pilosa Lehm., T. reflera Raf., and T. bracteata Small. There are other Tradescantias in the state but we did not find them in the area covered by this study. The outstanding differences between the three species are shown, somewhat diagrammatically, in figure 1. The middle row shows camera lucida drawings of the ovary and its pubescence. The lower row, reduced to scale from actual specimens, shows leaf size and number, length of node and number of nodes, etc. The upper row shows the gametic chromosome complement from camera lucida drawings of smears of pollen grain mitoses. Tradescantia reflexa Tradescantia reflexa is by far the commonest of the three species. It is tall and slender, glaucous throughout, with narrow, reflexed bracts. The ovary is completely glabrous or at most bears two or three tiny hairs at the base of the style. At Eureka, Missouri it was found on the limestone “‘glades”’ of the region. These are dry, stony hillsides with a semi-arid flora. They were apparently in this area the original home of the species, from which it has spread to tailroad rights of way, dump-heaps, and the like. There was no apparent morphological or cytological difference between these glade” populations and those along the railroad tracks. Several of the colonies included variants, which had they not been connected with the normal type by a complete series of ened mediates, might have been considered as taxonomically distinct. At Algonquin Station, Webster Groves, there were several very dwarf 4 lants less than a foot high, but there were also a number of inter- Mediates between these dwarfs and the rest of the colony. At cL Hamburg, Missouri, the opposite extreme was found, for several of 7 er Plants of T. reflera studied. For the present we hav: SIpr8 Mm under T. reflera. It is quite possible that they may Tesult of hybridization with T. bracteata, In all of the above — > __ the plants in this colony measured well over six feet in height. The : eps outstanding variation was seen at Hillsboro, Missour!, along a : tania outcrop. The plants of this colony were all characterized | ti et hairs on the calyx in addition to the normal tuft usually : . iaieke T. refleca. They were also differentiated by being early- — Wering and it is possible that they are varietally distinct fro oy ne e included 216 JOURNAL OF THE ARNOLD ARBORETUM [vou xm — these peculiar forms were examined cytologically (both PMC and | pollen divisions) and showed no divergences in chromosome num- bers. Nineteen plants were recorded, as shown in Table 1. All — were tetraploids (2n = 24). Many other plants were examined, and their chromosome numbers determined, but no definite record was made. Unfortunately the only plant showing a fragment chro- mosome was among these unrecorded individuals and it is not — possible to say from which locality it came. In spite of its variability, 7. reflera remains an easily recognizable unit over a very wide area. Colonies in south-western Michigan (Schoolcraft) and in southern Illinois (Ullin) seemed essentially the same as those examined in Missouri. The following collections available in herbaria represent plants morphologically similar to those we studied cytologically: Lansing, no. 3121, Mansfield, Mo; — Lansing, no. 3037, Cedar Gap, Mo.; Ridgway, no. 2116, Olney, Ill xamination of reduction divisions showed that J. reflera Was not only a tetraploid but was practically an auto-tetraploid. There was a very strong tendency for the chromosomes to conjugate in sets of four, the number of quadrivalents per PMC varying from 0 to 6. A count of ten PMC from a single smear of T. reflexa gave the fol- lowing frequencies: No. of times observed No. of II’s No. of IV’s 0 6 bt OS = 20 tO OD & © = 2 OS He Or 10 _ In the material examined the separation of quadrivalents was not iq at random. There was a very strong tendency (about 10 to 1) eh | adjacent chromosomes to pass to opposite poles. The data ey ; consistent with the hypothesis that T. reflexa is an auto-t aploid 4 in which there is a slight differentiation between the two diploid : sets of which it is made up. Its chromosome complement might be — diagrammed as follows: | - AA’ BB’ CC’ DD’ EE’ FF’ A’ BB’ CC’ DD’ EE’ FF’ _ Tradescantia bracteata Hoe As it occurs in Missouri this is a dwarf species seldom over a foot : (3 dm.) in height. The ovary is covered with short glandular ‘ (fig. 1). The flowering period is comparatively short and geet seed ripens the leaves die down and the plants pass thre ie : summer in a semi-dormant condition. The flowers are More” ail oie list 1932] ANDERSON AND DIEHL, TRADESCANTIA PROBLEM 217 liant in color than those of T. pilosa or T. reflexa. Pink-flowered plants and blue-flowered plants are usually found growing together and albinos are not uncommon. The following collections represent material essentially similar to that which we studied cytologically : Bush, no. 684, Watson, Mo.; Bush, no. 336, Grandin, Mo.; Davis, no. 4403, Whiteside, Mo. Tradeseantia bracteata grows in abundance on rich black soil along roadsides and in wet meadows in the bottomland of the Mississippi River at Portage des Sioux, Missouri. The chromosome numbers of six plants from this locality were determined as is indicated in Table 1. All were diploids (2n = 12). One had a small fragment chromosome. Tradescantia pilosa This is by far the most distinct of the Missouri Tradescantias. In the St. Louis area it occurs only in shady situations, usually in rich sandy soil at the base of sandstone cliffs or in pockets on the face of the cliff. Since sandstone areas in eastern Missouri are mainly confined to a long narrow outcrop of St. Peter sandstone, - pilosa is an uncommon species there. It was found in abundance TABLE 1 Chromosome Numbers of wild-growing Tradescantias from eastern Missouri. (Including a few plants of 7’. reflera from outside that region.) PMC Potten GRAIN MITOSIS = 42 n* eRe 3 n= 12 = 12 n=12 pene n= 8 0 n=12 a. n=2+f ait n=1 aa n=12 og n= 12 ail n= 2 me n= ale n=2 =< 19 n=2 any SC a ie ie n=12 Re n=112 ble n+ 12+f n=! nate a n=6 bag n=6+f a n= a n=6 LE n=6 <3 n= 6 218 JOURNAL OF THE ARNOLD ARBORETUM [vou xm at Marthasville, Hermann, Ashland, and Winfield, Mo. The fol — lowing collections, available in many of the larger herbaria, repre _ sent material essentially similar to that which we studied cytolog- — ically: Eggert, St. Louis, Mo.; Davis, no. 3604, Hannibal, Missouri; — Eggert, Hematite, Mo.; Palmer, no. 34,802, Pontiac, Missouri. ’ In general aspect T. pilosa is entirely different from any other Missouri Tradescantia. The stem is tall and zig-zag, the entire — plant is sparingly pilose. The flowering season is late (July to August). The ovary bears long scattered hairs with relatively — small glands at their tips. (Fig. 1). The chromosome numbers of 4 plants were determined; as reported in Table 1. All were tetra- ploids (2n = 24). . A COMPARISON OF VARIATION IN DIPLOID AND TETRAPLOID SPECIES As has been reported above, T. reflexa (in Missouri, Illinois, Wis- consin, and western Michigan) was found to be a tetraploid species. T. bracteata, on the other hand, was a diploid. That is, 7. bracteata like most normal animals and plants had its chromosomes in sets of i twos. Those of T. reflexa on the other hand, were in sets of fours. — This tetraploid condition should have a very marked effect upon the He nature of individual differences in the two species. It should m- crease not only the proportion of intermediates but the number of — intermediate types. An example may make this more clear. Let us consider the simplest possible case, a single factor differ- k ence, albinism, for example, as it might be expected to operate the diploid T. bracteata, on the one hand, and in the tetraploid, 4 YT. refleza on the other. The inheritance of albinism in Tra o) so far as we know, has not actually been studied but the circum: stantial evidence from forms existing in nature is all in acco opie | the hypothesis that as in practically all other flowering plants wn i: due to a single recessive gene. If we represent the gene for albinis™ by a, and its normal allelomorph by A, an albino plant of 7. bracteala : will be of the genetical composition (aa) and a pure-breeding full * colored plant will be (AA). Crossing the two will give us a hetero” — | f j ; 8 It is Particularly interesting that the tetraploid Tradescantias tould be more northerly than the diploids. Sax (1931) has reported : in the closely related genus Rhoeo, the artificial production of tetra- rt _ Ploi # by exposure to low temperatures. A similar geogta 18 | Position for tetraploid races and species to the north of their diploid < | relatives has been reported for a number of genera. Hagerup Sacoogs é collected six such cases in the Bicornes alone and has recently a5] 7 “arized the evidence on polyploid geographical races ess eed _ . Mularly interesting that Mangelsdorf and Reeves ee eel By with another American monocot of tropical affili i adeant ae dactyloides) have found that the plants collected a ean ef “Gs while those from the north and east are tetrapl 430: 228 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm It may be well in passing to point out that the differences between the seven diploid species from Texas are quite as great, on the whok, as are those between the three Missouri species. Polyploidy her, as elsewhere, has introduced complexity into inter-specific relation- ships, but species differentiation has taken place to an even greater extent in regions where polyploidy was absent. Table 5 summarizes the outstanding differences between these species. Camera lucida outlines of their ovaries are shown it — figure 8. While instances of inter-specific hybridization are not u- known, most of these Texas species are kept apart by habitat differences and maintain themselves as recognizable units over® — wide area. 2 TABLE 5—A TABULAR COMPARISON OF SPECIES DIFFERENCES IN TEXAS AND MISSOURI PUBESCENCE ON Sprcres Stem FLorRALLeEAvesS LEAVES SEPALS Ovary TEXAS SPECIES: T. texana short, weak long, equal long, vil- dense, a few glandular branched - lo . glandular hairs at top T. gigantea tall short, sub- glabrous dense, non- very dense, equal, dense- glandular non-glandulat | ly pilose me T. hirsuticaulis tall short, ual hirsut sub-gland- glandular # | uneq irsute ub-é on gnc i, T. humilis i short unequal hirsute glandular ine glant: T. occidentalis slender slender glabrous _ scattered glandular and q glandular non T.sp.(refleza?) medium _—long, unequal glabrous glabrous _ glabrous except for ise ib A tuft at apex T. reflexa ce to long,unequal glabrous glabrous glabrous except for tuft T. bracteata short Vers iy andular .. Gente | ery long, grea ica glandular sts rd subequal glan glandular T. pilosa tall zig-zag sub-equal Lensad glandular scatter pilose pilose glandulat CONCLUSIONS i It should be remembered that the following conclusions are | et ; more than working hypotheses and that they are put eek : tentatively at the end of our first year of intensive work. In we ning this ony! we had as our objectives (I) the ae species of T. a as they occur in nature and (II) the¢ at tion of the evolutionary processes whic are taking yi ae : the —— time. 1982] ANDERSON AND DIEHL, TRADESCANTIA PROBLEM 229 I. As regards the description of these species and their separation and classification we feel that they are a difficult group but by no means an impossible one. Their inter-specific relationships are not nearly so intricate as are those of such genera as Rubus and Crataegus, for instance. In this connection we have found the pubescence on the ovary a particularly useful character because it varies so little within species. A colony of Tradescantias may vary strikingly in size and general aspect from plant to plant and yet the pubescence on the ovary will be the same throughout the colony. The pubes- cence also varies widely from species to species. It may be dense, _ orsparse, or restricted to one part of the ovary, or completely want- . ing. The hairs may be long or short, and glandular or non-glandu- lar. Used in connection with other characters it is very helpful in _ Working out specific relationships. - In evaluating the evolutionary processes which are taking | place at the present time, we have evidence on three, fragmentation, _ polyploidy and hybridization. FRAGMENTATION, th every species in which we were able to examine a number of | different plants, we found individuals with supernumerary fragment : chromosomes. That is, in addition to the normal chromosome com- . plement for the species, these individuals had one or two fragment romosomes, much smaller than the rest (Plate 45, figs. A, K). In at least two cases these fragments paired regularly at the reduction "sion and were distributed to all the germ cells. We found frag- _ _yats occurring with roughly the same frequency in all the species re which we investigated. If, as seems probable, they affect the ex- PS ternal morphology of those plants which bear them, we have here a _ Mique case in which one of the causes of variation within species is : hot itself effective in forming new species. Had it been so we should hs i d entire species or races which were characterized by the Possession of supernumerary chromosomes. In these Species of Tradescantia polyploidy is apparently intra- With consequent division of those species possessing it into and tetraploid races. It apparently allows a northern ex- “ the range in those species in which it has occurred. It . : manyfold the variation between individual plants. Its ” __ @ffect upon variation in flower color can actually be “ated and a similar effect upon morphological characters 1s from the peculiar variability of the tetraploid species, T. Tf. virginiana, In the section of the genus which we have — originating new species though it multiplies the complexity of i intr. | specific and intra-specific relationships. rs HYBRIDIZATION. Although this undoubtedly occurs we have as yet found little — actual evidence for it. The colony from Hillsboro, described above, may perhaps have resulted from previous hybridization between T. reflexa and T. bracteata. An apparent example of hybridization z between T. humilis and T. reflera has just been discovered] in the © vicinity of Austin, Texas. SUMMARY Three species of Tradescantia are common in the region ale Louis, Missouri, two tetraploid species T. reflexa and T. , one diploid species, 7. bracteata. White-flowered and pink-flowe forms are frequent in T. bracteata while in the two te rap species they are rare. Furthermore various intermediate magenta ) shades are common in the tetraploid T. refleza but are not foundin T. bracteata. This is shown to follow logically from the fact . T. reflexa is practically an auto-tetraploid. Darlington’ s assumption of highly developed vegeta rem duction in tetraploid Tradescantias is found to be wi foundation in fact. The inter-clonal variation of a single colony is presented in Seven species of Tradescantia from eastern Texas were f be diploids. The eickptiateiry importance of fragmentation, polyploidy: hybridization is briefly discussed. ACKNOWLEDGMENT Much of the work reported in this paper was carried on at 7 Missouri Botanical Garden. ‘The authors are indebted 4 eet oo Dr. George T. Moore, for support and ener LITERATURE CITED -Daruixeton, C. PD. Chromosome behaviour and structutal e Tradescantiae. (Jour. | Cenbk 44 207-286. 1929.) Hacerup, O. Morphological and cytological studies of Bicornes- | Bot. Arkiv, 6 (1), 1928). | see O. Ueber Polyploidie in Beziehung zu Klima, Oekok onl * a E i B® <— oe Jour. ARNoLD ARB. Vou. XIII Puare 45 Piette ees ee 1 2 See a Pee SF et CHROMOSOMES AND Ovary Harrs or TRADESCANTIA SPECIES 1932] ANDERSON AND DIEHL, TRADESCANTIA PROBLEM 231 Rosz, J. N. Three New Species of Fa epee from the United States. (Contrib. U.S. Nat. Herb. 5: 204-206. Sax, Kart. omosome ring formation in Rhoeo discolor. (Cytologia, 3: 36-53. 1931.) ARNOLD ARBORETUM, Harvarp UNIVERSITY. EXPLANATION OF PLATE 45 7 ic ers : cervinus, Russula roseipes, Tricholoma album, Boletus sea ‘a 1 : : seen speciosus, Merulius tremellosus, and Polyporus betulinus. It 1922] HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES to ~) ~1 strange that Polyporus betulinus, although much firmer in texture than other members of the above group, is hardly surpassed as a host for Physarum polycephalum. The fungi which were attacked more slowly, but on which the plasmodium fed for a comparatively long time, are: Fomes applana- tus (Plate 49, fig. 10), Lenzites betulina, Polyporus adustus, P. fron- ate 2: : A. Sporophores of Boletus subtomentosus twelve hours after inocu- tipe with Physarum polycephalum, showing the plasmodium eroding the i 4 stroy ying the adjacent pores of the f ruitbody on the left, and the ok Over the stipe of the one on the right. B. Twenty-four hours later rus at the left is a soft disintegrating black mass. The pores of the 4 $ n the right are shout to be attackes * . o* te 4 . : 5 >: > (Plat + hirsutus, P. pargamenus, P. resinosus, P. Schweinrtzu ate . vo: Phy 8, figs. 3-4), and P. versicolor. The sporophores which AT hi um polyce phalum attacked virulently, but ofte ntimes from Cc » the Plasmodium moved so sluggishly that when putrefaction 278 JOURNAL OF THE ARNOLD ARBORETUM [vou xm began it disintegrated with the host, are: Collybia succosa (Plate : 48, figs. 1-2), Flammula polychroa, Hypholoma appendiculatum, H. velutinum, Mycena alcalina (Text-fig. 1), Pholiota praecor, Pleurotus sapidus, Psalliota campestris, P. placomyces, P. Kod- mani, Russula fragilis, Boletus felleus, B. subtomentosus (Text-fig.2), Strobilomyces strobilaceus, Polyporus sulphureus, Hymenochaete badio-ferruginea, Hydnum repandum, H._ septentrionale, Hydnel- — lum scrobiculatum, Clavaria aurea, C. fusiformis, C. Kunzet, and Tremella lutescens. The plasmodium of Physarum polycephalum failed to attack Claudopus sp., Clitocybe pithyophila, Pleurotus ulmarius, Tricholoma personatum, Poria corticola, and Lycoperdon sp. Positive statements regarding the intensity of the plasmodial attack upon a certain fungus host are difficult to make, since the maturity and condition of each sporophore tested may make @ difference. In some trials Hydnum septentrionale was not attacked after inoculation with the plasmodium of Physarum polycephalum, while in other trials the plasmodium consumed almost the entire sporophore. The same was true of Amanita phalloides and of Russula emetica. A good example of the effect of the maturity of the sporophore upon the parasitism of the Myxomycete was shown — by Physarum polycephalum upon Coprinus micaceus. If the plas- modium was placed upon the gills of a young pileus, the plasmodium grew rapidly at the expense of the gills and trama, but if the gills _ had changed in color from white to reddish purple or had begun | to deliquesce, the plasmodium was killed. Physarum tenerum Rex. rapidly attacked Merulius tremellosus, — but very slowly attacked Coprinus atramentarius, Lepiota brunntt, — Polyporus adustus and P. pargamenus. : Physarum virescens Ditm. was unable to attack Coprinus | micaceus, but made a slight growth upon Collybia succosa. Fuligo septica Gmel. failed to attack the sporophores of Clitocybe multiceps and Clitopilus prunulus when the typical yellow plas: modium was used as inoculum. When two white plasmodia, - whose aethalia formed in cultures have been identified as F ulig? : septica, were tested on fungi, they destroyed the tissues of somé fruitbodies. The sporophores of Amanita muscaria, Pi leurotus 4 petaloides, Tricholoma album, Boletus scaber and M erulius treme” — losus were rather rapidly attacked, while those of Amanita phallowdes, 2 -A. verna, Lactarius volemus, Lepiota naucina, Pleurotus ostrealusy Boletus speciosus, Polyporus Schweinitzii, and T remella lutescen* were attacked more slowly. The fruitbodies of the followine fungi were not attacked: Hygrophorus Peckianus, Russula emetic® 1982] ‘HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES — 279 Polyporus betulinus, P. sulphureus, Poria corticola, and Hydnum septentrionale. _ Leocarpus fragilis Rost. increased slightly in size upon sporo- phores of Collybia succosa, but fruitbodies of Coprinus micaceus, ; Mycena galericulata, Fomes applanatus, and Trametes pini inocu- > cet with the plasmodium were unattacked. ‘ e - Brefeldia maxima Rost. made a slight growth upon Collybia _Iygrophoroides and Fomes applanatus, but none on the pilei of - Coprinus micaceus. Lindbladia effusa Rost. failed to grow on Mycena galericulata and Coprinus micaceus. Lycogala epidendrum Fr. made a slight growth upon the apothe- Gum of a Peziza and upon the mushroom Mycena galericulata, ‘ bod It failed to attack Mycena Leaiana and Coprinus micaceus. pe tichia clavata Rost.; the plasmodium attacked sporophores "i Mycena galericulata and made a fair growth upon them, but it failed to attack Collybia hygrophoroides, C. suecosa, Coprinus — Micaceus, and Mycena Leaiana. Trichia decipiens Macbr. grew quite rapidly at the expense of the es of Mycena Leaiana, grew slightly upon Mycena galericulata nega hygrophoroides, but did not attack Coprinus micaceus. eg sp. seemed to attack Hypholoma sublateritium tat proved i innocuous on Pleurotus ostreatus, P. petaloides, ipidus, P. ulmarius, and Poria corticola. unidentified yellow plasmodium (BFU) proved to be ex- 4Y parasitic and rapidly destroyed Coprinus atramentarius, 1a sublateritium, Lepiota brunnea, Pleurotus ostreatus, emetica, Boletus scaber, Merulius tremellosus , Polyporus _P. adustus, Poria corticola, Daedalea confreqan and ochaete badio-ferruginea. number of Myxomycetes whose Leapaary utilize fungi for from being kn 7 a one species of plasmodium may P utricularis and Physarum polycephalum are known beg. at some species are very general parasites. Other ery gi. More resistant tissues of the eet Folypores ut we may safely conclude from the long list alae fungithat => 280 JOURNAL OF THE ARNOLD ARBORETUM [VoL. xm i account for the slower attack by plasmodia upon them. The plasmodia move over the surface of the pores digesting the basidial _ layer and sometimes completely destroying the pores, as seen in Plate 48, figs. 3-4. As a contributing factor of great importance — in the destruction of the sporophores, it should be noted that all the observed plasmodia remove the surface layers and leave the — Polypore with a moist slimy coating; a very favorable substratum — for the development of Mucors, bacteria, and other secondary invaders. This parasitism of Myxomycetes on fructifications of fungi is” not confined to the laboratory, for during one season (1931) twenty- two plasmodia were found in the forest apparently feeding on the — fungi with which they were associated. Birch and maple logs, — decayed by various Polypores and Agarics, proved to be a favored haunt of the majority of these plasmodia, but three were found — destroying fungi on an oak stump, one on a dead standing Chestnut — (Castanea dentata), and another on a dead standing Pine. Although 3 field observation of the parasitic habit of plasmodia first arouse? — our interest in this problem, study in the laboratory has given US — some details of the method of attack. DISCUSSION OF THE METHOD OF ATTACK nee; Microscopical examination of both living and sectioned material — shows that the attack of plasmodia upon all of the sporophores — examined to be both pathogenic and parasitic; pathogenic in that : the plasmodium may cause necrosis of the living fungous tssue — with which it comes in contact, and parasitic in that the fungous — tissues are digested and then absorbed to furnish nutrients for the Mi, oars ih a power of the plasmodium which is supposedly able to dry UP of | kill the mycelial filaments without digesting them. So : ¥ ‘ evidence found by the authors to oppose this view 1S: 1, ers hyphae with no signs of bacteria surrounding them can actually? seen to dissolve throughout their length when covered by the ei modium, 2, stained sections (Plate 48, figs. 6-7) have failed tore 1 be bacteria digesting the hymenial layer in order that the plasmoe” Ft nae ia el oe all of yeasts, which have been found associated with an actively oe 4 sitic plasmodium, does not become mycophagous when beg be) opportunity to digest fungous material. : ‘ 432] HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES — 281 The process of destruction appears to be of a chemical nature, asis evidenced by observing the individual hyphae in contact with the plasmodium gradually lose their sharp outlines and become dissolved. The time required for the dissolution of the fungous tissue seems to vary with the digestive secretions of the plasmodium, but no special study of this digestive process was made. The nature of the hyphal wall appears to play an important role in susceptibility to plasmodial attack, as is shown by their attack on hyaline hyphae more readily than on colored ones. No micro- chemical tests of the hyphal walls have been made with which _ this specificity could be correlated. The thickness of the fungous _ wall likewise appears to be directly proportional to the length of _ time necessary for its digestion by the plasmodium, as is shown _ by thin-walled hyaline hpyhae of the hymenium actually dissolv- _ away within a few seconds after coming in contact with the ad- j -Yaneing margin of the plasmodium, while thick-walled hyphae _ tither may require several minutes for their dissolution and may be found several millimeters behind the advancing margin before they break down, or they are not digested but are left as debris. In order to dispel any possible idea that digestion occurs in 3 advance of the plasmodium, it is emphasized that the tissues to - be digested must lie in close contact with the plasmodium. This _ Contact, however, may be either with the surface membrane or S with 4 vacuolar membrane. Although digestion of hyphae in ng is a possibility, it could not be definitely demonstrated in et Stained preparations of plasmodia parasitising sporophores, but the hyphal tissues appeared to be digested without engulfment ‘the ventral surface of the plasmodium (Plate 48, fig. 5). Phalum attacking the lamellae of Hypholoma sublateritium examined (Plate 48, figs. 6—7). The smooth, oblong-elliptical, -brown spores, measuring 3—4 x 6-7 microns, may be clearly closed in vacuoles in the matrix of the plasmodium. In rt te figs. 6-7, the spores are shown near the ventral surface Plasmodium and the nuclei near the dorsal surface. ied. ayy is highly vacuolate in its ventral portion where it is ting nutrients. Very large, dark-colored, t ss are many times not ingested but merely pushed aside meally by the plasmodium. ; ®ugh Lister (8) and others report that fungous tissues, nd other solid ingesta are engulfed and move the pro- 282 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm : toplasmic currents of the plasmodium, the authors have found | that the plasmodium feeding upon a sporophore carries on practi- cally all of its digestion of hyphae and spores close to the advancing — margin. Only occasionally what appears to be a bit of hyphaor | a spore may be seen coursing through the plasmodial strands — back of the advancing edge. 2 A study of sections of the pileus of Plewrotus serotinus attacked — by the plasmodium of Physarum flavicomum shows the gills be- _ coming more and more eroded until there remains only a small — shapeless mass of indigestible residue. In Plate 48, fig. 5 the — basidial layer has completely disappeared and the tissues under- neath the plasmodium are being dissolved. There is a sharp line — of demarcation between the plasmodium and the fungus. As soon as the basidial layer is dissolved on one gill, the Myxomyeete moves to that of the next. : SUMMARY Attention is called to the ability of plasmodia of Myxomycetes- to parasitise the fruitbodies of common wood-rotting fungi. Some species of Myxomycetes have been shown by field observations and laboratory tests to be very generally parasitic upon Hymeno — mycetes, while other species are moderately so, or as yet, NOR — parasitic. In the laboratory about eighty-five species of the z: Hymeniales, ranging from fragile fleshy Agarics to hard woody Polypores, were inoculated and most of them were found to be attacked by one or more of the thirty-three different plasmodia used in this study. : .; Plasmodia of the following species are now known to be my" cophagous: Badhamia foliicola, B. magna, B. rubiginosa, B. utr- cularis, Brefeldia maxima, Fuligo septica, Hemitrichia clavalt, Leocarpus fragilis, Lycogala epidendrum, Physarum flavicomum, © polycephalum, P. rigidum, P. tenerum, P. virescens, and Tridut— _ decipiens. Many other plasmodia, as yet unidentified, exhibited : various degrees of parasitism. Unaccounted for preference he : some fruitbodies is exhibited by the plasmodia. Mushroom poisonous to man are readily parasitized by plasmodia, but - i maturity of the sporophore does affect the parasitism of the Myxr 7 mycete. ae peel eek: The plasmodia bring about the destruction of the sporophor® _ in two ways: 1, by actively digesting the sporophores themselve> and 2, by leaving the moist, slimy, injured fungous tissues, fh which they have passed in an ideal condition for infectio® . bacteria and fungi which carry the destruction further. ! ; i : - 1989) HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES = 283 oa The process of destruction of the sporophores is one of digestion, _ the basidial layer being usually the first tissue attacked. Hyphal - tissues are digested at the highly vacuolate ventral surface of the plasmodium without enclosure within the vacuoles, while the ; spores are usually ingested and then digested within the vacuoles. i LITERATURE CITED Bs sa-% A. Clastoderma, novum Myxomycetum genus. (Bot. Zeitung a ae ccs, Fr, Ilanées mycologiques, sér. 2. (Actes Soc. Linn. de hi ae nr aseoy 211-273. 1891.) 3, Currie, M.E. A critical study of the aS of Ontario. ( Royal Canadian Inst. [Toronto] 12: cakes t 8-10. 1919.) Eunrorr W. T. Some observations upon Be gra hevnes of fone) ii utricularis Berk. (Trans. Brit. Mycol. Soc. 5: 410- im are & Jessre 8. Extiorr. The sequence of fungi and _ Myce 208. our. Bot. 58: 273-274. 1920.) 6. oa. J. W. Observations upon the — plasmodia of _ _ Fuligo septica. (Bot. Gaz. 31: 198-203, fig. 1. ¢ 7. Jann, E. Myxomycetenstudien. 2. Arte nm aus plumenae (Brasilien). - . _ (Ber. Deutsch. Bot. Gesell. 20: 989-20, Taf. 13. o@ A. Notes on the reper i f Badan utricularis and ; Brefeldia maxima. (Ann. Bot. 2: 1-24, ap 1888. 8 “ap tino) T. H. The Slime “Moulds Pthodors, 2: 75-81, pl. 16. et ee eee eee eases - (Se em Fasor, P. E. Sur la germination des spores, sur la nutrition et sur la | chez les Myxomycetes. (Compt. Rend. Acad. Sci. Paris, ree 173: 50-51. 1921. Fores Notes on Malayan Mycetozoa. (Trans. Brit. Sy ‘# 2. On the parasitic habits of the plasmodium of Physa- ., Tum viride var. rigidum Lister. (Trans. Brit. Mycol. Soc. 7: 299-300. Fig. 8. Fig. 9. prin eee at Fig. 10. A sporophore of Fomes applanatus map Se hours after mee the y shia eget JOURNAL OF THE ARNOLD ARBORETUM EXPLANATION OF PLATES 48 AND 49 Fruitbodies of Collybia apa’ ‘ean had Hees inoculated with — Physarum polycephalum twelve hours previously. The to sporophores on the left are held | erect mak: pins due re des uction " of they 86 stipes by the plasm: whith the tissues are rshietdey (ieee the ee eae : marcation between the darkly stained plasmodium Ai te tissue. pecien cut 10 { u thick and stained with Flea — e. ; Photomicrograph of the plasmodium of Physarum arum polycephalum — shown digesting the cated and iugeoting the spores from & gill of Hyp. yee #] on one side of the Sl fas toes. dipsctod but the menial Ie penne tee ghaae of the gill is shown at the bottom of the picture. Photomicrograph of another portion of the sme m and ol ; shown in Fig. 6. Here is shown a bit of the ‘ with iron-alum-haematoxyiin. eet et 3 dale, ‘a al Pleurotus serotinus inoculated with the plasmodium of Phyo Trent four, hours later the plasmodium has destroyed the gil, and saprophytic PNP ir pee y The fluffy white euctads layer of encahf is being lowing the eroded brown wn pore surface to show throug. PLATE 48 Jorn. ARNoLD ArB, VOL. XIII HORES POROP 5 P ARASITISM or P LASMODIA ON Jour. ARNOLD Ars. VOL. XIII PLATE 49 PARASITISM OF PLASMODIA ON SPOROPHORES 1982] CHESTER, THE PRECIPITIN REACTION IN PLANTS 285 STUDIES ON THE PRECIPITIN REACTION IN PLANTS Il. PRELIMINARY REPORT ON THE NATURE OF THE . “NORMAL PRECIPITIN REACTION” KENNETH §S. CHESTER Iv 1928 Kostoff first called attention to the fact that the aqueous extracts of the foliage of certain Solanaceous plants precipitate in the presence of certain other such Solanaceous extracts (3). This _ precipitating action was designated as a “normal precipitin re- action,” and the modification of such reactions after intergrafting _ underlies in part Kostoff’s theory regarding acquired antibody pro- _ duction in such plants. In 1931 the writer published a record of his tests of the “‘normal precipitin reactions” in a number of species of the Oleaceae (1). These tests showed that the species of Oleaceae ‘ested were in no case interreactive with the exception of nu- _ Merous varieties of Syringa vulgaris which were suffering from a _ Physiological blight due to graft incompatibility. Such varieties _ feacted strongly against a number of other Oleaceae including healthy, ungrafted plants of the same varieties as the reactive - Plants. Meanwhile Silberschmidt in Munich had also been studying the Kostoft phenomenon from the standpoint of technique. Silber- _ Schmidt has published an extended piece of research dealing _ Manly with improvements of the methods of extracting and testing ? 4), and he intended, at the time of publication, a continuation oS of his work dealing with the actual experimental results of his a tests, Early in 1932 there appeared a second paper by the writer (@)sh wing that the “normal precipitin reactions” between species * the Rosaceae, Saxifragaceae, and Caprifoliaceae are well cor- related with the systematic positions of the species considered. ; _ Thus far little attempt had been made to determine the bio- ical nature of the reaction. Kostoff and Silberschmidt both that the phenomena are of protein nature, and that they logous to the phenomena of animal serology. In 1952, , the writer stated that: ‘‘Some of the reactions, such . Prunus, Ribes, and Robinia, lead one to the suspicion Possibly some non-specific compound is acting In & rather X fashion to produce the precipitates” (2, page 7))- ihe of the chemical nature of the “precipitin reaction Y Open to investigation and was of utmost importance 1 reting and evaluating the results of the earlier experiments. pathological laboratories of the Arnold Arboretum s have accordingly been in progress during the past 286 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm winter to determine the chemical nature of the precipitating 3 property of the extracts. As these investigations have been fruitful — and as the results are of importance in directing future activities in this field, the present paper is designed to give a condensed account of these results, while a more detailed description of the work will appear in an early number of this Journal. | In the fall of 1931 collections of leaves of certain species of woody — plants were dried and pulverized. These relatively homogeneous stock supplies were used for all the tests of woody plants deseribed | below. The technique employed was in all important particulars ) the same as that previously described (2). Among the species — selected were Prunus Armeniaca var. ““Mikado,”’ Platanus acerifolia, Le curt Robinia fertilis, and X Ribes Carrierei. Of these four the Prunus | vf a l tests strongly against the other three which latter are mutually inter-negative. The results obtained from a study of these four — species were later found to apply to extracts of Hydrangea pan — culata var. grandiflora, Syringa vulgaris, Ligustrum obtusifolium, — L. ibota, and L. vulgare var. foliosum. Finally, the findings wert — applied to thirty-five species of Solanaceae, and accordingly the — results here recounted have been gained from a study of practically ; all of the possible inter-reactions of forty-four species of herbaceous § ‘and woody plants. As a working hypothesis the theory was first entertained that i the reactions might be of protein nature. Accordingly the earlier f part of the investigation dealt with an analysis of the reactions 0” the basis of such an hypothesis. However, the results obtained i from an analysis of the Prunus-Robinia-Platanus-Ribes reaction ‘ showed a very peculiar behavior if such an hypothesis were correct. : The findings in this connection are here enumerated: , r 1. Variation in the salt content of the extracts by the use of Cohn’ a phosphate buffers at constant pH of 6.0 had very little ret5) on the reactive potency of Prunus between the limits of #4 fe and 1.2 M. Between the same limits, however, there was # gradual fall in reactive potency of Platanus and Rohima ae : the weaker to the stronger salt concentrations. (i : 2. Variation of the pH of the extracts by the use of Cohn’s pi" ; phate buffers at constant salt concentration of .06 M et phe effect on Robinia between the limits of 5.2 and 8.4, and also te effect on Prunus between the same limits except for # ede decline from 8.0 to 8.4 and from 5.6 to 5.2. ‘i ving at : of - 3. Long continued heating (even to three hours autocl : 5 lbs. pressure) did not decrease the reactive strength oF @ a 1932] CHESTER, THE PRECIPITIN REACTION IN PLANTS 287 _ 4. Precipitation of the extracts of Prunus or Robinia by excessive alkalinity or acidity, followed by filtration and subsequent neutralization did not remove the reactive principles. 5, There was no significant effect on the strength of the reaction if the pulverized leaves were thoroughly extracted with strong alcohol, anhydrous ether, benzol, or carbon tetrachloride as a preliminary to extraction with water. _ 6, There was no diminution in reactivity after storing the extracts at 2° C. for as much as four months. Even bacterial or fungous contamination had little or no effect on the strength or specificity : of the reaction, when the extracts were subsequently cleared. __ 1. The progressive dilution of each of the extracts in turn showed 4m almost linear diminution in reactivity, the reaction dis- appearing in Platanus between dilutions of 1 : 256 and 1 : 512 and in Prunus between dilutions of 1 : 64 and 1 : 128. (Note: Normality of the extracts was arbitrarily chosen at 1 part _ dried tissue to 10 parts distilled water. The dilutions referred __ to here were further dilutions of such normal extracts.) : 4 The reactivity of the extracts was unaffected by continued estion with trypsin, pepsin, and yeast enzymes at appropriate _ PH values and temperatures. ‘Complete precipitation of Prunus by Robinia, Platanus, Ribes, oh: Hydrangea completely eliminated any further reaction of the ____thus-precipitated Prunus by any other of the latter four. — i os Fractionation of the extracts by the Rimington technique _ ‘orremoving carbohydrates showed that the reactive principle « Prunus was precipitated by neutral lead acetate and bea ‘“atirely recoverable from that precipitate on treatment with ie HS. The reactive principles in Robinia and Platanus, on the ther hand, were unaffected by treatment with both neutral and alkaline lead acetate and were completely recoverable in the nearly Molisch-negative filtrate from such treatment. z . “gth of the reactive principles thus fractionate varied tly with the strength of chloride and carbohydrate (used as *s of the degree of dialysis) and bore no relation vee Strength of protein as indicated by the Millon and xantho- € tests (used as indices of the degree of dialysis). ing the evidence presented, the probability eat ae s-Robinia-Ribes reactions are due to proteins 1s / ~ } # } f bt 288 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm very slight. Although plant proteins are more resistant to heat than animal proteins and accordingly the results in heating the extracts are inconclusive by themselves, yet it is inconceivable that proteins would not be removed or inactivated by treatment with acids and alkalis and with alcohol, by salt content and pH, by contaminations and enzyme action, and by the removal of various constituents of the extracts as must be the case if proteins are here involved. The identity of the various reactions of these species (item 9 above) argues against the protein hypothesis, and finally the experiments in dialysis show that it is extremely unlikely that the reactions are due to extremely small traces of proteil. The likelihood that these reactions are due to lipoids or to carbohydrates is also very dubious in view of the results given in items 5 and 10 respectively. | In the course of an experiment on hydrolysis of the extracts @ / small excess of CaCO; was added to a sulphuric acid solution of 7 Prunus. On neutralization it was found that its precipitating action was completely reversed. The presence of the calcium | sulphate thus formed, although only very slightly soluble in water, rendered the Prunus negative to Platanus, Robinia, and Ri ; and, as was later found out, positive to extracts with which it had formerly been negative. Attention was accordingly directed . to the inorganic constituents of these extracts with war \ reference to calcium compounds. This last was a most productive field. It was soon discovered and confirmed that the reactions in the woody plants under COP — sideration are due to the interaction of free calcium ion in cert? extracts (in this case in Robinia, Platanus, and Ribes) with free oxalate ion in the other extracts (represented for the present by j Prunus). Such a view has resulted both from chemical anaiy*" — of the precipitates and from studies of the behavior of the whole extracts with regard to presence or absence of calcium and oxalate — The analytical evidence supporting the view that these reactions are due to the interaction of such ions in first presented: 1. The precipites, after washing in several changes of watel: white, heavy, limey, easily centrifuged, and inorganic ™ oh a pearance. There is no charring on heating to 500°C. 2. The precipitates microscopically are in the form of regu : granules, not amorphous, identical in appearance with a : commercial samples of calcium oxalate. ere ie 3. Recrystallization of the precipitates (by solution ™ on) H.SO, and precipitation by neutralization with strong CaCO gives crystals of the characteristic size and shape of egh | 1982] CHESTER, THE PRECIPITIN REACTION IN PLANTS 289 crystals, and indistinguishable from crystals of a commercial sample of CaC,O, similarly treated. 4. Treatment of the granules of the precipitates with strong H.SO, under the microscope shows first a moderate solution followed by a very striking conversion of the remainder of the granules into the characteristic raphides of CaSQ,. This is a fairly accurate test for CaC,Q, and is precisely the behavior of a sample of commercial CaC,O, similarly treated. __ 5. The precipitates are insoluble in all ordinary solvents. They , are moderately soluble in strong H.SO, but not in weaker acids. : Their solubilities are thus equivalent to those of CaC,Qx. _ 6. If alcohol is added to the H,SO,; solution of the precipitates | there is a precipitation. (Test for calcium ion.) 7. The acid solution reduces potassium permanganate. (Test for u oxalates). 8. Ignition of the washed precipitate yields 33% of oxide. The theoretical yield for CaC.O, is 38%. The evidence thus given leaves no room for doubt that the _ Precipitates resulting from the addition of the Prunus extract 4 to those of Platanus, Robinia, and Ribes consist of calcium oxalate. This evidence is still further confirmed by experiments to be re- Ported below. At this time, however, certain questions arise: _ @) Is this calcium oxalate reaction the only one involved in the Combinations of woody plants under consideration? (b) How generally is the calcium oxalate reaction distributed through : ‘Plants with especial reference to the tests which have been reported a3 [ have been reported, and if so, what is their nature? The “wets to these questions will appear from the following con- : If the calcium oxalate reaction is the main or only reaction in beg Woody plants available for this study it should be possible to Aide at the extracts into two groups, a calcium-positive, oxalate- | regative group which is intranegative but positive to a second, w€-Positive, calcium-negative group. Such is possible. Prents “Mikado” is here the only representative of the* oxala 3 g oxalate, lacking calcium), while the “calcium comprises Platanus, Ribes, Robinia, Syringa, Hydrangea, three species of Ligustrum. With the exception of athe teaction to be mentioned later, the latter group 3s periec’'y ative. On the other hand, the members of this dereagel All produce precipitates identical in appearance with te Teaction” when they are added to weak solutions © ty BE s + 290 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm oxalates (K,C,0, .01 M; (NH,)2C,0, .02 M) but are inactive to” the addition of weak solutions of calcium salts (CaCl, .005 M; Ca(NO;), .005 M). Prunus, however, reacts positively to pure solutions of calcium salts of the concentrations given above but is negative to these oxalates. At this juncture it was felt advisable to extend these results to _ the Solanaceae for the dual purpose of obtaining more extensive data, and of obtaining data on the family on which the work of : Kostoff and Silberschmidt has been done. Accordingly, all the — possible interreactions of thirty-five species of Solanaceae as well as all their reactions with the woody plants under consideration were | carried out by the writer in collaboration with Dr. Thomas Whitaker. — The results of these tests were highly confirmatory. Testing the : various Solanaceae against weak solutions of two oxalates and of two calcium salts it was found that the species fell into three groups, # group (Ca- Ox+) comprising Atropa belladonna, Solanum tuberosum, Capsicum frutescens, Physalis peruviana, Salpiglossis sinuata, Datura ferox, D.metel, and D.innozxia which ive in varying degree to the calcium salts and negative to the oxalate: a group (Ca- Ox) comprising Datura Wrightii, Cyphomandra betacea and Browallia viscost” which were negative to both ions, and a group (Ca+ Ox-) com prising eighteen species of Nicotiana, Petunia violacea, Lycopersicum cerasiforme, Solanum capsicastrum, S. melongena, and S. migrum which were negative to free calcium ion but positive to free oxalate — a ; ; u ion. The extracts of group (Ca- Ox-++) were positive to those 2 group (Ca+ Ox-) but those of group (Ca- Ox-) were negative tol in both other groups. Moreover the strength of the * ‘precipitin reactions” bore a very close relation to the strength of the reactions c of the corresponding extracts with the pure salt solutions. (Ca- Ox-++) group was perfectly negative inter se and the (Cat ' Ox) group was also negative inter se except for a few weak principally involving Datura Wrightii and Nicotiana Rusbyi- y Thus a consideration of the Solanaceae affords very strong | ae z firmatory evidence as to the validity of the calcium oxalate : planation. A warning should be introduced at this point. The writer 1 well aware of the fluctuations of the salt content in plant juices s and makes no contention that the values for the reactions | - definite and alwaya exactly reproductile: ‘The readings WET with the observer, with the technique, and with the ae “ : and heredity of the plants investigated. However, are — developments of the content of such ye constit Laie characteristic of certain species i is evident from a 1982] CHESTER, THE PRECIPITIN REACTION IN PLANTS 291 ‘the literature. Moreover that the results here described are comparable is evident from the facts that the layerings and readings were all made by the same observer, that all the reactions described _ for any given species of plant in this study were made from one given extract, that the Solanaceae were all grown under uniform greenhouse conditions, and that all of the “precipitin’’ tests : described were made within a few days of one other. _ An important confirmation should now result from testing the “calcium” Solanaceous extracts against the woody “calcium” | extracts. If the calcium oxalate explanation be correct and com- plete, then such reactions should all be negative. In order to extend the limits of this experiment still farther, all the “oxalate” "4 extracts, woody and herbaceous, were treated with a slight excess ‘ of CaCl, and filtered.. Hence the calcium oxalate reaction was completely eliminated from consideration. Then all the possible _ combinations of the 42 extracts were again tested. This experiment brought out a very important fact. There is a second reaction in which Platanus, Robinia, and Ribes show a varying degree of _ Teactivity against all the Solanaceae. The reactions of Platanus, : Robinia, and Ribes are so well correlated here that it is most prob- : that there is a single substance (A) present in these three a “extracts which reacts with a second substance (B) in the Solanaceae. = Prunus, Syringa, Ligustrum obtusifolium, and L. ibota react against ‘Neither the (A+ B-) nor the (A- B+) groups and accordingly _ 4 assumed to contain neither reactive principle. - tis now possible to eliminate both the calcium oxalate and AB Teactions from consideration by considering only the interreactions Peekt oxalate-negative, A-negative extracts. If this group of teactions be studied it is seen that the great majority are eae with the exception of Browallia, "Salpoleti, Petunia, oe Seng Ligustrum vulgare, and Solanum nigrum (M- negative to both groups. still remain a very few weak reactions (2%) which may | as due to a substance (X) present in gee eerors: Physalis, Datura innoria, and D. m 292 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm which reacts with a substance (Y) present in the Oleaceae studied, in Browallia, and in Nicotiana suaveolens, but absent in the other remaining extracts These four reactions, the first proven, the second and third assuredly present and distinct but of unknown nature, and the fourth more problematical, since it is very weak, explain all of the thousand or fifteen hundred reactions considered in this study. Of all the positive reactions at least 57% are due to the calcium oxalate combination, approximately 23% to the AB reaction, 17% to the MN reaction, and less than 3% to the XY reaction. We are now in a position to return to the questions propounded above. First, with regard to the presence of any other reaction in the Prunus-Platanus-Robinia-Ribes complex, we may say definitely — that there is no other “‘precipitin”’ reaction demonstrable by the technique employed than the calcium oxalate reaction. Removal of the calcium oxalate reaction leaves these four extracts perfectly internegative. Second, as to the distribution of the calcium oxalate reaction among the experimental plants heretofore atagiet it may be said with certainty that this reaction is of wide bution, that it accounts for the majority of the reactions of the plants used by Kostoff, and that it must be eliminated before any immunological interpretation can be made of the data thus far — published. Kostoff published a table of ‘normal precipitin — actions” in the Solanaceae which includes 56 positive par 40 of these have been repeated in this laboratory. Of the # repetitions, in 7 cases there were no positive results, in 10 cases — the reactions were due to the MN combination, and in 23 cas$ — the reactions were due to calcium oxalate. Kostoff reported i: acquirement of “‘precipitin potency” in 12 instances. The normal reactions as repeated in this study eh that of the 8 repetitions 3 reactions were due to calcium oxalate, 3 were due to the ) gig ; action, and in 2 cases no positive results were obtained. | * found a decrease of “precipitin potency” in 4 cases. 8 of these have been repeated and in all 3 cases the normal reactions We — found to be due to the MN combination. 4 a8 Third, as regards the a and nature of other reactions it A the plants considered, it may be said with certainty that there * is at least one other reaction present, probably two, and possibly = three or more. The nature of these other reactions is gees vestigated. For the time being, it may merely be said that W" regard to the AB and MN reactions they are indubitably ie in nature since the washed precipitates are strongly « in ignition. The precipitates of these reactions are B; Peay different 1982] CHESTER, THE PRECIPITIN REACTION IN PLANTS 293 appearance from the calcium oxalate precipitate. They are brown in color, even after washing, are copious and flocculent, and easily pass into a non-filterable colloidal suspension in pure water. As further proof of the presence of more than one reaction, instances may be mentioned in which given extracts may be precipitated thoroughly to remove the calcium oxalate factor. _ Such calcium oxalate free extracts have lost none of their potency _ for precipitating in the presence of the opposite principle of the AB | or MN reactions. : A word should be inserted at this point regarding the expressions | “calcium” extracts, “oxalate” extracts, etc., frequently used in the foregoing pages. Such expressions do not imply that certain plants are free from demonstrable oxalate or calcium respectively, | since the extracts used reveal the excess of either ion after such | autoprecipitation of CaC,0, as may take place in preparing the extracts. It is very apparent, for example, that Prunus Armeniaca vars. “Mikado” and ansu fall into the “oxalate” class not because they lack calcium but because they contain more than sufficient Moxalate in the affected cells. That this is not generally distributed _ “Toughout the plant (as would be true of an immunological - ‘Substance resulting from grafting) is shown by the fact that the _ ““acts made from green areas of mottled green-and-yellow leaves : d) test just as do normal leaves of ungrafted te ae “cumulation of free oxalate in the blighted portions 0 ; rs harmony with the fact that such leaves when ground and mixed Water begin an autoprecipitation (of the calcium normally ent in the leaf by the oxalate) which if permitted to he cee d ‘he end results in an excess of unused oxalate which gives the escribed. wever, both oxalate and 294 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm x Regarding the second question, as to the specificity of the re- actions thus far observed, light was thrown upon this phase of the work by a consideration of the distribution of free calcium ion, ‘free oxalate ion, and the other reactive principles in these various ei fee es a species of plants studied. On purely a priori grounds it would not _ be unreasonable to expect that in general the distribution of such substances would show some agreement with the taxonomic i relationships of the plants involved. Thus it is well known that the presence or absence of calcium oxalate crystals in plant cells is of taxonomic significance. How does this apply to the Solanaceae and woody plants here studied? A dendritic chart was made including all the possible combinations of reactive substances believed to be responsible for the reactions observed. Thus the chart would first include a trifurcation representing presence of free calcium or free oxalate or absence of both. Each subdivision could then be further trifurcated according to presence or absence of the A and B principles, ete. There would thus be 81 possible combinations of reactive principles represented. If the assumption of specificity be correct, there should be a definite tendency for — related species to have related positions on such a chart. actual placement of the various species in their proper positions on such a chart brought out the validity of this assumption a striking clearness. 18 species of Nicotiana had been studied. All contain free calcium but not free oxalate, all contain the principle B but lack the principle A, N. Rusbyi alone contains the principle M, the other 17 species all contain N, while all the species of Nicotiana save N. suaveolens lack both X and Y. The possibility be! their being so constituted by chance is infinitesimal. Similarly ~ : Oleaceae considered are closely allied in all containing Ca**, al : lacking principles A and B except L. vulgare which has only a trace of this, all lacking both X and Y, and separating only with reg® to the presence or absence of M and N. Robinia, Platanus, # : : Ribes, in closely related families, all are found to lie in the en quarter of the dendritic system, being all positive for calcium ane for principle A, they being the only representatives of the A pri in the whole scheme. So, too, the 4 species of Datura as well a8 | those of Solanum are found to agree in three of the four reactive 3 principles, separating on only one. It is thus seen that the distribution of the reactive principles the 42 species of plants here considered is in good accordance ‘ sie the taxonomic positions of these species, which clearly explains ©" earlier findings, before the nature of the reactions had bee? vestigated, with regard to the specificity exhibited by the reactio™ iD. ig a : a Z . pig f § : _ 198] | CHESTER, THE PRECIPITIN REACTION IN PLANTS 295 SUMMARY 1. The present paper gives a condensed account of the writer’s investigations of the biochemical nature of the “normal pre- cipitin reaction” in plants. _ 2, Extensive experiments in the testing of various physical and chemical treatments have yielded strong evidence against the hypothesis that the reaction in certain woody plants, viz. -runus Armeniaca, Platanus acerifolia, Robinia fertilis, and Shs Carrierei, is of protein nature. 8. Conclusive proof is given that the reaction resulting in ap- proximately 57% of the precipitates observed in all the possible combinations of 42 species of Solanaceous and woody plants is o; the precipitation 6f calcium oxalate by the interaction of the ___ Tespective ions in pairs of extracts. This precipitation of calcium oxalate is so frequently the sole or main phenomenon in the reactions hitherto considered immunological in nature as to invalidate immunological interpretations laid on such reactions in which the calcium oxalate factor is not eliminated. 4. The remaining reactions are susceptible to interpretation on the _ assumption of the presence or absence of three other pairs of reactive substances. The nature of these reactive pairs is _ telatively obscure at present, but it is being investigated. One such pair (AB) is particularly characterized by the reactions ae of oxalate-free Robinia, Platanus, and Ribes with the Solanaceae. _ This reaction accounts for about 23% of the positive precipita- tions. Tt is shown to be distinct from the calcium oxalate _ feaction. A second pair (MN) is represented by the reactions } of oxalate-free and AB-free Nicotiana Rusbyi and Datura — Wrightii with most of the other Solanaceae. It is plainly distinct from the calcium oxalate and AB reactions and ac- counts for about na of the pee A very 1 the - Tepresented by those of Solanum capsicastrum, Atropa onna, Pi epee agi Datura innozia, vie ii rtain Oleace oh ’s earlier findings with regard to “ precipitin I reactions” th ysiologically blighted Lilac and to the sei ny ae _ normal precipitin reaction” are readily interpre 133 of the findings of the present study. now] tions and freedom gement is due for the many sugges yh as given the writer by Dr. Ronald Ferry of the 296 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm ; vard Medical School, and to Dr. Thomas Whitaker who codperated — in the testing of the Solanaceae. The writer is also indebted to — Professor E. M. East for placing the Solanaceae at his disposal — and to Professor J. H. Faull for numerous helpful suggestions. LITERATURE CITED 1. Cuester, K. S. (1931). Sete slight: A disease of Lilac related to the employment of certain understocks in propagation. pipet: Arnold — Arb. 12: 78-146.) (1932). Studies on the precipitin reaction in plants. The specificity of the normal precipitin reaction. (Ibid. 13: a) 3. Kostorr, D. (1928, 1929). Induced immunity in plants. (Proc. Natl — Bane: Pea. 14: 236-237).—Acquired immunity in plants. (Genetics — 14: 7 | 4. SILBERScHMIDT, K. (1931). Studien zum Nachweis von Antikérpern in Pflanzen. (Planta v Wiss. Bot. 13: 114-168.) PaTHOLOGICAL Vieteabons ARNOLD ARBORETUM,

ROSACEAE Neillia sinensis Oliv. f. glanduligera (Hemsl.), f. nova. Neillia sinensis Oliv. var. glanduligera (Hemsley in herb. : Neillia glandulocalyx Léveillé, Fl. Kouy-T'chéou, 348 (1915). OES A typo differt tubo calycis glandulis longe setoso-stipitatis lnstructo. : Cava. Hu peh: without precise locality, A. Henry, no. 5554a (type of N. sinensis var. glanduligera). Kweichou: route de Pin-fa & Tou-yun, J. Cavalerie, no. 2461, July 1905 (holotype of N. glandulocalye; merotype in A. A.). Go This form seems almost as common as the typical form with i glabrous calyx and the following specimens are referable to it: an Shensi: “monte Thae-pei-san,” J. Giraldi, July 20, 1897; _ Hupeh: without special locality, E. H. Wilson, Veitch Exp. no. _ 701 in part; (fruiting branch only); Ichang, E. H. Wilson, Arnold _ Arb. Exp. no. 86 in part (fruiting branches only); Hsing-shan BS hsien, E. H. Wilson, no. 189; Hunan: in monte Yun-schan : S ease urbem Wukang, H. Handel-Mazzetti, no. 12010. ahs oe ea media Schmidt var. monbetsusensis (Franch.) Cardot _ “Pud Nakai in Tokyo Bot. Mag. xt. 464 (1928). Spiraea Fauriei Léveillé in Fedde, Rep. Spec. Nov. vit. 281 (1910).— E ‘nov, : i ji a “HALIN: in rupibus littoris Kornakoff, U. Faurie, no. 515, oe _ (holotype of S. Fauriei; photo. of type and isotype 1 Th a note on the sheet of the type specimen Koi 5 ne identified Siig cs with 4 weskeresoins Franch.,and sasdi but § s 18 cited by Nakai (1. c.) under the specimens eeanaane “| hame given by Léveillé is not mentioned. = ec from vol. xn, 281 (1931); for preceding parts see *- beter 184-196. 300 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Spiraea Martini Léveillé in Fedde, Rep. Spec. Nov. 1x. 321 (1911); Cat. Pl. Yun-Nan, 244 (1917).—Rehder in Jour. Arnold Arb. 1. 258 (1920). Spiraea fulvescens Rehder in Sargent, Pl. Wilson. 1. 439 (1913). HINA. Y unnan:environs de Yun-nan-sen, dans la montagne, ravines, bords de torrents, E. Bodinier, no. 89, March 2, 1897 (holotype of S. Martini; merotype and photo. in A. A.); Mengtze, — barren clay hills, alt. 5000 ft. A. Henry, no. 10662 (holotype of S. fulvescens in A. A.). Spiraea Cavaleriei Léveillé in Fedde, Rep. Spec. Nov. 1x. 8% (1911); FI. Kouy-Tchéou, 301 (1915). soe Cuina. Kweichou: district de Tou-chan, J. Cavalerie hae herb. Bodinier, no. 2602, March—April 1899 (holotype; photo. in : . A.). 3 | This species is closely related to the preceding and both may be forms of one species. It differs from S. Martini in the distincly ovate leaves rounded at base, crenate and not at all lobed, less glaucous beneath, in the larger inflorescence with about 10 ee more flowers and an elongated rhachis about 1 cm. long; in the — larger flowers about 8 mm. across, in the slenderer pedicels up 1.2 cm. long, and in the small leaflets at the base of the inflores cence being absent or reduced to very small bracts. pee Spiraea pubescens Turezaninow in Bull. Soc. Nat. Moscou, V- a iraea ouensanensis Léveillé in Fedde, Rep. Spec. Nov. Vi. 197 a (1909). —Synon. nov. =f Korea: in montibus Ouen-san, U. Faurie, no. 315, July Bio : (holotype of S. ouensanensis; photo. and isotype in A.A). ; An isotype of Faurie’s no. 315 in this herbarium has the leaves — 13d rots and is in fruit. Nakai (Tokyo Bot. Mag. xu 165 466. 1928) distinguishes two forms, var. lasiocarpa and Med peee = carpa; the former seems to be the more common and S. #6 belongs to it. a = Spiraea ovalis Rehder in Sargent, Pl. Wilson. 1. 446 (1918). Spiraea microphylla Léveillé in Bull. Géog. Bot. xxv. 44 (1915); Oo Pl. Yun-Nan, 244 (1917).—Synon. nov. gee os . CHma. Yunna x raat des montagnes 4 Kiao-mé-t, =f a 3150 m., E. E. Maire, May 1912 (holotype of S. microphyll merotype in A. A.). cB: ; Pelee This species was known so far only from the type eentt collected in Fang Hsien, Western Hupeh. ae pedi o5! S 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 301 _ Spiraea Veitchii Hemsley in Gard. Chron. ser. 3, xxxmi. 258 Spiraca atemnophylla Léveillé in Bull. ae Bot. xxv. 44 (1915); Cat. Pl. Yun-Nan, 244 (1917).—Synon. nov Cova. Yunnan: rochers des inositegnes & Kiao-mé-ti, $200 m., E. E. Maire, Aug. (syntype of S. atemnophylla; photo. in A. ‘ A.); brousse de mamelon A Ta-hai, E. E. Maire (ex Léveillé; syntype _ of 8. atemnophylla); paturages du haut plateau de Ié-ma-tchouan, $200 m., E. E. Maire (ex Léveillé; syntype of S. atemnophylla). This species does not seem to have been recorded from Yunnan before. ee ee ee en ee _ Spiraea Mairei Léveillé in Bull. Géog. Bot. xxv. 43 (1915); _ Cat. Pl. Yun-Nan, 244 (1917). oe Cama. Yunn an: brousse des montagnes & Kiao-mé-ti, 3100 ma, E. E. Maire, May 1912, “arbuste fragile, buissonant; fleurs : sn soal (holotype; merotype in A. A.). This species is very similar to S. Teniana Rehd. in its inflores- ‘ence and the shape and serration of the leaves, but differs in the striped or slightly angled villous branchlets, in the leaves being on both sides with long accumbent hairs, in the longer en pubescence of the inflorescence and the pubescent Pe sais i sae : ; Spiraea japonica L. f. var. acuminata Franchet in Nouv. Arch. e es VIII. oy de eee n. 36) (1885 3 LK otha he oe Nov. rx. 322 (1911); Rie ouy-I¢. a "Spiraea Bedinier bigs 1819) -_ Staon no 11).—Synon. nov. ee Esquirolis 1 Léveillé 1. e. (1911); L a Fests _—Synon. nov. > -Cutwa, Kweichou: Mont du Collége, E. Bodinier, July 3, 3 1900 emntype of S. Bodinieri; photo. in A. A.); environs de Tou- A » 4. Cavalerie, no. 2682, June 1899 (syntype of S. Bodinieri); ie Lou-hong-koan, E. Bodinier, no. 1709, June 10 and July $: 897 (syntypes of S. Bodinieri var. concolor photos. in A. A.); tien, E. Bodinier, no. 492, (not 429) July 16, 1900, “arbuste; fa (syntype of S. Esquirolii; merotype in A. A 18 No specimen of S. Bodiniert with the data “Mont du Juillet 3, 1900, E. Bodinier” in the Léveillé herbarium, Eby elimination it must be the sheet labeled i in Léveillé’s hand- Spiraea Bodinieri without any further data. piraea Bodinieri var. concolor is not mentioned in Flore de chéou, but its type number is cited under 8. -sayeetyd sine aged ee tee Geog. Bot. xxv. Ti aes: q seater Bir Hrons in i 302 JOURNAL OF THE ARNOLD ARBORETUM (von. xm Cotoneaster horizontalis Decaisne in Fl. des Serres, xxt, 168 1877 : in 8 Chaffanjoni Léveillé in Fedde, Rep. ay Nov. xa. 1 pris FL. Kouy-Tchéou, 145 (1914).—Synon. n . Cuina. Kweichou: Kouy-yang, cascade ide Collége, Chaj- fanjon, May 12, 1928 (holotype of Diospyros Chaffanjoni; photo. — in A. A.). 3 Léveillé compares his Diospyros Chaffanjoni with D. vaccinioides Lindl. to which it has a very slight resemblance only in the shape of the leaves, but the latter are considerably larger in D. vaccinioides. — Cotoneaster Franchetii Bois in Rev. Hort. 1902, p. 379, fig. 159-161, 164. Cotoneaster Mairei Léveillé in Bull. Géog. Bot. xxv. 45 (1915); Cat -1. Yun-Nan, : 229 oly 17).—Synon. nov. C airei var. albiflora Léveillé, 1. c.—Synon. nov. . Cuina. Yunnan: rochers des montagnes derriére Tone. tchouan, 2600 m., E. E. Maire, June 1912 (holotype of C. Maireis photo. in A. A.); rochers des collines & Kin-tchong-chan, 2550 My E. E. Maire, May 1912, “arbuste dressé buissonant; fleurs blanches _ : sur fie rouge” (holotype of C. Mairei var. albiflora; merotype inA ‘9 Var. albiflora differs only slightly in its smaller leaves and pure white flowers from typical C. Mairei; the color of the flowers varies in most species of this group from pure white to more or ess flushed with pink in bud and on the back of the petals. Tit varietal name is not mentioned in the Catalogue des plantes ie Yun-Nan and does not appear on the label of the specimen from Kin-tchong-chan. ef Cotoneaster glaucophyila Franchet, Pl. Delavay. 222 (1890). Photinia rosifoliolata Léveillé in Bull. Géog. Bot. xxiv. 142 (1914); 7 Kouy-Tchéou, 349 Ga —Synon. nov. June Cura. Kweichou: Tin-fan, J. Cavalerie, no. 3838, re rag! ‘fleurs blanches” (holotype of Photinia rosifoliolata; merotyPe in A. A.) : Crataegus scabrifolia (Franch.) Rehder in Jour. Arnold ny xu. 71 Cine Crataegus Henryi Dunn Linn xxxy. 494 (1903). Craiapue Bodie Level in Bal Soe Bor Bot. France, LY. 8 Guing Yubueu ‘montagnes entre Ma-kay et Sets chou, E. eaten ibeeay 4, 1897 (holotype of C. Bodiniert; 1 Cotoneaster Blinii ie dC. Eaquirali Le (FL K Kooy Tee peo boone ‘donot belong te Secldindteh + the ineoniplete : does not allow t to place them They may Sores Pomoideae, resemb Seger eo somewhat Photinia or alain differing in the structure of the fruit, immature ; in C. Esquirolit. 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 308 4 in A. A.); Ma-chou, alt. 3000 m., E. E. Maire, April 1911-13 (in herb. Léveillé sub C. Bodinieri). 4 Both specimens are in bloom, while the type of Franchet’s ie scabrifolia is in fruit. Crataegus cuneata Siebold & Zuccarini in Abh. Akad. Miinch. - &1yv. 130 (Fl. Jap. Fam. Nat. 1. 22) (1846). Crataegus Argyi Léveillé & Vaniotin Bull. Soc. Bot. France, tv. 57 (1908). r fie in Mem. Acad. Ci. Art. Barcelona, ser. 3, XII. no. 22, p. Ro Or “age stephanostyla Léveillé & Vaniot, 1. ¢. (1908).—Léveillé, 1. e. a - Crataegus C Chantcha Léveillé in Fedde, Rep. Spec. Nov. x. 377 (1912); : Fl. Kouy-Tchéou, 346 (1915). : HINA. Kiangsu: Chang-li-hong, Chang-sun, Vou-né, Ch. e Ary [1846-66] (holotype of C. Argyi; photo in A. A.); Si-tcha- q Si-souo-se; Zi-se, montagnes, Ch. d’Argy, May 15 [1846-66] q (olotype of C. stephanostyla); merotype in A. A.). Kweichou: - fivirons de Gan-pin, E. Bodinier, April 29, 1897 (holotype of C. _ Chanicha; photo. in A. A.). ¥ The specimens of (. stephanostyla and C. Chanicha are in bloom — and that of C. Argyi is in fruit. All three represent undoubtedly ik the widely distributed and characteristic C. cuneata Sieb. & Zuce. a Crataegus pi pinnatifida Bunge in Mém. Acad. Sci. St. Pétersb. 11. : nen Pl. Chin. Bor. 26) (1831). 4 a coreanus Léveillé in Fedde, Rep. pee Noy. vit. ee, _ hol. I, 7700 (1000) ohakal: poe os Me t. 26, fig. a (1916). : circa Seoul communis, U. Faurie, no. 307, June 1906, C; ‘circa Chinnampo, U. Faurie, no. 308, Sept. 1906 (syntypes of _” “reanus; isotypes and merotype in A. A.). ; Cra coreanus has been already identified by Nakai (I. ¢.) say Cy Pinnati ifida var. psilosa Schneid., but this identification < le only as far as it concerns Faurie’s no. 307, while no. 308 ich is i in fruit belongs to the A reeet form (var. typica Schneid. — The two forms differ only in their glabrous or pubescent fence; they are hardly of varietal rank and do not lergsg Seographical ranges. : # Kochneana Schneider in Bull. Herb. Boissier, sér- 2, v1 : Cat. Pl. * Bien Léveillé in Monde PL. sér. 2, xvi. 28 (1916); un-Nan, 242, fig. 62 (1917).—Synon. nov. Flee ¥ Yunnan: plateau du Io-chan, 3400 m., E. . ; ; : : “grand arbuste rameux”’ (holotype of 8. Valbrayt 4 ae) 304 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm Sorbus Keissleri (Schneid.) Rehder in Sargent, Pl. Wilson. 11. 269 (1915). Sorbus Mairei Rehder & erat apud Léveillé, Cat. Pl. Yun-Nan, 242 (1917), nomen.—Synon orbus Aria var. a "Leveillé, Cat. Pl. Yun-Nan, 242 (1917), pro synon. Sorbi Cuina. Yunnan: rochers des montagnes 4 Kiao-mé-ti, alt. 3100 m., E. E. Maire, “‘arbuste buissonant, haut 1 m. ou 2 m.” (holotype of S. Aria var. Mairei; photo. in A. A.). Léveillé cites (I. ¢.) “S. Aria Crantz var. Mairei Lévl. in Fedde Rep. m1. 341,” but neither there nor elsewhere did I find this name, though it appears on the label of the type specimen. I may have written to Léveillé with whom I corresponded at that time, that his Sorbus Aria var. Mairei did not belong to S. Aria, but I had no part in making the combination S. Mairet. Sorbus Hemsleyi (Schneid.) Rehder in Sargent, Pl. Wilson. 1. 276 Ce. Pirus Koehnei Léveillé in Fedde, Rep. Spec. Nov. x. ake (1912); Fl. Bete rare) 351 Huis): non a Koehnei Schneid. (1906).— Cail heel hed: Pin-fa, montagnes, J. Cavalerie, no. 2376, June 15, 1905 (ex Léveillé; syntype of Pirus Kochnei); J Esquirol, no. 385 (syntype of Pirus Koehnei; photo. in A. A.). Photinia villosa (Thbg.) De Candolle, Prodr. 1. 631 Sagi TRehder 2 n Sargent, PL Wikoe Se ies amen hh | Pirus por iss errs caimowiczana Léveillé 1. ¢. (1912).— sls vee Hie eae Léveillé, 1. ¢. (1912).—Rehder, Te (1915). | P g.) Decne. var. typica (Schneid.) Nakai, FL Sylv. Kor ‘vi. 28 he eeei Pourthiaea villosa v brunnea (Lévi.) Nakai I. ¢. 29, t. 7 (1916). Korea. riche dent eines in silvis Hallaisan, 800 m., E. no. 2819, Oct. 1909 (holotype of Pirus REECE merotype and isotype in A. A.); in silvis Haitchenam [?], E. Taquet, no. 282) Aue 1909 (holotype of Pirus sinensis var. Maximowicziana; ; photo, and isotype in A. A.); in silvis Hallaisan, 900 m., E. Taquet, 2° 2815, June 1909 (holotype of Pirus spectabilis var. albescens; photo. and isotypein A. A.). : Pirus brunnea does not seem to be sufficiently different from typical Photinia villosa to consider it a distinct variety aS done by Nakai, while he refers Pirus spectabilis var. albescens to ect thiaea villosa var. typica. Pirus sinensis var. M armen does not mention. The latter species and P. brunnea are re by fruiting specimens; P. er aliecen sn boot 4 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 305 _ Photinia villosa var. coreana (Decne.) Rehder in Jour. Arnold Arb. m1. 45 (1920). Pirus mokpoensis Léveillé in Fedde, Rep. Spec. Nov. vit. 200 (1909). vse). villosa var. coreana (Decne.) Nakai, Fl. Sylv. Kor. vi. 29 Korza: cirea Mokpo, U. Faurie, no. 1556, May 1907 (holotype __ of Pirus mokpoensis; isotype in Herb. Arnold Arb.). _ Pirus mokpoensis was first identified with Photinia villosa var. coreana by Nakai. Photinia Beauverdiana Schneider in Bull. Herb. Boiss. sér. 2, -¥L-819 (1906). x tinia Cavaleriei Léveillé in Fedde, Rep. Spec. Nov. 1v. 334 (1907); q Fl, Kouy-Tchéou, 349 (1915 5).—Synon. nov. | Cara. Kweichou: Ly-po, J. Cavalerie, no. 2631, May ; 1809 (holotype of P. Cavaleriei; photo. in A. A.). Photinia parvifolia Schneider, ‘Ill, Handb. Laubholzk. 1. 711, fig aoe 2-9 (1906). a oe Rehder & Wilson in Sargent, Pl. Wilson. 1. 189 wo) —Synon. no Crataegus Cavaleriei “Léveill é, Fi. pedi eee 346 (1915), quoad . . cit. -1303; non Léveillé & Vaniot V omarovit Léveillé & Vaniot in *rodde, Rep. Spec. Nov. 1x. _ 8 (i9i0); Fi. ‘Kouy-Tehéou, 66 (1914).—Synon. nov Cua. Kweichou: Pin-fa, montagne, J. Casclerie, no. 1303, May 3, 1902 (? paratype of Viburnum Komarovii; photo. in A. A.); » J. Cavalerie, no. 1303, Oct. 1908 (holotype of Viburnum _ Komarovii; merotype in A. A.); no. 1892, Sept. 1908 (ex Léveillé, FL Kouy-Tehéou), : i here seems to be considerable confusion in Léveillé’s numbers. fos Léveillé’s herbarium in the cover of Crataegus Cavaleriei there a 'S@ flowering specimen from Pin-fa, May 3, 1902, with the original 4 _humber 93 crossed out and 1303 written over it, but the fruiting : 3 Specimen under no. 93 upon which Léveillé’s description is based 7. 18s to Malus Sieboldii (Rgl.) Rehd. As Viburnum Komarovii Spin ‘describes, (l. ¢.) under no. 1303 flowering and fruiting aS: » though he gives only October 1908 as collecting date pra i a an additional number 1892, Sept. 1903, from Pin-fa which © not seen. Apparently the flowering specimen in the cover cited May Viburnum Komarovii; no reference to a ieaees unis Y 3, 1902 appears under Crataegus Cavaleriet or Vi- Komarovii. the abundant material now at hand I cannot find that P. | Rehd. & Wils. is en | distinct from P. parvi- Original description, but in the Flore du Kouy-Tchéou — Tataegus C Cavaleriei is the flowering specimen described, but — 306 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm folia to be*maintained as a species, since the extreme forms are connected by intermediate specimens which makes it difficult to separate clearly the two groups. Photinia seer aioe Rehder & Wilson in Sargent, Pl. itbevas 1. 190 (19 rus Feddei Leeatit 5 in Fedde, Rep. Spec. Nov. xu. 189 (1913); Fl dc tallde wero 350 (1915). Cuina. Kweichou: Pin-fa, J. Cavalerie, no. 2533, Oct. 11, 1905 (holotype of Pirus Feddei; photo. in A. A Photinia crassifolia Léveillé, Fl. Kouy-Tchéou, 349 (1915), nomen seminudum.—Cardot in Bull Mus. Nat. Hist. Paris, xxv. 398 (1919). Photinia Cavaleriei Léveillé in Fedde, Rep. Spec. Nov. x1. 66 (1912); non Léveillé (1907). Photinia eb pee Levl. var. denticulata Cardot in Lecomte, Not. Syst. m1. 372 (1918). CHINA. Kvel chou: Tin-fan, J. Cavalerie, no. 3571 (in part) June 1909 (holotype of P. Cavaleriei; merotype in A. A.); Gan-chouen, J. Cavalerie, no. 3571 (in part) April 1912 (paratype of P. crassifolia; merotype in ; This species was first pula ted as P. Cavaleriei based on a specimen collected at Tin-fan in 1909 by Cavalerie and numbered 3571; in 1915 Léveillé enumerates under the name P. crassifolia the preceding specimen with the addition of a specimen from Gan-chouen collected in 1912 and also numbered 3571, but without reference to the previous name and description. The species varies in the pubescence of the inflorescence and in the denticu- lation of the leaves; in the specimen from Tin-fan the inflorescence is densely villous except the calyx, and the leaves are practically entire while in a flowering specimen from Gan-chouen the inflores- cence is nearly glabrous, though the base of the peduncle and the branchlets are densely villous, and the leaves are entire, but 23 distinctly denticulate in a specimen from the same locality with an old densely villous inflorescence. The latter apparently is the specimen upon which Cardot based his var. denticulata, while the flowering specimen may be of the collection of 1910 also from Gan- chouen mentioned by him, but not referred to by Léveillé. Photinia serrulata Lindley in Trans. Linn. Soc. xm. 103 (1821), ; excl. syn. Crataegus glabra bier Stranvaesia Argyi Léveillé . Art. Barcelona, Sét- 3, = 560 (Cat. PL Kiang Sou, 20) (i916), pro synon. 8. Calleryanae - 848-66] tre Kiangau: without locality, oh yleds {i (holotype of Stranvaesia Ae photo, in suse 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE = 307 by | Stranvaesia Argyi seems to be only an herbarium name given by _ Léveillé to a specimen without any other data except Léveillé’s P label “Stranvaesia Argyi Lévl.” He published it later erroneously __asasynonym of S. Calleryana. Photinia Bodinieri Léveillé in Fedde, Rep. Spec. Nov. iv. 334 pes —Cardot i i Bull. Mus. Nat. Hist. Paris, xxv. 402 (1919). Hiptage Esquiro ee Fedde, dy igs Nov. x. 372 (1912); et. Kouy-Tehéou, 27 oe _—Synon. serrulata Léveillé, Fl. Ko ay tebeue. 349 (1915), quoad synon. et specimen cit no. 2256.—Non Lindle ) HINA. Kweichou: environs de Kouy-yang, mont. du bad Collége, E. Bodinier, no. 2256, May 18, 1898 (holotype of Photinia __ Bodinieri; photo. in A. A.); Choui-t’eou, route de Tin-fan 4 Lo-fou, bs alt. 900 m., J. Esquirol, no. 2097, May 4, 1900 (bel arbre, fleurs bys blanches; holotype of Hiptage Esquirolii; merotype in A. A.). Ven Ih his Flore du Kouy-Tchéou Léveillé refers his Photinia Bodini- _ erito P. serrulata Lindl., but as Cardot (1. c.) points out, P. Bodi- nieri can be readily distinguished by the larger flowers and the more - connate styles. ris Photinia Mairei Léveillé in Monde des PI. ser. 2, xvii. 28 faded ae ioe Pl. Yun-Nan, 230 (1917). ne A. Yunnan:rochers, brousse des montagnes & reer ti, 3100 m., E. E. Maire, May 1912 (holotype; merotype in A. A.). _ OF this plant I have before me a rather meagre specimen with a 4 a crowded paniculate inflorescence not leafy at base; the leaves me we elliptic-obovate, crenate-serrulate and 6-7 cm. long. The miorescence recalls that of Eriobotrya, but the nearly superior Temoves it from that genus. I do not know any species ? in Eriobotrya or in Photinia with which it could be compared. ’ Stranvaesia Davidiana Decaisne in Nouv. Arch. Mus. Paris, 179 (1874). He ‘Pirus Cavaleriei Léveillé in Fedd Rep. Spec. Nov. x1. 66 (1912); Kouy-Tehéou, 95 350 (1915).—S ake a nov. Cmxa. Kweichou: Pinfa, J. Cavalerie, no. $569, Oct. (holotype of Pirus Cavaleriei; merotype in A. A.). | Cavaleriei Lévl. comb. nov. a a Cavaleriei Lé « teeny Fedde, Rep. Spec. Nov. x. 372 wee | Kouy-Tehéou, 2 ort (1914). ‘riobotrya Brackloi Hand.-Mazz. var. atrichophy ila Handel Mose ee oe Akad. Wiss. Wien Math. Naturw. KI. ¥1922, no. 12, p. Ce. Forts, 16, p, 2) (1922).—-Syuon. 80¥. ee Kweichou: Pin-fa, montagne en pente, J. Cava- «20. 3220 May 20, 1907 (arbre de 9 m. de hauteur; fl. blanches type of Hiptage Cavaleriei; widieit in A. A). 308 JOURNAL OF THE ARNOLD ARBORETUM [von. xm In the original place of publication the citation of specimen and locality is lacking, but is given in Flore du Kouy-Tchéou (I. c.) As Eriobotrya Brackloi var. atrichophylla Hand.-Mazz. becomes a synonym of E. Cavaleriei (Lévl.), typical E. Brackloi must be considered a variety of that species and should bear the name Eriobotrya Cavaleriei var. Brackloi (Hand.-Mazz.), new comb. (E. Brackloi Handel-Mazzetti in Anz. Akad. Wiss. Wien Math.- Naturw. Kl. 1922, no. 12, p. 102 [Pl. Nov. Sin. Forts. 16, p. 2] [1922]). Eriobotrya Seguini (Lévl.) Cardot apud Guillaumin in Bull. Soc. Bot. France, txx1. 287 (1924), ““Seguinii.” Symplocos Seguini Léveillé in Fedde, Rep. Spec. Nov. x. 431 (1912); Fl. Kouy-Tchéou, 408 (1915). Bere pseudo-Raphiolepis Cardot in Lecomte, Not. Syst. m. 371 Cuina. Kweichou: environs de Ou-la-gay et de Hoang-ko- chou, J. Seguin in herb. Bodinier, nos. 2262, April 1898, and 2617, March 1, 1899 (syntypes of Symplocos Seguini [in herb. Edinb.] and of Eriobotrya pseudo-Raphiolepis [in herb. Paris]; photo. and fragments of no. 2617 [herb. Edinb.] in A. A.). This species was first described by Léveillé as Symplocos Se- guint, but the herbarium specimens bear the name Lindera Seguin Lévl. in his own handwriting. On duplicates of the same numbers in the Paris Herbarium Cardot based his E. pseudo-Raphioleprs. The connection was discovered when Guillaumin writing his paper “Observations sur les Symplocos d’Extréme-Orient” examined the type specimen of Symplocos Seguini and found that it was not a Symplocos, but an Eriobotrya. Amelanchier asiatica (Sieb. & Zuce.) Endlicher apud Walpers, Rep. u. 55 (1843).—Nakai, Fl. Sylv. Kor. v1. 19, t. 1 (1916). Pirus Taqueti Léveillé in Fedde, Rep. Sp. Nov. vir. 199 (1909). Prrus Vanioti Léveillé, op. cit. 200 (1909). Korea: Hallaisan, in silvis, U. Faurie, nos. 1559, 1560 (ex Léveillé), 1561, May to July 1907 (syntypes of Pirus Taquett; isotype of no. 1561 in A. A.):; same locality, E. Taquet, mG 103, Oct. 1907 (syntype of P. Taqueti); Quelpaert, in silvis, U. Faurie, no. 1557, May 1907 (holotype of Pirus Vanioti; isotype in herb. A. A.). } Pay Wighgeeehe) Pirus Taqueti and P. Vanioti had already been identified with Amelanchier asiatica by Nakai (1. c.). Malus Sieboldii (Reg.) Rehder in Sargent, Pl. Wilson. u. 293 y (1915). ; - : : ce 1988], REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 309 Crataegus Cavaleriei Léveillé & Vaniot in Bull. Soc. Bot. France, ak 58 Beyonsl: Fl. may Tehéou, 346 (1915), excl. no. cit. 1303.—Syn Pirus ‘subcrataegifolia Léveillé in Fedde, Rep. Spec. Nov. vir. 199 (1909). inia rubro-lutea Léveillé in Fedde big ot eae Nov. rx. 460 (1911); Fi. KouyTehéon, 349 (1915) _—Synon. n Crataegus it Léveillé in Fedde, Rep. Tied. Nov. x. 877 (1912).— Synon. yg Pirus Esquiroliit Léveillé in Fedde, Rep. Spec. Nov. xu. 189 (1913); Fi. Kouy-Tchéou, 350 (1915). Symon. nov. M Toringo Siebold, Cat. Rais. 1. 4 (1856), nomen.—Nakai, Fi. Sylv. Kor. vr. 35, t. 10 (1 916). Korga. Quel paert: Hallaisan, U. Faurie, no. 1558, June ___ 1907 (holotype of Pirus subcrataegifolia and syntype of Crataegus Taqueti); in silvis Hallaisan, 900 m., E. Taquet, no. 2828, Oct. | 1909 (syntype of Crataegus Taqueti; isotype in A. A.); in silvis - Yengsil, 1000 m., E. Taquet, no. 4220, Aug. 12, 1910 (syntype of Crataegus Waguel:: isotype in A. A.). Cama. Kweichou: Pin-fa, montagne, J. Cavalerie, no. a) 98, July 23, 1902 (holotype of Crataegus Cavaleriei; photo. in A. A.); same locality, J. Cavalerie, no. 1304, April 2, 1902 (holotype of Pirus Esquirolii; photo. in A. A. ); without locality, J. Cavalerie, a ee Nov. 13, 1907 (holotype of Photinia rubro-lutea; merotype | mAA), talus Steboldii seems to be widely distributed in southeastern My - Tanging west to Kweichou, as the following additional bs specie ns in the herbarium of the Arnold Arboretum show: Che- Stang (N. T. Liou, no. 442), Kiangsi (H. H. Hu, no. 943), Kwang- a ag Tsiang, no. 1381), Kwangsi (R. C. Ching, no. 5935), Hunan del-Mazzetti, no. 11787, Pl. Sin. cur. Handel-Mazzetti,no. 94) and veichou (Handel-M azzetti, no. 285, Y. Tsiang, nos. 5006 and 5521). mee does not seem to occur in northern China and its Chinese zig shows no connection with the range of the species in Korea Pe Japan, but I can see no marked difference between the plants the two areas and have to consider them conspecific in spite Beographical separation. . roma (Franch.) Schneider in Fedde, Rep. Spec. Nov. r Bodinieri Léveillé in Bull. Geos. Bot. xxv. 44 (1915); Cur Pl. Yeas Nan, 229 (1917).—Synon oo seals prés de la thoatibee de Kouy-Tchéou a e-li, E. Bodinier, April 9, 1897 (holotype of Cotoneaster Bodi- photo. and fragments in A. A.). easter Bodinieri was first identified with D. Delavayi by ae to a note on 1 the type staeret mn cn ey 310 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm ynia rufifolia (Lévl.), comb. n Firws ( "Gy rufifolia Léveillé in Bull Géo one xxv. 46 (1915); Cat Yun-Nan, 231 (1917), sphalmate “rubifolia. M. ia pile Schneider in Bot. Gaz. txm. 400 (1917).—Synon, nov. prion’ docynioides (Schneid.) Rehder in Jour. Arnold Arb. n. 58 1921 Cuina. Yunnan: flanc des coteaux arides 4 Lou-pou, 3050 m., E. E Maire, June 1912, “fleurs blanches; fruits j jaunes, oblong” (type of P. rufifolia; merotype in A. A. ). Pyrus Calleryana Decaisne, Jard. Fruit. 1. in textu ad t. 8 (1872). Pirus Mairei Léveillé in Fedde, Rep. Spec. Nov. x1. 189 (1913); Cat. Pl. Yun-Nan, 231 (1917). Cuina. Yunnan: Tche-hay, haies de la plaine, 2500 m., E. E. Maire, March 1911 (holotype of Pirus Mairei; merotype in A. A.). Rosa multiflora Thbg. var. adenophora Franchet & Savatier, Enum. Pl. Jap. 1. 134 (1875), nomen; 1. 345 (1879).—Nakai, FI. Sylv. Kor. VIL. eat, t. 5 (1918). Rosa na Léveillé in Fedde, Rep. Spec. Nov. x. 432 (1912). Korea: “Corea media” U. Faurie, no. 330, July 1906 (holotype of R. Nakaiana; photo. in A. A.). Rosa multiflora var. quelpaertensis (Lévl.) Rehder & Wilson in Sargent, Pl. Wilson. 11. 335 (1915 3 | fiosa mokanensis Léveillé in Fedde, Rep. Spec. Nov. vit. 340 (1909), pro parte typica.—Willmott, Gen. Rosa, u. 511, t. (1914). Rosa quelpaertensis Léveillé in Reddo, Rep. Spec. Nov. x. 378 (1912). ae serene var. quelpaertensis W: Willmott, gree Rosa, u. 512, t. 1914 Rosa multiflora var. microphylla bata: Fl. Sylv. Kor. vu. 30 (1918), pro parte.—Non Franchet & Sava Koad. Quelpaert: Ma E. Taquet, no. 778, June 8, 1908 (in part; syntype of R. mokanensis; photo. and isotype in A. A.); in sepibus Hogno, E. Taquet, no. 2870, May 1909 (holotype | of R. quelpaertensis; isotype in A. A.); in petrosis secus vias, U. Faurie, no. 1567, May 1907; in sepibus, E. Taquet, no. 5587, June 1911; Hoatien, E. Taquet, no. 5590, June 1911 (nos. 1567, 5587 and 5590 in herb. Léveillé sub R. quelpaertensis). The type specimens of R. mokanensis consists of two branches, One representing R. multiflora and one R. Wichuraiana; the former must be considered the type of R. mokanensis, as the character “peduncu- lis glandulosis” applies only to this form. Nakai refers R. quelpaertensis to R. multiflora var. microphylla Franch. & Savatier, but this variety is described as having ovate leaflets not cuneate at base, while in var. pecan vent the leaflets are generally obovate and cuneate. 1932] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 311 Rosa multiflora var. cathayensis Rehder & Wilson in Sargent, Pl. Wilson. 11. 304 (1915). epee var. hypoleuca Léveillé, Fl. Kouy-Tchéou, 354 (1915), CHIN. Ds chou: Gan-chouen, commune, J. Cavalerie, no. sos8. “4. bicighes” (holotype of R. macrophylla var. hypoleuca; photo.in A. A.). Yunnan: haies, plaine de Long-tou, alt. 2400 m., E. E. Maire, “‘fl. rosées” (as R. clavigera forma in herb. Léveillé; photo. in A. A.). The flowers are white according to the collector and borne in ___ few-flowered corymbs; the leaflets are rather small and densely grayish pubescent beneath, but not whitish as the name seems to imply. Another specimen collected by Bodinier in 1888 on the “Pehod chan” and named R. macrophylla var. hypoleuca in léveillé’s herbarium but not in his handwriting belongs to R. it Koehne or a related species. Rosa multiflora var. carnea Thory in Redouté, Roses, 11. 67, t. ___ (1821).—Rehder & Wilson in Sargent, Pl. Wilson. 11. 305 (1915).— 8 Byhouwer in Jour. Arnold Arb. x. 86 (1929). , i Léveillé in Bull. Géog. Bot. xxv. 46 (1915); Cat. Pl. Yun- | Nan, 235 (1917). | Rosa Blinii Léveillé, 1, c. (1915); 1. c. 234 (1917). < Citra. Yunnan: haies, plaine de La-kou, alt. 2400 m., E. E. ae Maire, May [1910-14], “‘épineux, buissonant, haut 0.80 m.; fl. | roses ou rouges, inodores”’ (holotype of R. Lebrunei; photo. e in A, A); haies de la plaine & Tong-tchouan, alt. 2500 m., E. E. : Maire; Mai 1910-14 “petit rosier, épineux, fl. rouges, abondantes a -aagl ‘(type of R. Blinii; photo. in A. A.) 4, Rosa Wichuraiana Crépin in Bull. Soc. Bot. Belg. xxv. 189 (1886) 6) —Rehder r & Wilson i in Sargent, Pl. Wilson. m1. 385 (1915). & Rochebrune apud Crépin Bot. : in Bull. Soe. DAS ae x 823 { as7i), pro parte.—Nakai, Fl. Sylv. Hee. vu. 28, t. 2 Rosa Tague Léveillé in Fedde, Rep. Spee. Nov. vit. 199 (1909), pro i : me mokanensi Léveillé in Fedde, Rep. Spec. Nov. vu. 340 (1909), Ae “(oi6) eth oH Sylv. oh eye ee: * 33 “018), ‘tantum joe! synon. saat : Koura. Quelpaert: Hallaisan, E. Taquet, no. 102, October 312 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm The type specimen of R. Taqueti consists of two branches, one belonging to R. Wichuraiana and one to R. acicularis; according to the description “stipulae . . . margine glanduloso-fimbriatae; inflorescentia spicata flexuosa; fructus rubri, globosi, ut pedicelli glandulosi; styli hirti, columnares” the branch of R. Wichuriaiana should be considered the type. The type specimen of R. mokanensis also consists of two branches, of which, as shown under R. multiflora var. quelpaertensis, that representing var. quelpaertensis must be considered the type; the other specimen is an unusually small- leaved form of R. Wichuraiana. Taquet’s nos. 5588 and 5589 are named R. Fauriei Lévl. in Léveillé’s handwriting, but the type of that species belongs to R. Mazximowicziana Reg. and the paratype to R. acicularis Lindl. Nakai refers R. Wichuraiana as a synonym to R. Luciae Franch. & Rochebr., which, though very closely related, may be distin- guished by its more upright habit, thinner and narrower, usually acute, mostly 7 leaflets and smaller flowers. Rosa Maximowicziana Regel in Act. Hort. Petrop. v. 295, 378 (1878).—Nakai, Fl. Sylv. Kor. vu. 26, t. 1 (1918). rosa Fauriei Léveillé in Fedde, Rep. Spec. Nov. vit. 199 (1909), excl. specimine Faurie, no. 99. Korea: Quen-san, U. Faurie, no. 328, July 1906 (syntype of R. Fauriei; photo. in A. A.). The description of R. Fauriei is based almost exclusively on Faurie’s no. 328 which is the first specimen enumerated, and con- stitutes the type of species; from the second specimen, Faurie’s no. 99, only the description of the fruit was drawn and this specimen belongs to R. acicularis Lindl. The name R. Fauriei does not ap- pear on Faurie’s no. $28 in the herbarium of Léveillé. Rosa Brunonii Lindley, Ros. Monog. 120, t. 14 (1820).—By- houwer in Jour. Arnold Arb. x. 87 (1929). : Rosa clavigera Léveillé in Fedde, Rep. Spec. Nov. xi. 338 (1914); Cat. Pl. Yun-Nan, 234 (1917). Cuina. Yunnan: haies, plaine de Long-tan, alt. 2500 Mm.» E. E. Maire, May [1910-14] “grand rosier; feuilles vert luisant, pubes- centes au dessous; fl. blanches” (holotype of R. clavigera; photo. in A. A.). $ Rosa Rubus Léveillé & Vaniot in Bull. Soc. Bot. France, LV. Ly : (1908).—Willmott, Gen. Rosa, mu. 507 (1914).—Léveillé, Fl. Kouy-Tchéou, 354 ( 1915).—Byhoawer in Jour. Arnold Arb. *- 90 (1929). Rosa R: var. yunnanensis Léveillé 55 (1908); Cat. Pl. Yun-Nan, 235 (1917). in Bull. Soe. Bot. France, 1¥- oe ied nace ag h Mee, ee | 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 313 Cama. Kweichou: route de Pin-yang, L. Martin in herb. Bodinier, no. 2603, May 12, 1899, (holotype of R. Rubus; photo. in 4 A). Yunnan: montagnes, an bord de la plaine de Lo-pin- tehéou, E. Bodinier, no. 2603,! April 6, 1897, “branches sarmen- a aaa aad of R. Rubus var. yunnanensis; photo. and mero- “The ee labels in Léveillé’s herbarium of R. Rubus and R. Rubus var. yunnanensis seem to have been interchanged; they do not bear Léveillé’s names which are on separate slips in Léveillé’s handwriting, but the name R. Rubus var. yunnanensis does not ap- *) ged at all on the specimen; instead a slip with the the name R. a , @ species identical with R. microcarpa Lindl., has been e. pansy by mistake on that sheet. The specimen with the label of _ Bodinier’s no. 2603 agrees with the description of var. yunnanensis 7 in the characters “petala apice rotundata; folia utrinque viridia, -argute dentata,” while the sheet with Bodinier’s label of the Yunnan Specimen contains the plant which has crenate-serrate leaflets ‘@ayish beneath and emarginate petals, all characters given by Léveillé for typical R. Rubus. To the Arnold Arboretum Léveillé had sent in 1915 as R. Rubus a specimen which agrees with R. Rubus yunnanensis; this shows that even at that time confusion existed _ ow that Léveillé having more material of var. yunnanensis sent a % specimen of the variety instead of the type. "y Tet Pe ee ee ee ee E aera) IS TERE ree es Tee een BO ee Rosa Gentiliana Léveillé & Vaniot in Bull. Soc. Bot. France, caile 55 (19 8).—Willmott, Gen. Rosa, Wilson in Sargent, Pl. Wilson. 1. $12 (1915) —Léveillé in Mem, Acad. Ci. Art. Barcelona, ser. 3, x11. no. 22, p. 20 (Cat. Pl. Kiang- , i sat, Tl. Seu-Tchouen, 164, t. 59 (1918). ag NA. _Kiangsu: without precise locality, Ch. d Argy [1846— 64]; (holotype; fragments in A. A.). Lhd two oo belrigege cited above are obviously based on the ~ n which apparently is the type, but does not seem to be in Laveilé’s herbarium now. have seen Ro material of this species except some fragments te é in 1915 to the Arnold Arboretum? The species is oak ae in eastern and Central China, but the peculiar © of the leaflets stressed by Léveillé represents apparently an : form not constant for the species. - gps Bamber in pencil pinpaek, not in Bodinier’s, “ego 'Sbecies are not now in the Elerbasiam us, B- adenoclada snd poss of ‘t when she was G Rosa and probably became Lé ile hank aig work Tue GENUS 314 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Rosa adenoclada Léveillé in Fedde, Rep. Spec. Nov. x. 481 (1912); Fl. Kouy-Tchéou, 353 (1915).—Willmott, Gen. Rosa, 1, 517, t. (1914), Rosa Gentiliana Rehder & Wilson in Sargent, Pl. Wilson. 1. 312 (1915) quoad syn. R. adenoclada.—Non Léveillé. Rosa Gentiliana var. adenoclada Léveillé, Cat. Pl. Yun-Nan, 234 (1917), nomen. Cuina. K weichou: Grotte de Gai-kio, J. Esquirol, no. 2100, May 3, 1910 “couleur rouge”’ (holotype). The material of this species in Léveillé’s herbarium is very meagre consisting of a single leaf, a piece of a leafless branch and some detached flowers and leaflets, but there is a good figure of a flowering branch in Miss Willmott’s book. It is a very puzzling Rose and apparently does not belong to R. Gentiliana where it was placed by my former colleague, Mr. Wilson, and myself. It differs from that species in the pilose, more or less free styles, in the red color of the flowers and in the slightly bristly and stipitate- glandular branch. It may possibly be a hybrid of R. Rubus Lévl. and R. chinensis which would account for the color of the flowers and the character of the styles. It is also somewhat similar to R. lucidissima Lévl. Rosa Willmottiana Léveillé in Fedde, Rep. Spec. Nov. xt. 299 (1912)— Willmott, Gen. Rosa, 521, t. (1914). 014). Kosa Charbonneaui Léveillé in Fedde, Rep. Spec. Nov. xu. 338 (191 ). Cuina. Yunnan: brousse et haies a Long-ky, 700 m., E E. Maire, June 1911, “Eglantier épineux a long rameaux; fi blanches” (holotype of R. Willmottiana; photo. in A. A.); plaine de Long-tan, 2500 m., E. E. Maire, May 1913 “rosier épineux + long rameaux; fl. d’abord blanches, puis sanguines” (holotype R. Charbonneaui; photo. in A. A.). beets: Rosa Willmottiana represents a form with rather large eaten i up to 8.5 em. long and not reticulate beneath, while R. Charbon- - neaut has much smaller more finely serrate leaflets reticulate 7 beneath. uy Rosa microcarpa Lindley, Ros. Monog. 130, t. 18 (1820)— : Rehder & Wilson in Sargent, Pl. Wilson. 1. 314 (1915). ae : Rosa sorbiflora Focke in Gard. Chron. ser. 3, xxxvl. 227, fig. 96 (1908)- x Léveillé, Fl. Kouy-Tchéou, 354 (1915). 76 Rosa Chaffanjoni Léveillé & Vaniot in Bull. Soc. Bot. France, £¥- > 1908). ne Dodiniers Léveillé & Vaniot, 1. ¢. (1908).—Willmott, Gen. Rosa, 1, 485, t. (1914) —Léveillé, Fl. Kouy-Tehéou, 353 (1914). 1932] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 315 Rosa Esquirolii Léveillé & Vaniot in Bull. Soc. Bot. France, tv. 56 (1908).—Willmott, Gen. Rosa, 11. 485, t. (1914).—Léveillé, Fl. Kouy- Tchéou, 353 (1915). Rosa Cavaleriei Léveillé in Fedde, Rep. Spec. Nov vi. 61 (1910). Cova. Kweichou: environs de Kouy-yang, mont du Collége, haies, prés des villages, J. Chaffonjon in herb. Bodinier, Rosa Banksiae Aiton, Hort. Kew. ed. 2, 1. 258 (1811).— Byhouwer in Jour. Arnold Arb. xr. 91 (1929). ae f. aculeata Léveillé, Cat. Pl. Yun-Nan, 234 (1917), nomen. Rosa Banksiae f. subinermis Léveillé, 1. c. (1917), nomen. Fosa Banksiae f. albiflora Léveillé, 1. c. (1917), nomen. * HINA. Kweichou: Kouy-yang, mont. du Collége, dans les haies, prés des villages, J. Chaffanjon in herb. Bodinier, no. $259, May 15, 1898 “fl. blanches, souvent doubles’; Gan-chouen, : Roe walerie, no. 3943, May 1910. Yunnan: Tong-tchouan, #3 haies de la plaine, alt. 2500, E. E. Maire, April [1910-14], “fl. blanches doubles”; environs de Yun-nan-sen, E. Bodinier, March 41897, “8. tantét simples, blanches, trés souvent doubles.” 2 a2 are no specimens in the Léveillé herbarium named f. é Pro and f. subinermis; these names were apparently taken a oo f. subinermis fl. pleno and f. aculeata fl. pleno albo lt Bot. Gard. Edinb. v. 65, 66 [1913]). Also of f. albiflora ae i E JS no Specimen and this name may belong either here or to 4a © White, single-flowered R. Banksiae f. normalis Regel. ' Rosa Banksiae f. lutea Lindl. Bot. Reg. xm. t. 1105 (1827).— Pe) Saat Jour. Arnold Arb. x1. 92 (1929). : Cre ae f. luteiflora Léveillé, cat. Pl. Yun-Nan, 234 (1917), nomen. 2500 - Yunnan: haies de la plaine 4 Tong-techouan, alt. . ae E. Maire, April [1910-14], “fl. jaunes, doubles, odor- i8 no specimen labeled f. Iuteiflora in Léveillé’s herbarium, name doubtless belongs to the yellow double flowered form Tpresented in his herbarium. : 316 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm i Rosa lucidissima Léveillé in Fedde, Rep. Spec. Nov. rx. 444 (1911); Fl. Kouy-Tchéou, 354 (1915).—Willmott, Gen. Rosa, u. 519, t. (1914).—Rehder & Wilson in Sargent, Pl. Wilson. 1. 320 (1915), sub R. chinensis f. spontanea. Cuina. Kweichou: Pin-fa, précipices, J. Cavalerie, no. 990, April 13, 1903, “fl. rouges, odorantes” (holotype; photo. in A. A.); Gan-chouen, Hin-y-fou, J. Cavalerie, nos. 3927, 3942, June 1912. This peculiar Rose belongs apparently near R. laevigata, but differs in the less prickly and bristly branchlets and peduncles, in the adnate stipules and the red flowers; it probably is a hybrid of this species and of R. chinensis from which it is easily distinguished _ by the 3-foliolate leaves and the bristly receptacle, pedicels and branchlets. Cavalerie’s nos. 3927, 3942 designated by Léveillé in his herbarium as R. lucidissima var. lack the bristles and rep- resent obviously a form approaching R. chinensis, but differing in the 3-foliolate leaves with more coriaceous leaflets. Rosa luci- dissima is very similar to X R. anemonoides Rehd. (in Jour. Arnold Arb. m1. 13. 1921), supposed to be a hybrid between R. laevigata and R. odorata; it differs from R. lucidissima chiefly in the larger pink flowers and in the stipules being adnate only about one half their length. Rosa odorata Sweet, Hort. Suburb. Lond. 119 (1818).—Rehder & Wilson in Sargent, Pl. Wilson. n. 338 (1915).—Léveillé, Cat. Pl. Yun-Nan, 235 (1917)—Byhouwer in Jour. Arnold Arb. x1 % (1929). Rosa gechouitangensis Léveillé in Fedde, Rep. Spec. Nov. xt. 299 (191). aon faery [ste] Léveillé, 1. ¢. (1912).—Willmott, Gen. Rosa, ™ Rosa t vanensis Léveillé in Fedde, Rep. Spec. Nov. x1. 300 — (1912).—Willmott, Gen. Rosa, 11. 523, t. (1914). : Cuina. Yunnan: Ge-choui-tang, 3 flancs des cotealt — calcaires, 2450 m., E. E. Maire, April 1911, “Rose thé semidouble, — a long rameaux” (holotype of R. gechouitangensis; photo. in A. A); z Ou-long, haies des tertres, 2500 m., E. E. Maire, April 1911, “petit — rosier épineux; fl. roses” (holotype of R. oulengensis; photo. in A. i A.); haies de Tong-tchouan, alt. 2500 m., E. E. Maire, April 191, “rosier de Bengale; fl. roses? (syntype of R. tongtchouanensis); | haies de La-kou, alt. 2400 m., E. E. Maire, March 1911 “ros ‘ épineux, 4 long rameaux; fl. blanches doubles, inodores” (syntyPe 4 of R. tongtchouanensis; photo.in A. A.). a The three species of Léveillé’s cited above are all double-flowere! j ; forms of R. odorata and therefore must be either cultivate¢ * 7 7 escaped from cultivation. — ; 3 y Ree: _ 1982] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE: 317 Rosa odorata var. gigantea (Collet) Rehder & Wilson in Sargent, PL Wilson. n. 338 (1915).—Léveillé, Cat. Pl. Yun-Nan, 235 (1917). a : in Peter rep Arb. x1. 94 Laat . ex Rehder & Wilson in Sargent “Vien. Il. 389 ( aha et i Level Cat. Pl. Ye un-Nan, 235 asin, c Cua. pine unnan: environs ate Vaisdieich: dans la mont., bord des ruisseaux, F. Ducloux, March 28, 1897, “grande liane _ Sarmenteuse; gr. fl. blanches” (holotype of R. Duclouxii; photo. A. A.), . mA, : ne air var. Gmelin (Bge.) C. A. Meyer, in Mém. Acad. Sci . St. Pétersb. Tc 6, (Ueber Zimmtros.) (1847).—Nakai ae Tokyo Bot Mag. xxx. "341 (1916); Fi. Sylv. hat vu. 37, t. 10 (1918). : - a Fait Leveilé i in Fedde, Rep. Spec. Nov. vit. 199 (1909), quoad Rowe Ts et Tau Léveillé in Fedde, we AES Nov. vit. 199 (1909), pro Parte, quoad specimen oan Rosa Korsakovionsis Léveill iv Redde, | Rep. Spec. Nov. x. 378 (1912).— __, Willmott, Gen. Rosa, 11. B17, t t. (1914). acicularis var. ¢ Nakai in Tokyo Bot. Mag. xxx. 241 (1916) et Fl. Sylv. Kor. vir. 38, t. 11 (1918) pro parte, synon. citato exclud. OREA: secus vias regionis interioris, U. Faurie, no. 99, Sept. 4, _ M01 (eyntype of R. Fauriei; photo. in A. A.); Hallaisan, E. Taquet we, 102, Oct. 1907 (in part, as to branch without fruit; syntype of a Slant Photo. in A. A.); in sepibus Hallaisan, 1700 m., E. Taquet, is 74, 7, 1908, et Hallaisan, 1700 m., sed in horto missionis at: E. Faquet, no. 4228 May 10, 1910 (both as R. Taqueti in veillé). : eirea Korsakof, U. Faurie, no. 570, Aug. 3 (holotype of R. ‘ie orsakoviensis). 2 le of R. Fauriei is Faurie’s no. 328 which represents R, vana Reg. The type of Rosa Taqueti is identical with » the type specimen, Taquet’s no. 102, consists of snches, one with fruits on which apparently the description i under R. Wichuraiana p. 312) and which represents R. “ana and two sterile specimens which belong to R. acicu- te only Part of the description which is clearly applicable to Specimens i is “caules niger et lucidus,” the rest ap- ~— or both. Besides the type there are two ns, Taquet’s no. 774 and 4228 labeled R. Taqueti in iting ‘which both belong to R. acicularis, but are * original description. Apparently Léveillé later as to the real identity of his R. Taqueti and ap- SEP to the wrong species. The same mistake was made Seg therefore his name though meant for a form of ecomes one according to the pene brinioe 318 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm synonym, a synonym of R. Wichuraiana. The form distinguished as var. Taquetii by Nakai represents a form devoid of bristles. Rosa Marretii Léveillé in Fedde, Rep. Spec. Nov. vu. 981 (1910).—Willmott, Gen. Rosa, 11. 495, t. (1914).—Nakai in Tokyo Bot.. Mag. xxxvi. 63 (1922). Rosa rubrostipullata Nakai in Tokyo Bot. Mag. xxx. 242 (1916); Fl. Sylv. Kor. vu. 40, t. 13 (1918). SAGHALIN: circa Korsakof, U. Faurie, no. 571, Sept. 10, 1908 — (holotype of R. Marretii; isotype in A. A.), no. 572, Aug. 1908 (as R. Marretii in herb. Léveillé; as R. sp. in A. A.). ree Faurie’s no. 572 not cited with the original description is in bloom; it has like the specimen of no. 571 in the herbarium of the ; Arnold Arboretum slightly broader leaflets and lacks the peculiar | ascending prickles at the base of the branchlets present in the type specimen. Rosa Marretii differs from R. davurica Pall. chiefly in the green, glabrous or glabrescent and not glandular under side of the leaflets and in the ascending, not recurved prickles. Rosa Davidi Crép. var. elongata Rehder & Wilson in Sargent, | Pl. Wilson. 11. 323 (1915).—Byhouwer in Jour. Arnold Arb. x — 99 (1929). Rosa Parmentieri Léveillé in Fedde, Rep. Spec. Nov. xm. 339 (1914); Cat. Pl. Yun-Nan, 235 (1917). ee 64 Cuina. Yunnan: brousse de Io-chan, $400 m., E. E. Maire, ; June 1913, “rosier épineux, buissonant; fl. blanches” (holotype of t R. Parmentieri; photo. in A. A.). | Rosa sertata Rolfe in Bot. Mag. cxxxix. t. 8473 (1913).—By- j houwer in Jour. Arnold Arb. x. 100 (1929). os | lochanensis Léveillé in Fedde, Rep. Spec. Nov. xu. 339 (1914); Cat. Pl. Yun-Nan, 234 (1917). ee | Cuina. Yunnan: Mont Io-chan, alt. 3300 m., E. E. iets ui June 1913 “‘rosier 4 rameaux fins, épineux, buissonant, haut 0.80 ™ ij fl. rouges” (holotype of R. iochanensis: photo. in A. A.). : Rosa omeiensis Rolfe in Bot. Mag. cxxxvim. t. 8471 epee oO Rehder & Wilson in Sargent, Pl. Wilson. u. 331 (0) a by Cat. Pl. Yun-Nan, 285 (1917).—Byhouwer in Jour. Arnold At. — x. 102 (1929). ‘ae Kosa Sorbus Léveillé in Fedde, Rep. Spec. Nov. xu. 338 (1914). | Rosa sericea f. aculeata Focke path illé, Cat. Pl. Yun-Nat, (1917), nomen. Rosa sericea f. eglandulosa Léveillé, 1. ¢. (1917), nomen. 2 Rosa sericea f. inermis Léveillé, 1. ¢. (1917), nomen. ee Cuina. Yunnan: brousse 4 mi-mont du Io-chan, alt. age oi i E. E. Maire, June 1913, “grand rosier épineux buissonant, 4 fl. blanches,”’ (holotype of R. Sorbus; photo. in A, A.). 4 ; i ; Ne 35 et I at _ W882] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 19 Rosa Mairei Léveillé in Fedde, Rep. Spec. Nov. x1. 299 (1912); Cat. Pl. Yun-Nan, 235 (1917).—Willmott, Gen. Rosa, um. 521 (1914)—Rehder & Wilson in Sargent, Pl Wilson. 11. 343 (1915).— Byhouwer in Jour. Arnold Arb. x. 103 (1929). Cana. Yunnan: collines arides autour de Tong-chouan, 2600 m, E. E. Maire, April 1911, “‘rosier buissonant, 4 fortes épines tri- angulaires; fl. Erortes ” (holotype; photo. in A. nae 58 (1908). NA. Kiangsu: Song-kiang-fan, Ch. d’ Argy (5 specimens, syntypes of P. Persica var. lasiocalyx; photo. in A. A.); Tao-chow; oo Ch. d’ Argy (holotype of P. daemonifuga; photo. in Prunus persica var. lastocalyx seems scarcely different from the type and P. daemonifuga was reduced to synonymy by Léveillé himself in 1916 (1. c.). Prunus Persica var. longistyla Léveillé in Bull. Bot. Soc. France, | a5 58 (1908); in Mem. Acad. Ci. Art. Barcelona, ser. 3, 560 (Cat. | M1 Kiang Sou, 20) (1916). - Gowa. Kian Su: without locality, Ch. d’Argy ? (holotype; begs inA. A), 2 is: ‘*ccount of its long style this variety resembles P. kansuensis _ Mehd. which differs, however, in its slenderer branches, glabrous vine buds less pubescent sepals glabrous inside, white or nearly begs tather smaller flowers and i is only known as a wild plant from iz "ota frontier at altitudes of 6000-8000 ft.; it also eres y from P. Persica in its stone not being pitted. | ey 1 5 Variety is apparently a form of the cultivated Peach, be the length of the style is an unusual character in this species. cy MUS japonica var. Nakaii (Lévl.) Rehder in Jour. Arnold Ee : a oe (i921). Los) «ps Nakai Léveillé in Fedde, Rep. Spec. Nov. vu. 198 : a a" aehne Day sie gt FL Pilon. 1. 267 (1912).—Nakai, cont t Sylv. Kor. v, 36, t. 22 (1516) 6 Ck *n-san, in lacunis montium, U. Faurie, no. 334, July he Molotype of P. Nakai; isotype in A. A.). Pek ponienis Koehne in Sargent, Pl. Wilson. 1. 200 (1912). myrtacea Léveillé j Bot. xxv. 45 (1915); Cat. Pl. = 234 (1917). wens betes 320 JOURNAL OF THE ARNOLD ARBORETUM [vou. xu Cuina. Yunnan: brousse des montagnes 4 Pe-long-tsin, 3200 m., E. E. Maire, June 1912 (holotype of P. myrtacea; photo, — in A. A.). Prunus odontocalyx Léveillé in Bull. Géog. Bot. xxv. 45 (1915); Cat. Pl. Yun-Nan, 234 (1917). Cuina. Yunnan: haut plateau de Ta-hai-tse, 3200 m., E. E. Maire, May 1912 (holotype; merotype in A. A.). This species I have not been able to identify with any previously described Prunus. It apparently belongs in Koehne’s Series Oxyodonta of his subsection Ceraseidos, but differs from the species known to me in its longer and narrower leaves (not yet fully un- folded) quite glabrous except a fugaceous silky pubescence along the midrib and in the 2-3-flowered racemes on a short peduncle 3-5 mm. long with narrow oblong-lanceolate caducous bracts to 1 em. long. Prunus Padus Linnaeus, Spec. Pl 473 (1753).—Nakai, Fi. Sylv. Kor. v. 18, t. 3 (1916). Prunus Fauriei Léveillé in Fedde, Rep. Spee. Nov. vit. 198 (1909). Korea. Quelpaert: Hallaisan, 1500 m., U. Faurie, - 1549, June 17, 1907, “rara, e basi ramosa, 2 m., alta 5—6 m. lata (holotype of P. Fauriei; photo. in A. A.). Nakai (op. cit. 15) cites P. Fauriei as a synonym under P. d Buergeri Miq. (= P. Buergeriana Miq.), but under P. Padus he cites “P. Fauriei Lévl. in litt. fide Taquet.” Koehne (in Fedde, Rep. Spec. Nov. xm. 135. 1913) states that P. Fauriet hardly 2 differs from P. Padus, to which I agree with the original specime? from the Herbarium Léveillé before me. Prunus Padus L. var. seoulensis (Lévl.) Nakai, Fl. Sylv. Kee v. 19 (1916). Prunus seoulensis Léveillé in Fedde, Rep. Spec. Nov. vu. 198 (1909). Korea: Namsan Seoul, U. Faurie, no. 331, July 1906 (holotype of P. seoulensis; isotype in A. A.). Koehne (in Fedde, Rep. Spec. Nov. xu. 134. 1913) states 1h P. seoulensis hardly differs from typical P. Padus and that 4 possibly belongs to var. commutata Dipp. Prunus pubigera (Schneid.) Koehne in Sargent, Pl. Wilson. ' q 67 (1911). Prunus (Padus) Vanioti Léveillé in Bull. Géog. Bot. xxv. 45 (1915) Cat. Pl. Yun-Nan, 234 (1917). : Cuina. Yunnan: flane aride des montagnes 4 Ma-kon® 2700 m., E. E. Maire, April 1912 (holotype of P. Vanioti; merotyP® in A. A.). 2 p il 1092] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 321 Prunus Maackii Rupr. var. diamantina (Lév].) Koehne in Fedde, Rep. Spee. Nov. x. 134 (1913). Prunus diamantina Léveillé in Fedde, Rep. Spec. Nov. vu. 198 (1909). _ Korea: in Monte des Diamants, U. Faurie, no. 332 (holotype of P. diamantina). Nakai (Fl. Sylv. Kor. v. 17. 1916) refers P. diamantina to P. _ Maackii as a synonym without recognizing it as a variety. he LEGUMINOSAE _ Albizzia kalkora (Roxb.) Prain in Jour. As. Soc. Bengal, txvi. ai (807), __ Albizzia Esquirolii Léveillé, FI. Kouy-Tchéou, 224 (1914).—Synon. nov. c. Cua. K weichou: Tou-chan, J. Cavalerie in herb. Bodinier, ‘lo. 2634, June 1899 (ex Léveillé; syntype of A. Esquiroliz); same locality, J. Esquirol, no. 431, June 1905 (syntype of A. Esquirolii; __ Acacia Delavayi Franchet, Pl. Delavay. 194 (1890). Peps Acacia Cavaleriei Léveillé, FI. Kouy-Tchéou, 222 (1914), nomen.—Synon. Ee Cava. Kweic hou: Mou-you-se, J. Cavalerie, no. 2041, June _ 1904 “tige lianeuse court sur les murs” (holotype of A. Cavaleriei; - Photo. in A. A.). oo Thave not seen the type of A. Delavayi, but Cavalerie’s specimen oo: w 'is in bloom seems to agree well with Delavay’s description ex- “pt that the only leaf present has 6 pairs of pinnae and the leaflets ae up to 1 cm. long; Franchet describes also the fruit, but Cava- : § specimen has only flower-heads mostly in bud. ;. Cnginal description of Acacia Cavaleriei I have not been able ate; possibly Léveillé did not publish a description. | is Gagnepain in Lecomte, Not. Syst. m. 181 - Kouy-Tchéou, 227 (1914). rt Léveillé in Fedde, Rep. Spec. Nov. xt. 31 (July 1, 1912). densiflora Franchet, Pl. Delavay. 191 (1890).— Bou. S°UY-Tehéou, 227 (1914). “ma Cavalerie: Léveillé in Fedde, Rep. Spec. Nov. x1. 31 (1912). tion .. ©! Chou: Lo-fou, J. Cavalerie, no. 3676, Aug. and » 1909 (holotype; photo. in A. A.); Pin-fa, J. Cavalerie 322 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm no. 659; descente du fleuve, J. Esquirol, nos. 507, 890 (ex Léveillé, Fl. Kouy-Tchéou). Bauhinia Cavaleriei was referred by Léveillé as a synonym to B. densiflora in his Flore du Kouy-Tchéou. ia yunnanensis Franchet, Pl. Delavay. 190 (1890). Bauhinia altefissa Léveillé, F1. sony ebay ey, 226 (1914); Cat. Pl. Yun- Nan, 152 (191 6).—Synon Cuina. Kweichou: Rica y: -se, J. Cavalerie, no. 3908, June 1912 (syntype of B. altefissa; photo. in A. A.); without locality, J. Esquirol, no. 511 (syntype of B. altefissa; photo. in A. A.). Bauhinia altefissa agrees perfectly with B. yunnanensis, but Es- quirol’s no. 3056 “‘derriére le camp de Lo-hou, Juillet 1911” cited by Léveillé under the original description of B. altefissa does not belong here; the specimen is sterile and probably belongs to B. densiflora Franch. Bauhinia aurea Léveillé in Bull. Soc. Bot. France, tiv. 368 (1907); Fl. Kouy-Tchéou, 226 (1914). HINA. K weichou: rochers et bois des environs de Lo-fou, J. Cavalerie, no. 2614, Nov. 1905 (holotype; photo. in A. A.); Pin-fa, J. Cavalerie, no. 2614; Lao-ten, E. Bodinier, no. 43; Ou-la-gay, J. Seguin; confluent des deux riviéres de Lo-kouy; J. Esquirol, no. 3226, 1898-1912. (All enumerated by Léveillé, FI. Kouy-Tchéou-) This species resembles in the size, tomentum and general outline of the leaf B. Vahlii Wight et Arn., but differs in the somewhat acutish not rounded lobes, and in the narrow, not open, basal sinus; also the pod which I have not seen agrees according to the descrip- tion with that of B. Vahlii. I have not seen the additional specimens cited in the Flore du Kouy-Tchéou. A similar form also only fruit has been collected in Kweichou by Y. Tsiang (no. 7289) ane Gan-wu, Lo-hu on the Kwangsi border, but the leaves are hau date to nearly truncate at base, while the lobes are acutish. Pterolobium punctatum Hemsley in Jour. Linn. Soc. xxi. 207 (1887). Prosopis Esquirolii Léveillé, Fl. Kouy-Tchéou, 242 (1914). “pow, J Cuina. Kweichou: de Tong-fong-tcheou & igen : Esquirol, no. _ 166, Aug. 1904 (holotype of Prosopis Es merotype in A. A.). Sophora japonica Linnaeus, Mant. 1. 68 (1767).—Léveillé, Fi. Kouy-Tchéou, 243 (1914); Cat. Pl. Yun-Nan, 161 (1916). Ormosia greenery Léveillé, Fl. Kouy-Tchéou, 240 (191 4) Sree Sophora Mairei Léveillé in Bull. Géog. Bot. xxv. 48 (1915); 20 Mairei Pamp. (1910) gery nov. Cmva. Kweichou: Lao-ouang-tchai, J. Esquirol, no. 3840, teers 1912, “arbre 6 m.; fl. blanches” (holotype of Ormosia Esqui- _ rolii; merotype in A. A. ). Yunnan: laine de Tong-tchouan, alt. $00 m., E. E. Maire, Aug. [1911-14], “‘gros et grand arbre; fleurs _ blanches inodores” (holotype of S. Mairei; merotype in A Sophora glauca De Candolle in Ann. Sci. Nat. ser. 1, rv. (1824) 8 (1825).—Gagnepain in Lecomte, Not. Syst. m1. 117 (1915).— Level Cat. Pl. Yun-Nan, 161 (1916). ; Indigofera Mairei Léveillé in Fedde, Rep. Spec. Nov. x11. 190 (1913). " | Cua. Yunnan: couteaux calcaires arides, derriére La-kou, if: 450 m., E. E. M aire, June 1911, “fleurs violettes” (syntype of x “Indigofera Mairei; photo. in A. A.); plaine de Kiao-kia, alt. 400 m., _ ELE. Maire, July 1911, “‘fleurs violettes” (syntype of Indigofera Pies photo. in A. A.). Indigofera Mairei has been identified already by Gagnepain (I. c.) 3 = Sophora glauca. ‘Sophora glauca var. albescens Rehder in Sargent, Pl. Wilson. : am. 447 (1917), ne Cavaleriei Léveillé, Fl. Kouy-Tchéou, 242 (1914).—Synon. nov. 3 ioe. Kweichou: Hia Obate-dentate with 4-5 large teeth, straight side tlet bro)’ 2 apex, with a minute inflexed lobe at base; > °vold, glabrous, slightly ribbed. “shan, alt. 500 m., common, H. F. Chow, no. 41730 uct species of the section Eucarpinus characterized “ Tather small loosely veined, coarsely and nearly ~ strate leaves with mucronate teeth and coarsely “<-dentate fruiting bracts. Carpinus Chowii is — Lae 48-50, 224-208; xm. 151-156. 334 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm apparently closely allied to C. Turczaninowii Hance var. ovalifolia Winkler, but the latter differs in more closely doubly serrate leaves and much more finely serrate but not lobulate bracts. \\ Carpinus Chuniana, sp. nov. Arbor; ramuli graciles, pubescentia villosa flavescente vestiti; gemmae globosae, 1 mm. diam., perulis paucis rotundatis imbri- catis. Folia membranacea, firma, ovata vel elliptico- ad oblongo- ovata, 8-11 cm. longa et 4-6 cm. lata, acuta vel breviter acuminata, | basi cordata, irregulariter et breviter duplicato-serrata vel sub- simpliciter serrata, supra costa puberula excepta glabra, subtus costa nervisque pilosis et axillis barbulatis exceptis glabra, utrinque glanduloso-punctata et reticulata, nervis utrinsecus 15-16 et 5-8 mm. distantibus trabeculis satis congestis conjunctis, petioli 8-10 mm. longi, villosuli. Infructescentia densa, 8-11 cm. longa — et ad 4 cm. lata, rhachi sericeo-villosula, pedunculo ad 3 em. longa villosulo; bracteae dense imbricatae, late semi-ovatae, ad 2.5 cm — longae et 1.3 cm. latae, acutae, in latere convexo breviter dentato- serratae, in latere recto remote serrulatae et basi leviter inflexae, reticulatae, utrinque in costa hirtellae; nuculae leviter compressa, — subgloboso-ovoideae, 4 mm. longae et latae, conspicue 8-costatae, glabrae et glandulis resinosis aurantiacis conspersae. Tree; young branchlets slender, yellowish villous-pubescent; buds globose, 1 mm. in diam., with few imbricate rounded scales. Leaves membranaceous, firm, ovate or ovate-elliptic to ovate — oblong, acute to shortly acuminate, cordate at base, shallowly and irregularly doubly or nearly simply serrate, glabrous except pubert- _ rn lous along the midrib above, glabrous except pilose along the : midrib and veins and with axillary tufts of hairs beneath, glandulos — punctate and reticulate on both surfaces, 8—11 cm. long, 4-6 &™ broad, veins 15-16 pairs, 5-8 mm. distant; petiole 8-10 mm. ay ‘ villous-tomentulose. Infructescence dense, 8-11 cm. long, © ; cm. broad, rachis sericeous-villosulous, peduncle slender, to 3¢™ long, villosulous; bracts densely imbricate, broadly semi-ovate — acute, convex side shallowly dentate-serrate, straight side remotely i : serrate and slightly inflexed at base, to 2.5 cm. long, 1.3 em. broad, reticulate, sparsely hirtellous along the midrib; nutlet eer! a compressed, broadly rounded-ovate, 4 mm. long and ribbed, punctate with orange resinous gl. glands, glabrous. “ : Kwanetune: Lokchang, C. L. Tso, no. 20872 (type), May 3) : 1929. A distinct species of the section Eucarpinus, characterized ¥ the large reticulately veined leaves, long, dense infructescence *” large bracts shallowly dentate-serrate on the convex side and HU, NOTULAE SYSTEMATICAE AD FLORAM SINENSEM 335 otely dentate-serrate down to the base on the straight side. (orpinus Chuniana is closely allied to C. Fargesiana Winkl., but differs in the much larger cordate leaves and in the bracts being either lobate-dentate on the convex side nor dentate-serrate to : the base on the straight side. bis kwangsiensis, sp. nov. ‘ ; ramuli juniores glabri. Folia coriacea, lanceolata, circiter 6 em. longa et 1.5 cm. lata, basi anguste cuneata et in petiolum marginatum circiter 1.5 cm. longum decurrentia, apice use acuminata, obtuse remoteque serrata et leviter revoluta, lucide Viridia, subtus fuscescentia, utrinque glabra et leviter lata rete leviter elevato, costa supra manifeste subtus leviter is globosus, pedicellatus, 7 mm. diam., niger. ub; young branchlets glabrous, leaves coriaceous, lanceolate, y cuneate and decurrent at base, obtusely acuminate at notely and obtusely serrate and slightly revolute along the Shining green and glabrous and reticulate with elevated ) above, glabrous and finely reticulate with slightly elevated beneath, about 6 cm. long, 1.3-1.6 em. broad, petioles about 1.5 cm. long. Flowers unknown. Fruiting panicles song, 4 cm. in diam.; fruits globose, pedicellate, black, 7 mm. ‘3 Me-kom, Seh-feng Dar-shan, S. Nanning, alt. 800 ng, 00. 8360 (type), Nov. 3, 1928. allied to R. indica Lindley, from which it differs in the lanceolata, sp. nov. Famuli glabri. Folia coriacea, lineari-lanceolata, 2.5- et 5-10 mm. lata, obtusa vel obtuse acuminata, * m petiolum 2-8 mm. longum marginatum de- a, valde remote et obtuse serrata et leviter revoluta, et rugosa, subtus laxe reticulata. Fructus globosus, Pedicellatus, niger. | ts glabrous. Leaves coriaceous, linear-lanceolate, and decurrent at base, obtuse to obtusely acumi- viming green and glabrous and rugose with elevated Slabrous and slightly reticulate with prominent very remotely and obtusely serrate and slightly the margin, 2.5-6.5 em. long, 0.5-1 em. broad; *$ mm. long. Fruits globose, black, pedicelled, 336 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Kwanest: Seh-feng Dar-shan, S. Nanning, alt. 775 m., R. ¢. Ching, no. 8060 (type). Oct. 21, 1928. Closely allied to R. gracilis N akai, from which it differs in its very narrow, remotely serrate, linear-lanceolate leaves and in its smaller fruits. ; “0 Torricellia angulata Oliv. var. intermedia (Harms), var. nov. Torricellia intermedia Harms in Bot. Jahrb. xx1x. 507 (1900).—Wange- rin in Engler, Pflanzenr. Iv.—229, p. 33 (1910). Szecuuan: Nanchuan, T’ai-ho-tung, Bock & Rosthorn, no. 154%, Sept. 1891 (type of 7. intermedia Harms); Mt. Omei, side of thicket, alt., 1800-2000 m., F. T. Wang, no. 23297, July 10, 1931 (tree); between Pai-kuo-wan and Moso-ying, C. Schneider, no, 620 (arborescent shrub to 4 m.). Yunnan: “laine de Tché-hai,” alt. 2500 m., E. E. Maire (Arnold Arb. distr. no. 294); without precise locality, G. Forrest, no. 10910. Hunan: “ad minas Hsi- kwangschan prope urbem Hsinhwa,”’ alt. 300-900 m., H. Handdl- Mazzetti, no. 11773, May 9, 1918. Kwanest: Yeo-mar-shan, N. Hin-yen, alt. 1225 m., R. C. Ching, no. 7237, Aug. 31, 1928. This variety differs from the type’in the leaves being coarsely mucronate-dentate betweeen the lobes and in the usually sparingly pubescent or puberulous veins of the under surface. Though Wilson’s no. 4611 from Taning-hsien, eastern Szechuan is some- what intermediate between the type and the variety, as it has _ few teeth between some of the lobes, the specimens cited above are decidedly different from the type in their leaves being ¢ . toothed all around between the lobes and merit to be distinguished as a variety. Fan Memoria Instirute or Bio.ocy Perrine, Cuina. " 198] REHDER, NEW SPECIES, VARIETIES AND COMBINATIONS _ 387 _ NEW SPECIES, VARIETIES AND COMBINATIONS FROM _ THE HERBARIUM AND THE COLLECTIONS OF THE ine ARNOLD ARBORETUM! = | | ALFRED REHDER - Whane hypomalaca, sp. nov. Frutex gracilis metralis; ramuli leviter flexuosi, hornotini fulvo- Ps hirsuto-villosi, leviter angulati, annotini rubro-fusci. Folii lamina e -ambitu triangulari-ovata, 3-4.5 cm. longa et 2.5-3 em. lata, basi __ Subcordata vel rotundata, inciso-lobulata utrinque lobulis 3-4 acutis erratis infimo 5-8 mm. longo, in caudam dimidiam laminam ae- en el us, lobi lanceolati, 5 mm. longi, petala rotundato-ovati, sepalis paullo 15-20; ovarium unicum vel raro duo, in stylum basi sparse et longe pilosum > * mm. longum, > Ovula 4-5, ril [1910-14 ?] “arbuste delicat, buissonnant, fl. in shady thickets on the Li-ti-ping, lat. 27° 12’ N Forrest, no. 18888, June 1917 (shrub 3-4 ft.. white). : a, weparently closely related to N. sinensis Oliver ee WwW. W. Sm.; from the former it is easily distin- 338 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm guished by the densely pubescent branches, rhachis and petioles, by the smaller subcordate and more deeply lobed leaves densely pu- bescent beneath and less so above and the shorter pedicels; from N. villosa it differs chiefly in the much smaller leaves, shorter peti- oles, narrower stipules, the presence of petals and the glabrous ovary. From N. thibetica Franch. which also has pubescent leaves, it differs chiefly in its spreading and pilose, not velutinous pubes- cence, more deeply lobed smaller leaves, shorter petioles, the calyx- tube not silky outside and in the glabrous ovary. Securinega suffruticosa (Pall.) comb. nov. C lum ? suffruticosum Pallas, Reise Russ. Reich. ut. pt. 1, p. 424 (1776), nomen. on ? suffruticosum Pallas, op. cit. m1. pt. 2, p. 716, t. B, fig. 2 1776). Xylophylla ramiflora Aiton, Hort. Kew. 1. 376 (1789). : Osyris alba Georgi, Beschr. Russ. Reich. ur. 1341 (1800); non Linnaeus Ross 583 XXXII . Phyllanthus japonicus Mueller Arg. in Linnaea, xxxur. 52 (1863), in pe j iflor ndolle, Prodr. xv. pt. 1. 44 (1866).—Rehder in Jour. Arnold Arb. viz. 191 (1926); vin. 152 Lay? fluggeoides Mueller Arg. in De Candolle, Prodr. xv. pt.1- nega japonica Fi. Jap. 292) (1867), in Some time ago Mr. H. C. Skeels drew my attention to the fact | that Securinega ramiflora (Ait.) Muell. Arg. had an older specific epithet, namely “suffruticosa” going back to Pharnaceum ? suffritt- cosum Pall. of 1776. Fischer and Meyer based the type of their new monotypic genus Geblera on Pharnaceum 2 suffruticosum, of which Pallas gives not only a fairly accurate, though incomplete descrip- tion but also a figure of a flowering branch; this figure with nt description, taking into consideration the region where the plan was collected, shows clearly that the plant described by Pallas is — identical with Securinega ramiflora. I am unable to find specific differences between S. ramiflora and S. fluggeoides, as I have already stated in this Journal (va. an aq and vit. 152); all the differences given are variable and do not hold in the numerous specimens before me. Also Securinega japo™ ica Miquel, at least partly, belongs here, excluding the name brig ing synonym, the specific epithet having been taken from peapertt : : japonica Baill. which is a synonym of Phyllanthus fleruosus (S og & Zucc.) Muell. Arg. Fol ica Miquel in Ann. Mus. Bot. Lugd.-Bat. m1. 128 (Prol. apibed 882] REHDER, NEW SPECIES, VARIETIES AND COMBINATIONS — 339 Vitis Wilsonae Veitch apud Gard. Chron. ser. 3, xivr. 236, fig. _ 101 (1909), nomen seminud.—Rehder in Sargent, Pl. Wilson, m1. 498 (1917), pro synon. V. reticulatae Gagnep. s ‘itis reti Pampanini in Nuov. Giorn. Bot. Ital. xvi. 429, fig. 13 (1910) —Rehder, Man. Cult. Trees Shrubs, 603 (1927)—Non V. reticulata (Thwaites) M. A. Lawson. Vitis reticulata Gagnepain in Lecomte, Not. Syst. 11. 12 (1911); in Sargent, _____ Pi Wilson. 1. 103 (1911).—Non V. reticulata (Thwaites) M. A. Lawson. This species was described at approximately the same time inde- _ pendently under the same name by Pampanini and by Gagnepain, _ but based in each case on different specimens; by Pampanini on ce Silvestri, no. 1438 and by Gagnepain on Farges, nos. 539 and 124 El and on Wilson, Veitch. Exp. no. 1151. Both authors overlooked or = disregarded the older homonym Vitis reticulata M. A. Lawson Which they probably considered as based on the invalid name 4 r Cissus reticulata Thwaites, changed by Planchon to Cissus retivenia _ ‘account of the earlier C. reticulata Bl., which was cited by . Miquel (in Ann. Mus. Bot. Lugd.-Bat. 1. 81. 1863) only as a a ‘ynonym of Vitis geniculata (Bl.) Mig. and did not become a valid tame until Planchon published a description in 1887. There exists, : however, a still earlier name, (. reticulata Willd. apud Roemer & é Schultes, Syst. mr. Mant. 248 (1827), which was referred to Cissus Peeried Lam. by Kunth. Vitis Wilsonae, the earliest name given ecies, was not taken up, as it was considered a nomen nudum dum having been published with an insufficient descrip- a figure of a leaf. It was identified with V. reticulata m 1917 and is now the only available name for this Ms ana Stenophyllus Harms f. dilatatus, f. nov. © Tecedit foliolis semper 3 (in specimine viso tandem) anceolatis vel oblongo-obovatis petiolulo brevi incluso .. et 2-3.5 cm. latis —Frutex inermis, glaber, ramulis thas. han 8 i: Yuan-ch’ii distr., Shui-wang-ping, in silva 900 m., Harry Smith, no. 6563, July 21, 1924. Apparently an extremely broad-leaved form of A. ‘ough at first glance it looks very distinct on account eader leaflets, and particularly if compared with the of the species, A. stenophyllus £. angustissimus Rehd. “TD. EX. 99. 1928) with leaflets only 3-5 mm. wide. Possible that these two extreme forms could belong Sons I can find no other character to distinguish it mus. From A. Wilsonii Harms, with which this distinctly stalked and in longer pedicels. 340 JOURNAL OF THE ARNOLD ARBORETUM [vou. xu Campsis Tagliabuana (Vis.), comb. nov. (= C. chinensis X radi- cans). Tecoma hybrida hort. ex Dippel Handb. Laubholzk. 1. 48 (1889), pro syn. T. grandiflorae. in in Jardi ‘ ; LV. 315 (1899).—Rehder in Sargent, Trees & Shrubs, 1. 93, t. 47 (1903). Campsis hybrida Zabel in Ruempler, Gartenb.-Lex. 166 (1901).—Rehder * en” Stand. Cyel. Hort. 1. 652 (1914); Man. Trees Shrubs, 790 a intermedia Schelle in Beissner, Schelle & Zabel, Handb. Laub- holz-Ben. 435 (1903). It seems strange that Visiani’s name Tecoma Tagliabuana which was published with a full Latin description and the definite state- ment that it is a hybrid between T. radicans and T. grandiflora, has been completely overlooked by all the later authors and no mention of this name is found in botanical or horticultural litera- ture, until Meunissier resuscitated it in an article in Revue Horticole in which he gave complete data and a colored plate of this hybrid. According to Visiani the hybrid was raised by the brothers Tag- liabue apparently some time before 1859, but it does not seem to have been distributed under Visiani’s name. The hybrid, however, has appeared in gardens under various other names. The first references I find are those given in 1872 by K. Koch (Dendrol. u. 308) as T. Princei coccinea grandiflora and in 1877 by Lavallé (Arb. Segrez. 176) as T. radicans coccinea grandiflora. : It is also probable that T. radicans atropurpurea mentioned in grandiflora atropurpurea belong to it. Also the following names are referable to forms of this hybrid: Tecoma grandiflora aurantia and T. grandiflora Princei (Dippel, Handb. Laubholzk. 1. 48. 1889); T. chinensis aurantiaca (Koehne, Dendr. 522. 1893); Campsis radicans f. atropurpurea Voss, C. chinensis f. aurantiaca Voss and f. Princet Voss (Siebert & Voss, Vilmorin’s Blumengart. 1. 801. 1894); r radicans Princei, T. grandiflora rubra Hort. Sahut and 1. grandiflora v. 208, 209. 1898); T. radicans grandiflora atropurpurea and ames may appear in horticultural lit- erature or in catalogues under Bignonia or under Campsis instead € | J Tecoma. os a . same specific and varietal names yd: es Lasianthus Labordei (Lév1.), comb. nov. dau 4 "Canthium Laborde Léveillé in Fedde, Rep. Spec. Nov. xu. 178 (1914); 4 Fl. Kouy-Tehéou, 364 (1915). ~ Per hiis es ae Cuina. Kweicho u: district de Tsin-gay, mont de Kao tebay, REHDE , R, NEW SPECIES, VARIETIES AND COMBINATIONS — 341 ag in Herb. Arnold Arb.); Si-mi-yao, Pa- -na, pene Sone Se Mise-shan, Kiang-kow, alt. 450 m., Y. Tsiang, no. 7487, 1930 (small tree in light wood); foot of Van-ching-shan, Yin- alt. 500 m., Y. Tsiang, no. 7591, Dec. 12, 1930 (low shrub, ae of Ksi, Tan-ling, ihe S. pRoreer alt. 500 fats: 4 fac: irons aids show. It apparently is hee: - to Lasianthus and seems most closely related to L. longicauda ok. f., from which it differs chiefly in its much narrower leaves em. meg and usually 8-16 mm. broad, with less prominent > veinlets and in the glabrous or nearly glabrous he Giole plant is generally quite glabrous, only on no. ight pubescence is noticeable on the tips of the branchlets, ; Bey riick and on the very young calyces. No. 7487 differs on hey Dice. Tess about 1 mm. pig ? ay us longicauda Hook. f. I have seen no Indian material, Pea ‘unnan specimens (Henry nos. 9035 and 10633) determined and one Yunnan specimen collected by Delavay at termined by Pitard as L. japonicus Mig. which to L. longicauda and not to L. japonica Miq. ers in the quite glabrous more caudate leaves, the branchlets, less pubescent calyx and in the corolla- ui s inside except at the base, while in L. japonica sure (Iconog. FI. Jap. 1. t, 28) the corolla- inside to their very tips. (Lo be arene 342 JOURNAL OF THE ARNOLD ARBORETUM [vou xm THE CRATAEGUS PROBLEM Ernest J. PALMER IN LOOKING OveER the field of American taxonomic botany, as judged by the manuals, reports, and local plant lists that have been published since the beginning of the present century, it must be apparent to any one that the biggest unsolved problem and the one about which the greatest difference of opinion exists is that of the genus Crataegus, especially as regards the validity of the large number of species that were proposed, mostly during the first decade of the century. There can be little wonder that when the twenty or twenty-five supposedly well-known species of the older manuals began to expand by scores and hundreds, until more than a thousand new species and varieties had been published, the first feeling of surprise on the part of the general students of botany and interested laymen, later assumed something like an attitude of dismay and incredulity, as they realized the hopelessness of attempting to become ac- quainted with or to identify living plants and collections of this genus from technical descriptions, in many of which only very slight characters were indicated to differentiate the proposed species. In Gray’s Field, Forest and Garden Botany, published in 1857, twelve species and two varieties of Crataegus were listed for the 5 : 5 z 2. 5 f : a 3 g 2 E = range. ce ae ee, a Up to the year 1899 about 175 names, including varieties and forms, a number of them merely on garden lists and without botanical descriptions, had been pu lish d for North Americad PALMER, THE CRATAEGUS PROBLEM 343 Fe ieee. A large proportion of these were segregates of supposedly ry f ee species, such as C.. crus-galli, C. coccinea, C. punctata, @. glandulosa, C. tomentosa, and others. _ About ten years later what may be called the period of expansion for the genus began. In 1899 and 1900 Beadle and Ashe published _ independently descriptions of several new species from the south- tastem United States, followed quickly by many others, and Sargent a year or two later began describing many new forms, -_ mastly found in the Northeastern and Central States. Altogether . Ashe has see at least 177 species and 3 varieties of Gretongus, Mr. C. D. Beadle 143 species and 1 variety, and _ Professor at, more than 700 species, 22 varieties and 5 forms. ok small number of Sargent’s names were new combinations or _ Were proposed as substitutes for various reasons for his own pre- Viously published species or for those of others. __ The treatments of the genus that have appeared 1 in the manuals Me ¥ ‘> 2, By ey : y provisional and partial, with frequent shifting of ground in regard to the species recognized and the passing over entirely of many others. This is not surprising nor a matter for adverse ; Me but it serves to illustrate the difficulty of the problem. + compilers of local lists have in the meantime given up “any attempt to enumerate the Thorns of their regions, eat vonage mentioned the fact that various unrecognizable eccur; some have referred their readers to the manuals or ta’ treatments, or have submitted collections to a few Bis Ants of the genus who were willing, or who had ‘Ys to undertake identifying them. The Arnold Ar-_ been ¢ - upon to pass upon a number of such m spite of the large number of types-and the great material in the herbarium, as well as an extensive Mictincns, even when flowers or fruit were present, = cases it is quite out of the question to determine them ens or from a single leaf or flower, as can often © wy re as the Willows, Oaks, Maples, Roses, and oa this situation a feeling seems to have arisen among _, ~ and uncritical students of plant life, and even S, especially those not mainly concerned 2 hopeless confusion exists in this genus ph ols the SPAS wvesert species are 344 JOURNAL OF THE ARNOLD ARBORETUM [vou xm without standing and should be disregarded. It has also been suggested that the genus is in a state of mutation or instability or that the different forms have been so frequently hybridized with each other that no clear specific lines can be drawn as in other genera, and that almost any individual tree might be made the type of another so-called species. Such a situation as the present is obviously unsatisfactory and unscientific, but much of the uncertainty and skepticism, under- standable as it is, does not seem to ‘be entirely warranted by the facts. For it is probably true that morphological characters peculiar to and inheritable in segregable groups of individuals that we call species are generally as constant and well marked in this genus as in many other genera of the Rosaceae or other families, such as Rubus, Rosa, Prunus, Cotoneaster, or Rhododendron. The main difference and difficulty in regard to Crataegus as contrasted with such other genera is that of size, the very large number of species that must be dealt with in any adequate general treatment, their wide geographical range, and the practical difficulties of studying or collecting them in the field, with the uncertainties of seasons and irregularity of fruit crop, or of bringing together under cultivation a collection at all adequate for a general revision. ‘But in addition to this it will } generally agreed, I believe, that matters have been greatly complicated by the way in which the genus for several years. In view of the great amount of work done upon the genus by : one mj -—S”s«é PALMER, THE CRATAEGUS PROBLEM 345 the late Professor Sargent and the accumulation of material and _ literature in the herbarium, plantations, and library here, it seemed evident that the Arnold Arboretum was the logical place for initiating and carrying out such a work, and indeed, that it could _ wareely be done with adequate facilities anywhere else. And for this reason, after being urged to do so from a number of sources, the writer rather reluctantly undertook this difficult and somewhat re elapsed since its beginning the plantation is mito’ a collection at the present time of about 1,400 of record, representing nearly 700 species and varieties, Percentage of which are of Old World origin. Other he seed was taken has been in progress are merely individual variations, as well as _ nclusions as to the relative value of number of stamens and styles, 346 JOURNAL OF THE ARNOLD ARBORETUM fron. xm color of the anthers, shape and color of the mature fruit, &e., as specific criteria. j To examine and make notes upon this large number of trees both in the flowering and fruiting stage and to make the necessary comparisons with the herbarium material and with the published descriptions, is at best a slow process and one attended with some — practical difficulties. Most of the trees.have now grown to fruiting — size but not all of them produce flowers or fruit each year, some seasons being quite unfavorable to the less hardy forms. Others are just coming to maturity and some flower for the first time each year. Since the flowers are very transient and it is possible to — determine the color of the anthers for only a short period after they open, the season often advances so rapidly that it is difficult | to keep up with them in this stage. Additional field work which is urgently needed in this genus in many parts of the range also presses at this season and for several years it has made it impossible for me to be at the Arboretum in spring. : iy It should also be understood that large as the collection here ato and furnishing as it does by far the best experimental data so far available, it is by no means complete or fully adequate for a study of the whole genus. Some species of the'southern states have not proved hardy in this climate and for various reasons many which ‘ have been started have not survived and others have not been a secured. It can easily be understood that a certain amount of error was almost unavoidable in the various steps from collecting — herbarium specimens and seeds in the field, planting and germinating 4 these, transplanting the seedlings to nursery rows, and finally ef to a permanent place in the collection, as well as later in keepiif t records and in having the plants properly labeled. A small pees centage of such errors have been detected in the collection her 2 Some of these are so obvious as to be unmistakable and sometimes . correctable, but in a few cases they may leave open to baste! tions of possible variability, which we would like to settle. ee) addition to this the first idea of the plantation, which seems © — of each of the several hundred aupposed species to fruiting 9% but it would have been most helpful if this had been done in 2 re The main region of distribution for Crataegus in North Amerie! : extends from the Atlantic seaboard to beyond the Mississl | PALMER, THE CRATAEGUS PROBLEM 347 river, and the border of the plains. The western boundary may be roughly taken as running from northwestern Minnesota, through eastern Nebraska, Kansas, and Oklahoma, to the mouth of the Colorado river in Texas. Many of the species and several of the groups disappear much farther east, and several sub-regions may be recognized. West of this region Crataegus is a genus of minor importance and such scattered species as occur are mostly quite different and distinct from those of the east, and the problem is comparatively simple. Some of the western species are associated with the Sonoran floras and range southward into Mexico, and others belong to the Rocky Mountains and Pacific floras. A considerable part of this main area has been pretty thoroughly explored, but it should be remembered that there are still large wetions of the country in which very little collecting has been done and where the Crataegus flora is still inadequately known. Most of the material studied by Sargent and other recent authors was brought together as a result of intensive collecting by local students in a comparatively few limited areas. A large amount of material was left undetermined in the herbarium here, and while much of this has recently been placed with described species, a tumber of collections remain, with either complete or partial 3 material, that cannot be so disposed of, and reluctant as one might be to add to the number already published, it will be necessary to | some of these that seem clearly distinct. fay : The sections of North America where the Crataegus flora 1s 5 Pest known and from which material is fairly adequate are the St. ae hee valley, most of New England, New York, Pennsylvania, 3 Gu of the Southeastern and Middle-western States and the 8 el But while g there is every reason to believe that some new species ‘ “ilturn up in these regions, it is not likely that the number that ips and the ideal classification of Oralaegus, 98 in © bia etP® that cannot be definitely settled on purely mort breeding, and cytological study of - - : responses of if 4 pPosed species will ultimately throw much light _ some and varieties — bi i 348 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm fication and conception of species. Investigations along some of — these lines are now in progress at the Arnold Arboretum and elsewhere, but the practical difficulty of applying such tests to hundreds of forms of such slow-growing plants as Crataegus is obvious. According to Mr. W. H. J udd, propogator for the Ar- boretum, the seeds require from two to six years to germinate. And on an average ten years more must elapse before they grow to fruiting size. But valuable as such investigations are, they have not yet pr- gressed far enough, at least in the case of Crataegus, to be used as the basis for any scheme of classification.! And indeed it scarcely seems likely that they can ever be so used except in a limited way, or that any classification based mainly upon other than obvious morphological characters can be devised that will be practicable and usable by the large and increasing number of people of all degrees of scientific attainment who are for various reasons I- ‘terested in the study of plants. Of course a taxonomic arrangement based upon morphological characters should, approximately follow lines of genetic relationshi , but in certain cases it probably does not do so consistently, and as in keys used in the manuals to aid m locating species, an artificial arrangement is sometimes more useful and usable than one that is concerned only with natural relationship. In considering any scheme of classification it is well to under- stand clearly what the writer means by the term species or other subdivisions of the group, for there is still nothing like uniformity in the use of such terms. If we begin by recognizing that there is nothing inviolable or Sacrosanct about a species, but that na merely a convenient unit to be employed in describing a group of plants or animals having a number of recognizable characters common, and that the limits of such a unit must be based upe? tbe best judgment of some competent observer who has studied i) : group, it is evident that a considerable degree of latitude may be oh exercised in the use of the term. The number of species in a°Y group will therefore vary according to whether the author takes * : narrow or a broad view of the species. The decision must be more ; or less arbitrary, and naturally there will be room for pneted®” : j of opinion about any proposed classification, and in the nate © things it must be somewhat provisional and subject to revisio® It might be more accurate to refer to such units as morphological ie "species, since they may differ greatly in limits and numbers iP a the species concept of the geneticist or cytologist. _ eit ‘cake 4 A system of classification in which only species are recognized, es Pra ee en 5 + 1982] PALMER, THE CRATAEGUS PROBLEM 349 one in which consequently only Linnaean binomials need be employed, has its obvious advantages in simplicity, especially in small groups. But to attempt to treat such a large genus as Cra- faegus in this manner would result either in an enormous and unwieldy number of species, or specific lines would have to be drawn so broadly and with so many exceptions as to make them almost useless. A conservative treatment, such as seems desirable here, should regard as species all readily recognizable forms that can be distinguished by a group of characters or at least by more than one clear difference, with descriptions broad enough to allow for a reasonable amount of individual variation, to be determined by observation. Under varieties would be placed such subdivisions of the species as differ from the typical form of the description in a single clearly recognizable character or in one or more minor details. And in some cases it might be desirable to give names to forms in which a single distinguishing character less clear or con- stant is found. Under such a treatment the number of species might be consider- ably reduced from the present total, but it would still be large, probably much larger than in any other genus of woody plants in the American flora, unless we refuse arbitrarily to recognize as species forms quite as well marked as many of those generally so treated in other genera. Such a course would go far tow defeating its own purpose. Although no satisfactory general treatment of Crataegus has yet appeared since the publication of the large number of recently Proposed ade towards a better “hderstand have been real contribution, but in regard to others there seems to be very Mt cies, based upon Our as large as could have been expected. Still it i : _ ait the number of species in a genus by law or rule, will be better to proceed slowly and to collect 25 much data ible before attempting a revision. te é 4 350 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Sufficient progress has been made in observing the living col- — lections at the Arboretum, in additional field work and in rear- ranging and critically studying the material in the herbarium, to make it possible to draw a few general conclusions. q The extreme complexity of the genus, the existence of many — distinct forms in all parts of North America, many of them being — quite local in range, and which were unrecognized and unaccounted for in the older manuals and floras, has been amply demonstrated. The recognition of this fact and the great amount of work done in making them available for study both through published descriptions and cultivation constitutes a great contribution on the part of Sargent and his co-workers and contemporaries. It cannot be too strongly emphasized that nearly all of the large number of forms described were based upon real differences. Whether these dif- ferences were sufficient to be made the basis for specific distinctions is an open question about which there can properly be much difference of opinion. In the opinion of the writer it might have been better in many cases to have treated them as varieties oF forms or merely as variations of polymorphic species. The specimens cultivated at the Arnold Arboretum show gen- erally that the distinctive characters of the wild specimens, which were taken as the basis for the new species hold true often to minutest particular. There are a few exceptions to this. Besides the plantations at the Arnold Arboretum a very large number of species were grown and records kept of them by the Park Department of Rochester, New York. The collection there is available for study, and those who have been in charge of it report similar results, and indeed some of them are much mee inclined to recognize and to insist upon the distinctiveness of the species than we are. In addition to thosé species that can probably be more properly treated as forms or varieties, some others were published of which the names were invalid, or which cannot be maintained for mer reasons. In some cases names were preoccupied or the same name 1 was used more than once for different species by the same oF = x different authors. Most of these cases have already been cor . , but apparently a few still remain. Duplicate descriptions undoubt- f edly sometimes appeared of the same plants, either by the same - i by different authors working independently. This was alt : unavoidable from the mass of material that was being | oe the rapidity with which the descriptions appeared, and the small | amount of tabulation that had been done, as well as because M™ characters of some of the sections into which it has been proposed to divide the genus were not properly under stood. 1988] PALMER, THE CRATAEGUS PROBLEM 351 There is a very wide range in the variability of the leaves, flowers, fruit and several others characters in Crataegus, and the species can be arranged in several natural sections or groups. There have been anumber of attempts at dividing the genus into these sections, with considerable variation as to the number, arrangement and limits of the groups. The oldest arrangement is that of Loudon in 1838.! In this treatment, he arranged the species of which he gave descrip- tion in 15 sections, of which one (XV. Pyracantha) is not now included in the genus. Four others, his sections V. Nigrae, X. Azaroli, XI. Heterophyllae, and XII. Oxyacanthae, are exclusively Old World groups. His other sections are clearly recognizable aud most of the names are retained, with a few changes in the Placing of species. The sections of American Thorns were I. Coceineae, II. Punctatae, Il. Macracanthae, IV. Crus-galli, V1. Douglasii, VIL. Flavae, VII. Apiifoliae, IX. Microcarpae, XII. Parvifotiae, and XIV. Mevzicanae. To these a number of new sections have been added by later authors. Mr. E n in his treatment of the genus in the 7th edition of Gray’s Manual (1907) arranged the American species that came within the geographical range of that work into 17 sections, and mn the 2nd edition of Britton & Brown, Illustrated Flora (1913) he has I sections, some of them slightly changed in scope from the er treatment. | ‘Professor Sargent proposed several new sections, and in his last Teatment, in the 2nd edition of the Manual of the Trees of North America (1922) he arranged the arborescent species into 20 natural sroUups, which did not take into account the Mexican species nor ee two exclusively shrubby groups, Parvifoliae (Uniflorae) “ %. iflorae. In the course of his studies of the genus he abandoned his soups Lobulatae and Flabellatae and seems to have changed his com the groups Coccineae and Rotundifoliae, as well as his ave accepted Beadle’s name Silvicolae as having priority over ta Medioximae, and Macracanthae of Loudon for Tomentosae. the en, Flora, has carried the division of the genus into sections 9 papit ” : degree, basing some of them, as it seems to me, apen eee ‘pay 424 82 more than specific value, and in the case oF MH | Ne ee and Berberifoliae scarcely that in some cases. “© af & ‘third “are recognized in this work, which covers less than ‘The ~ , in his generally excellent treatment of the — “outheastern United States, as it appears in the 2nd edition * the area of the United States. their dis- 352 JOURNAL OF THE ARNOLD ARBORETUM (vou xm tinctiveness and in the constancy of the characters by which they — can be recognized. It is generally possible to definitely place most _ species in the proper group with the help of keys or by one familiar ; with them, but in most cases there is an indistinct border line where one group merges into another, and perhaps into more than _ one in different directions or in different areas, and it is difficult ‘ to say whether certain species belong with one or with the other. | This may indicate either hybrids between species of the different groups, which is sometimes probably the true explanation, or it may indicate merely the relationship of the groups to each other and that certain intermediate species connect them. For example, in the section Crus-galli, which is generally one of the best marked of the larger groups, some species approach in the character of their fruit and foliage those of the section Punctatae, which is usually placed next to it, and in other species, the thinner or slightly lobed leaves and smaller fruit, which may finally become mellow or succulent, or in the thinner, scaly, bark of the trees, an approach to the Virides is found. The Virides group has good distinguishing characters in most cases, but besides certain species that seem I some ways intermediate with the two previously mentioned, some of the species assigned to the Pulcherrimae in the southeastern states have certain resemblances to this section, and in south- eastern Texas species have been found that seem intermediate between Molles and Virides. There has been much uncertainty for the typical form of that species suggests some large-lea or more distinct things. It is not an unnatural consequence = : later authors have been much confused and very hazy iO a interpretation of this species, and indeed it has since been made t include either typically or as varieties, forms of many diverse species 4 of American thorns. The group Coccineae, as finally used by a gent, includes several species of the northeastern United State® and Canada, having rather large, thin leaves, mostly glabrow : when mature, flowers in which the styles are usually less than five, : and fruit with rather small, sessile calyx, and flesh that becom — pulpy or succulent at maturity. On the one side this group PF proaches the Tenuifoliae and on the other the Molles, and a ‘ é 192] PALMER, THE CRATAEGUS PROBLEM 353 as in other cases, it is difficult to decide definitely to which of these groups certain species belong, a description broad enough to cover such a wide diversity of forms as have been included under this group and the Molles, as well as those of the small Dilatatae group of Sargent, would lose much of its taxonomic value. The same situation seems to obtain with most of the other groups, which it is not necessary to take up in detail here. Attention may be called, however, to the fact that there are several small groups, such as Aestivales, Brachyacanthae (Brevispinae), Microcarpae and Cordatae, which appear to be so clearly distinct that they probably tepresent species that originated quite early in the history of the genus, It seems probable that Crataegus was a comparatively early offshoot from the Pomaceae stock, but that its principal develop- ment has taken place since the Glacial periods of the Quaternary and even in very recent times. It may have originated at rather high latitudes in Eurasia in a period of favorable climatic conditions and been dispersed from a circumpolar center southward into both hemispheres before the close of the Tertiary era. This view seems ‘to be supported by its present wide distribution and that of as- Sociated plants, and by the fact that the genera most closely related to it amongst living plants, such as Osteomeles, Cotoneaster, P yracantha and Mespilus, are confined to the Old World. That ne groups found in the-two hemispheres are all, or nearly all, {ute distinct would also indicate a considerable period of isolation. | Paleontological evidence seems to be rather meager for a genus c abundant and widely distributed at present, but a number of _ Ss species have been described, based upon either leaves or : fruit, from both the Old and New World. A comparison of these “ith living species will be of value and may throw light upon the Present distribution and relationship of the different groups. we: _Afew generalizations can be made as to the geographic distribu- ae ‘large proportion of the species that have been described appear _ quite local and in not a few cases the only material nee” ‘tom the type locality or from a single tree. Such Pica Bis they come from regions that have been pretty well 44 d tog - y, naturally fall under suspicion as to their validity, an them will probably prove to be only divergent ‘ “te ~ 8. Many species, however, and some of them well mar i ~ are of quite restricted range. This may be accoun a Ways: either they are comparatively recent forms that have 354 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm not been able to spread far from the point of origin, or they may be relics of ancient and disappearing types. In the case of Crataegus the former seems much the more probable explanation. Several considerations lead to this conclusion, but it need only be mentioned here that in the case of relic species the range is usually interrupted and they crop out in peculiarly protected or favorable spots, often widely separated from each other. This is not the case in Crataegus, with a very few exceptions, so far as known. Most species of Crataegus are more or less pronounced calciphiles, and they are found in the greatest diversity and abundance in limestone regions. Many of them are limited to such outcrops, and the soil factor is probably an important one in determining their distribution. There are, however, some striking exceptions to this. There is a marked zoning of the groups from north to south, and to the westward they are limited by aridity, like other mesophytic plants, as they approach the plains. But there are various other minor causes, besides soil and climate that serve to limit them and that have brought ‘about the present distribution. This is emphasized by the fact that in spite of the limits of the groups north and south in nature, the species in the main seem quite adaptable in the matter of climate, many of those native to Florida and southern Texas having proven hardy in New England The genus as a whole is not adapted to spreading in forest areas, and most species are essentially plants of prairie openings, borders of woods, copses, pastures and glades, or along the open banks and bluffs of streams. In primitive times such habitats in Eastern North America, where other conditions were favorable, were 20t of wide extent. The great unbroken forest that extended from the Atlantic coast, across the Mississippi valley, to the edge of the plains, afforded only occasional opportunities for the growth of such small trees and shrubs, and in places where they might seek to gain a foot-hold they were probably held in check by frequent fires and by grazing animals, since the young shoots, before the spines are sufficiently developed to afford protection, are eaget eaten by the ruminants. The seeds are heavy and are not likely to be transported far except when carried by water or in the stony — achs of birds and other animals. The latter seems to be at i : portant means for their dissemination at present, but so bog* favorable localities were lacking this would avail little, and ae 4 advance was also probably impeded by the slow germination © the seeds. eee | Since the clearing away of the forest on such a large scale, vastly greater areas have become available and there has undoubtedly 1932] PALMER, THE CRATAEGUS PROBLEM 355 been a great increase in their development and in the spread of species in recent years. The distinct calciphiles and rock growing species, or those of more specialized requirements, are usually the most limited in their range, and the more widely distributed ones, such as Crataegus crus-galli, C. punctata, C. viridis, C. apiifolia, C. spathulata, C. Margaretta, and ('. tomentosa, are such as. grow in alluvial ground or that follow the courses of large drainage basins. In general it may be thought that the more widely distributed species are the ancient ones, or that at least they are not of very recent origin. The criterion of age and origin is probably, however, only partly reliable in this genus, since other factors, as pointed out, have served to aid or retard their extension. In the writer's opinion the groups Aestivales, Brachyacanthae and Merxicanae are probably most closely related amongst American Thorns to the primitive types, and none of these have a particularly wide geo- graphical range. In the case of Aestivales, at least, this may be due to the peculiar and limited habitat. And it is probable that all of them are retreating or disappearing groups. A striking exception to the usual soil preference and habitat of most, of the groups is afforded by the large association of forms which has been classed under Flavae, although Mr. Beadle divides them into twelve or fourteen sections, of which it may be desirable to maintain some. These are restricted to the southeastern ay inated by slender branchlets, which are either zig-zag OF pendulous, and thickly set with single or small cluster ik _Sxdusively southern groups are the 2 Hea Microcarpae, Brachyacanthae, and Aestivales. ne of the Tal plants, which e2 356 JOURNAL OF THE ARNOLD ARBORETUM [vo. xm be properly placed when brought into comparison with each other. How far the matter of hybridization enters into the problem of the complexity and proper classification of Crataegus is a puzzling one, and one that cannot be settled definitely without a great deal more experimental work being carried on systematically through many years. From what is known of the rather frequent cases of hybridization amongst other Rosaceous genera, and from the fact that several supposed hybrid species of Crataegus, as well as a cross with the closely related genus Mespilus, are in cultivation, and others are found growing spontaneously that suggest such an origin, it is almost impossible not to believe that natural hybrids do sometimes occur. Anyone who has been in a large growth of Crataegus in early summer, where many diverse forms are growing together, and several of them blooming simultaneously, with the wind carrying pollen, and with swarms of busy insects flitting from one ways in which the integrity of species is protected in nature, and if this were not so, the humerous distinct forms that are found = many large genera could never have been segregated or maintained. Slight physiological differences are often enough to insure immunity, may make cross-pollenization in the wild state unlikely or impos- sible. : pe But in addition to the theoretical considerations that lend support to the belief in Crataegus hybrids some positive morpho- logical evidence is found both in the intermediate characters of as well as in the high percentage of pollen sterility. Pees Standish,! who examined th pollen of 171 species, states that thirty-five of this number had normal pollen, sixty, from 10 to” 50% sterility, forty-one, from 50 to 75%, and thirty-five from 75 to 100%. This would seem to indicate a large percent of hybrids or chromosome irregularity. $= § = § bee Longley,? who made studies of the chromosome structure, classes as triploids, and eleven as tetraploids and triploids. A considerable ‘aNDISH, L. M. What is happening to the Hawthorns? (Jour. Heredity eae 1932] PALMER, THE CRATAEGUS PROBLEM 357 degree of variability in the seedling offspring of triploid or tetra- ploid species might normally be expected. But this, as stated above, has not been the experience with the species grown here and elsewhere, where records have been kept. The fact that so large a percentage of the forms come true to the parent type searcely seems to lend support to the theory that they are of hybrid origin, and if that is admitted to be the case upon other evidence, it would seem to indicate that these triploid forms have developed a type of apogamous reproduction, as pointed out by Sax,' and as a consequence do not develop the variations of usual hybrids. My colleague; Dr. Karl Sax, has kindly furnished me with the results of some preliminary experiments which he has conducted to test this. The anthers and stigmas were decapitated before opening in 25 to 50 flowers of a number of selected species. Of the 39 species and varieties reported on, 16, or 41%, set fruit, ranging from one or two in most cases, to five in Crataegus erecta, seven in C. pruinosa, and eleven in C. Oxyacantha plena. In an experiment of this character it is also possible that injury to the flowers may have reduced the chance of setting fruit in some cases. This would seem to indicate clearly that apogamous reproduction does some- times occur, but as in the cases of the other lines of investigation upon chromosome characters and sterility, these experiments have not so far been carried out on a sufficiently extensive scale to be conclusive in regard to many species, although they point to general Conclusions. Ags an illustration Crataegus pruinosa appears to _ * well marked species with a wide range. Standish reports 60% pollen sterility in the specimens of this species tested. Longley ae > be a triploid, while Moffett? reports it as a diploid. hs ‘ possible that the specimens selected were not in some cases PUure- bred C. pruinosa. And it can easily be seen that any modification Bix! the scheme of classification, such as changing the limits of ibe Section, transferring species from one to another, or combining _ “eral so-called species into one, might materially alter results ~ eonclusions based upon them. ‘To further ee _ ““#Ndish found a high percentage of sterility amongs : 3 ft Intricatae group, and she concludes that it yeast allied ie Leas and that all of the species of Intricatae may geek i hybrids between two species of the latter group oF aeeege yet "species and some other parent. However, the ranges sfoups are quite different from those shown on the 12g KARL. The origin and relationship of the Pomoideae. (Jour. cease 22. 1931.) sehavior in Pomoideae. alae A. A. A preliminary account of Chromosome behavier = ogy, Ix. 100-110. 1931.) koe ey 358 JOURNAL OF THE ARNOLD ARBORETUM [vou. xnr to support this conclusion. The two groups do overlap in the northeastern states, but the wide range southward and westward shown for Coccineae results from following the treatment in Gray’s Manual, which includes Molles with that group, and if this were done consistently the range would be more than twice as extensive. Attention may be called to the fact in this connection that Longley found the three species of Molles examined to be diploids and all of the twelve Coccineae to be triploids. The range of the Intricatae group also extends westward through Ohio, Indiana, Kentucky, Michigan, southern Illinois, the Ozark region of Missouri and Arkansas to eastern Oklahoma, and many of the species are found outside the range of any species of Coccineae or in a few cases of any Molles species. Moreover, on morphological grounds, I can see little evidence of a close relationship between the I ntricatae group and the Coccineae, as understood here. Whatever the origin of certain of the Intricatae species may have been, the group as a whole is one of the better, although not one of the best, marked natural sections, and it appears to be more closely related to the Pruinosae and Rotundifoliae than to Coccineae. There is at least ground for suspecting that conclusions as to the status of some of Two points, however, seem clear: the need of a sound taxonomi¢ basis for all lines of botanical work, and that any classification : to be of practical value must be based mainly upon morphological _ characters. present genera of the Pomoideae might be classed as genetic species under one genus. In at least one case two genera should be combined under one species.” and ardent “combiners” would scarcely approve of treating all of the forms of Crataegus of both hemispheres, including the genus Mespilus, as one polymorphic i o Hybridization, therefore, seems to be a factor that must be reckoned with, but until much more experimental work has ete done it is impossible to say how large a proportion of the recorded species, and in most cases which particular ones, originated ™ this way. , Where hybrids have arisen between two quite distinct and well 1982] PALMER, THE CRATAEGUS PROBLEM 359 marked species or between those of two different groups, it may sometimes be detected by their intermediate characters or peculiar behavior, but this could only be done safely by careful field study and with data as to the species growing in the vicinity where the supposed hybrid originated, and as to the possibility of cross- fertilization. A hybrid between two closely allied species of the same group would be so obscure that they could scarcely be de- tected or identified on morphological characters, and to try to account for forms in this way, without experimental evidence, can be little more than a guess. On the evidence available, as well as on theoretical grounds, it would seem then that the astonishing number of forms of Crataegus that have been detected and described are of diverse, and many of them probably of recent origin. First, there is a very large number of species and varieties that have arisen by what may be called, for want of more precise knowledge, the normal processes of evolution, and these can naturally be accounted for by the instability of triploid and polyploid species: second, there is probably a small number of hybrids between diploid, or between diploid and poly- ploid species, which may be expected to follow the Mendelian law of variability: and third, there is a perhaps larger group of hybrids between mostly triploid or polyploid forms, having pollen wholly or partly sterile, but which produce seed apogamously and there- fore reproduce very closely all of the characters of the parent Plants, thus fulfilling all of the evident requirements of normal ! ea disposition then should be made of the forms that 8 een described as species, of various degrees of distinctness an sans bably of hybrid 2. stability, a considerable number of which are pro origm? _ Tame and treat it in all respects as a species, giving 4 (those less distinct. Where the evidence 1. a apg ) oes so, it will be desirable to indicate those species cael Ae hybrid origin, but the general student who 1s interested m ee - > sina basis for classification need not be greatly concern and where these seem sufficiently distinct to retain th "he @ oye { ‘ that will neither seek to reduce arbitrarily P bly true of all taxonomic work. But it would seem that such progress has now been out along conservative species by throwing recognizable, nor to T as synonyms forms that are clearly ier le problem of fons ve JOURNAL OF THE ARNOLD ARBORETUM range, may at least do something tow vee _ where it will not be a hopeless task for 1993] + SAX, CHROMOSOME RELATIONSHIPS IN POMOIDEAE _ 363 CHROMOSOME RELATIONSHIPS IN THE POMOIDEAE Karu Sax With plate 50 Taz Rosacear include four tribes or subfamilies which “are all closely linked together by their floral characters” (Rehder 1927.) The basic chromosome numbers are 8 and 9 in the Spirae- ‘idee, 7 and 9 in the Rosoideae, 8 in the Prunoideae, but it is 17 in all the genera of the Pomoideae. Polyploidy is found in all of these subfamilies although in the Pomoideae it is limited to tri- ploids and tetraploids with very few exceptions. According to Darlington and Moffett (1930) the 17 pairs of chromosomes in Pyrus are made up from a basic number of 7 by 4 duplication of 4 pairs and a triplication of 3 pairs of chromosomes. The Pomoideae are considered as complex autopolyploids and their norphological characters are attributed to this reorganization of 7 basic chromosomes. These conclusions are based pn the aig: that there is a tendency for the chromosomes to be associated | Stoups of two or three pairs at meiosis and the fact that the basic mosome is 7 in the more important genera of the Ro- ‘ideae. These authors also describe quadrivalent and sexivalent _ Gromosomes at meiosis in diploid species. The writer (Sax, 1931) and Moffet (1931a, 1931b) have found | aes genera of the Pomoideae have 17 chromosomes 8 bed = their number. The different genera are closely allied seventh i te eical characters and their breeding and 8 pe ee ea They have undoubtedly had a common origi 2% ao mosome behavior should be similar in all genera. ed tee peat genera there is a tendency for the bivalents to be groupe Wee first meiotic division, but different genera seem to Vary : ee Pehle The chromosomes of Crataegus and ingraat! 2 ae W secondary association to such an extent that “ae the . thas are difficult, but in Sorbus, Aronia and Amelanchier cag, MOsomes of diploid species are well differen’ ree drain there is some evidence of secondary erat set ns in an titetPretation of the quadrivalent and sexivalent ass0° age egy Species may well be questioned. In cay ? ge shows by Debknnte e ee mine ia. contact between oe pene | . _ 18, although one diakinesis figure (text-B8- *'/ ns “xivalent, four quadrivalents and six bivalents. The con 364 JOURNAL OF THE ARNOLD ARBORETUM [vou xm necting strands at late metaphase are about .05 micron in diameter and the difficulty in following these strands in side views in a _ Metaphase figure containing 17 chromosomes makes any con- clusions concerning multivalent associations rather questionable, It is probably significant that “multivalent” associations from side views were found far less frequently than those derived from polar views. Later investigations by Moffett (1931b) show that there is little, if any, real multivalent pairing in diploid Pomoideae. At dia- kinesis he found that “in the great majority of divisions examined 17 bivalents were observed.” The occasional “* quadrivalents” shown in diploid species probably do not represent true pairing of four chromosomes but is simply a grouping of two bivalents due to secondary association. If the Pomoideae are autopolyploids with sufficient homology and chiasma formation to permit multivalent chromosome as- sociation, the chromosomes in a triploid should form either multi- valents or bivalents or both, but few if any univalents. Theres, however, a large proportion of univalents in triploids. Only two or three univalents were found at metaphase in Pyrus by Darling- ton and Moffett, but at anaphase as many as 9 were found. In Pyrus minima (= Sorbus minima), Moffett shows as many as 12 to 15 univalents at metaphase and anaphase. The number of lagging univalents at anaphase is undoubtedly a better index of chromosome association than counts made at earlier stages, 2 simply confirms the conclusion that the multivalent associations found in diploids do not represent true pairing of chromosomes. _ The writer has studied several triploid forms in the Pomoideae and has found a large proportion of univalents in all cases. °” baronia alpina, a hybrid between Sorbus Aria and Aronia arbuti- foha (Rehder 1926), is a triploid. Judging from the appearance of this tree it has two sets of Sorbus chromosomes and one set a Aronia chromosomes. At the first meiotic division there are about 17 bivalents or trivalents and from 6 to 15 univalents. +° views of the first meiotic division are shown in Figures 1 and # the The univalents are usually found around, or at one side of, MW’ bivalents and trivalents and usually not in the same plane as show? in the side views. (Figs. 3 and 4.) It is not possible to distingus> between bivalents and trivalents in polar views and even from sil® views of the divisio fig Po ee ee moe” nee t asily differentiat There are, however, usually 17 pairs or multivalent ygennre ot : of chromosomes so that the deficiency of univalents pert ae accounted for on the assumption that some trivalents are fog ie 1932] SAX, CHROMOSOME RELATIONSHIPS IN POMOIDEAE 365 A few trivalent chromosomes can be observed from side views. If pairing occurs between the extra 17 chromosomes we would expect about 24 bivalent chromosomes and one univalent. At anaphase the univalents lag behind and ultimately divide. The most usual number of lagging univalents found is 11 or 12. (Figs. 5 and 6.) This observation is in accord with the numbers found at metaphase. Sorbopyrus auricularis bulbiformis is also a triploid and as pre- viously described (Sax 1931) it has about 12 univalents at the first meiotic division. (Fig. 7.) Malus theifera is another triploid which shows some univalents at the first meiotic division. (Fig. 8.) The divisions in the pollen mother cells are very irregular and practically no pollen grains are found in the mature anthers. The division in the megaspore mother cell is more regular and apparently trivalents are formed with only an occasional univalent chromosome. This species sets an abundant crop of fruit, the seeds are fertile and the seedlings are all true to type. The fact that this species is a triploid and has no functional pollen, but breeds true from seed, indicates that apomictic development occurs. The details of embryo origin and development have not yet been determined. The evidence from triploid Pomoideae clearly indicates that Ploids derived from parental types with 8 or 9 chromosomes. #3 ther subfamilies of the’Rosaceae the basic chromosome qumbers for almost all genera are 7, 8, or 9. In the Spiraeoideae the basic ; preys is 8 for Spiraea and Exochorda and 9 for Physocarpy’s, entactina, and Sibiraea. Hexaploid species have been found in : -__ im the Rosoideae the basic chromosome numbers are 7 and .- cg The larger genera such as Rosa, Rubus and Potentilla have er ’ amber of 7 chromosomes, but the monotypic genera arts ima and Neviusia each have 9 pairs of chromosomes: wines of \ 10, 11) The writer (Sax 1931) previously reported 8 pairs tly ‘romosomes in Rhodotypus and Neviusia, but there are appare® "S closely associated that 366 JOURNAL OF THE ARNOLD ARBORETUM (vou. xm counts are difficult suggests that some chromosome duplication may have occurred in these genera. If the Pomoideae have been derived from other subfamilies in the Rosaceae, it would seem that the Spiraeoideae and Rosoideae were involved. My colleague Mr. Rehder tells me that the Pomoideae are taxonomically more closely allied to the Spiraeoideae than to either of the other two subfamilies. Crosses between earlier types of Spiraeoideae or per- haps between primitive forms of Spiraeoideae and Rosoideae may have been the basis for the origin of the Pomoideae. Hybrids between forms close enough to cross would indicate some chromo- some homologies although perhaps not close enough for chromo- some pairing. Doubling of the chromosome number in the Fi hybrid would insure fertility and an allotetraploid would be pro- duced with 17 chromosomes. The recent production of allotet- raploids in different families shows that such an origin of a new type of plant is quite possible. i In such an allotetraploid there might well be sufficient affinities to produce some secondary pairing of bivalent chromosomes although real pairing to form quadrivalents would rarely or never oceur. Some secondary pairing may have been present in one of the contributing diploids as suggested by the chromosome behavior in Rhodotypus and Neviusia. Lawrence (1931) has presented good evidence that secondary association of bivalents does occur in many genera, apparently due to remote affinities between chromosomes which are too well differentiated to permit multivalent chromosome pairing. As Lawrence points out such allopolyploids would have a high survival value due to hybrid vigor and a high degree of fertility. The differentiation of gene* and species within the Pomoideae can be attributed primarily mutations or minor changes in the 17 pairs of chromosomes. SUMMARY & Sorbaronia alpina, Sorbopyrus auricularis and Malus theifera #° triploids. At the first meiotie division in the pollen mother ce™> — there are about 17 bivalents and trivalents, and from 6 to peed valent chromosomes. The fact that about 12 univalents are usualy — , found in triploid forms of Pomoideae shows that this subfamily ae is not an autopolyploid with a basic number of 7 chromosomes § several writers have suggested. u ' a The basic chromosome numbers jn the other subfamilies the Rosaceae are 7, 8, and 9. The Pomoideae may have originat from one or perhaps two of these subfamilies by hybridizatio? between different primitive forms followed by chromosome 4 aret: 2. #h, eet re. as 2 ®a,' e° sf eS eg» i ES] Se See Coaee' e BY ee ' ce pees * Meese rth “5 ? ‘ *e SB im Be 2? 2 Be vs eS A ose ‘ Se 2ee a @ eo 8 . we q 2 : d Piate 50 12 AB. CHROMOSOME RELATIONSHIPS IN THE PoMOIDE Ops ewes hee SMe + Pelee ae te Pp tgee hee 1982] SAX, CHROMOSOME RELATIONSHIPS IN POMOIDEAE ~ 367 in the F, hybrid. Remote chromosome affinities are indicated by secondary association of bivalents in the Pomoideae. True multivalent chromosome pairing rarely, if ever, occurs in “diploid” species. The available evidence seems to indicate that the Po- moideae are allopolyploids. Arnotp ARBORETUM Harvarp UNIVERSITY LITERATURE CITED Daruneron, C. D. & A. A. Morrerr (1930). Primary and secondary chromosome balance in Pyrus. (Jour. Gen. 22: 129-151. Lawrence, W. J. C. (1931). The secondary association of chromosomes. i ) Res e. Roy. Soe. 10 8, A (1927). Meanie) of roanivated trees and shrubs. pp. 930. Mac- Ka 0., New York. i Arnold 4 (1931).. T is origin and relationships of the Pomoideae. (Jour. 12: 3-22.) EXPLANATION OF PLATE 50 Figures 1, 2, 3, 4, 5, and 7 are from permanent smear peepee of ane mother cells. ot from ie mes : from 9 to 1-6. Sorbaronia alpina, firs tie division, showing from i : pina t meiotic " g 15 univalents at different stages and about 17 berelente and ie | vg cot h Soop auricularis bulbiformis, first metaphase, with 14 univalen - Figure 8. Malus theifera, a triploid with univalents and bivalents oF A as ea aa os : al llen is produced. | Figure 9. Nei onan nts oe fexyaose! polle: of chromo- ——-Bigure 10, “wom scandens, first metaphase, 9 pails eS Figur e ia Kerria japonica, first metaphase, 9 Tienes eee ee a aay metaphase, 9 of chromosom: of pairs of + Physocarpus monogynus, first me metaphase, 9 pairs dy er are | preparations. Magnification X 3000, packing Figure 8 i 2100. SrA ea ih Sauteed ele als 368 ~ JOURNAL OF THE ARNOLD ARBORETUM [vou. xm CHROMOSOME PATRING IN LARIX SPECIES Hatty Jouivetre Sax With plate 51 and one text figure Tue Conirers show a remarkable uniformity in chromosome number. With few exceptions all the genera have twelve pairs of chromosomes (Tischler 1927). No polyploid species have been reported, and the meiotic divisions seem to be very regular. Chro- mosome behavior in species hybrids has never been described in the Gymnosperms. This study was undertaken to determine whether or not there is any cytological basis for the uniformity in chromosome numbers in the Conifers, and to analyze the chromosome behavior in : species hybrid. The meiotic chromosomes have been examined in Pinus, Tsuga, Taxus, Picea, Pseudolarix, Cedrus, and Lanz. — The meiotic chromosomes are similar in all the above genera. The present analysis will be confined to two species of Lariz and at F, hybrid. A more detailed analysis of chiasma formation and chromatid relationships at meiosis will be published later. Larix eurolepis Henry is a hybrid between L. Kempferi Sarg. and L. decidua Mill. This hybrid is intermediate in appearance betwee _ the parental species, but possesses considerable hybrid vigor and : matures its cones earlier than either parent. An account of ” : origin and characteristics of this hybrid has been recently published by Anderson (1931). ES bit Larix Kempferi is a native of Japan, while L. decidua is indig- enous to northern and central Europe (Rehder 1927). ae distribution of the two species would indicate that the parental types had been separated for long periods of time. of The material for the present paper was taken from specimens ® — division in the hybrid was between that of the two parents. qian |: the last week in February the pollen mother cells were four’ : dividing in L. Kempferi. L. eurolepis produced dividing pole? — mother cells throughout the first week in March. The ages t a L. decidua occurred the following week. Other specimens of L a decidua, from a different part of the Arboretum and from 4 i tee slope, were a little more advanced in this respect. Material ¥* ‘ obtained over a period of several days in each case. eee SAX, CHROMOSOME PAIRING IN LARIX 369. Permanent smears of the pollen mother cells were made. A mumber of fixatives were used. Darlington’s modification of Flemming’s fixative proved to be the best for Larix. The crystal violet-iodine stain gave the best results for work on the chromosomes. Within a given genus chiasma frequency in the meiotic chromo- somes is evidently a good index of chromosome affinities and provides an accurate method for measuring the degree of chromo- some pairing in species and generic hybrids. Aase (1930) noted the greater frequency of univalents and open (single chiasma) types of bivalents in Triticum X Aegilops hybrids than in the parents. Although Aase did not use the term “chiasma” in her description of chromosome pairing, the different types of bivalents Were observed and discussed. Darlington (1931b) and Hollings- head (1982) have corroborated Aase’s observations in T riticum hybrids, and they have discussed the problem from the stand- point of chiasma frequency. THE MEIOTIC CHROMOSOMES. The twelve bivalents from one nucleus of Larix decidua are Min figure 1, These were drawn from a side view of the metaphase figure. Each chromosome was drawn separately so that details of pairing might be observed. The chromosomes of L. eurolepis, the hybrid, are shown in figure 2, and those of L. Kzmpferi. in figure 3. __ The nodes are undoubtedly chiasmata, and the cross formed : aoe exchange of partners among the four chromatids can be served in many chromosomes. Most of the pqananner te chiasmata are terminalized with difficulty, and the chromosomes Ct tetween the chiasmata. In the smear preparations , - i. observed, although, with the same fixative, paraffin sections ‘ie ag spindle fibers. : | ; _ ‘ting of bivalent chromosomes in Larix an : _ _, Sto be effected only by chiasmata,—a condition Nee of La {1931a) ielieves to be truc for all genesa. THERE > beet t chiasmata at diakinesis and early metaphase indicates prophase other Conifers 370 JOURNAL OF THE ARNOLD ARBORETUM fvou. xm CHIASMA FREQUENCY IN LARIX SPECIES AND F, HYBRID Chiasma frequency can easily be observed in Lariz at diakinesis and metaphase of the meiotic divisions. Counts were made in pollen mother cells where all twelve bivalents could be clearly observed. The data are summarized in Table I. TABLE I Chiasma frequency in Larix species and F, hybrids = tod + e feg 8 S| 8s 5 Number of Chiasmata Be] a ce a] 8: g 5 25 oe 5 o8 0 tes Zid eS 46 jan) Pe Oe ay Ee ae ee 5 EES Ae EELS Seer Larix decidua 37 19 | 279/110] 36 | 444 | 2.36 + 02 SES UNS ere 4% | 68%) 25%| 8% | Larix eurolepis 49 | 204] 198| 66 | 1 | 612) 245+. SLOPES EES ea | 8% | 48%| 32%|10.7%| .001% Sascis Larix Kempferi 88 | 223/139] 98 | 4 | 552 | 247+ 8 fee Ces a =, 40%| 25%| 18%|.007%| it In Larix decidua the chiasma frequency ranges from one to four per bivalent with an average of 2.36. In L. Kempferi the range is from one to five with an average of 2.47 chiasmata pe bivalent. No univalents or precocious divisions were found these species. In the F, hybrid (L. eurolepis) the chiasma frequenY is from zero to five, and the average per bivalent is 2.45. Siu hybrid is intermediate between the two parents, although the chiasma frequency is nearer that found in L. Kzmpfert. differences in mean chiasma frequency are of doubtful puna although these differences between the parents and between ™ decidua and the hybrid are about three times the probable errr- Univalents or precocious divisions of one bivalent were foun - about eight per cent of the pollen mother cells of the hybrid. - ie. mee EES | 4 practically all cases observed, these single were P aa to opposite poles, so that a normal distribution of cl aid dea be expected ay The distribution of chiasma frequency is shown graphi e ve in Text figure 1. The curve of chiasma frequency in the spre ' 168, PP ml SAX, CHROMOSOME PAIRING IN LARIX 371 which Haldane (1931) attributes to interference in chiasma forma- tion. éo4 50 5 : - : ees. — 40. i 4. 6 K. 5 - 7 Ms ey Ba ‘ I: \ > 30 A \ i \ v gr . - ] Nee ‘je bi g 20. Xo uv ‘ ‘. v os . * > Mies fe é * 10 a 7. y . - / AS » Z te \. ‘ Loe. Ne a 2 ear 4 5 Number of chiasmata per biv alent. as the similarity i j 3 y between larity in chiasma frequency ied Counts TABLE II | Be) Pollen sterility in Larix species and F: hybrid ey pope .|) peerem Gs | ahd a tl mic oe 372 JOURNAL OF THE ARNOLD ARBORETUM DISCUSSION The uniformity in chromosome numbers for most genera of the Gymnosperms, and the absence of recorded cases of polyploidy, can perhaps be attributed to the type and frequency of chiasma formation at meiosis. With an average of about 2.4 interstitial chiasmata per bivalent, any autopolyploids produced would be expected to form closely paired tetravalents. The segregation of homologous chromosomes in such polyploids would probably be too irregular to produce a high degree of fertility, and the polyploid would have small chance of survival. Very few polyploid species occur in nature where the chromosomes pair as quadrivalents at meiosis. The only known case where a high percentage of quad- rivalents is found in fertile species is in the tetraploid Tradescantias — (Anderson and Diehl 1932). In these species the spindle fiber — attachment point is approximately median, and the chiasmata are _ largely terminal in both diploids and tetraploids. The chromosome — morphology and types of chiasmata formed might permit rather free movements of the homologous members of a quadrivalent, so that regular chromosome distribution and a high degree of fertility might be expected. With a high frequency of interstitial chiasmata, regular assortment and fertility would not be expected in autopolyploids. The absence of polyploid species of Conifers may also be dependent on other factors, in addition to the type i and frequency of chiasma formation. completely homologous, but the average chiasma frequency found in the hybrid is not lower than that found in the parents. ee About nine per cent of the pollen in the F, hybrid is morp logically imperfect as compared with two or three per cent eRe ey in the parental species. The increased pollen sterility of ™° — hybrid can not be due to irregularities in chromosome distribute? — ? at meiosis. SAX, CHROMOSOME PAIRING IN LARIX 373 SUMMARY A detailed analysis was made of chromosome pairing in Larix Kempferi, Larix decidua, and a hybrid between these two species (Larix eurolepis). The chromosome number and the type of chiasma formation is similar in the three species and in the re- mainder of Conifers studied. The comparatively high number of chiasmata, together with the prevalence of the interstitial chiasmata formed here, may account largely for the great uniformity in chromosome number and general stability in the group. The chiasma frequency was remarkably similar in all three species. The average chiasmata frequency per bivalent was 2.36 for L. decidua, 2.47 for L. Kempferi, and 2.45 for L. eurolepis. Although there were a few cases where there was weaker pairing in the homologues in one bivalent of the hybrid, chromosome distribution appeared to be regular, and the average chiasma cy was as high as that in the parents. The pollen sterility Was somewhat greater in the hybrid than in the parents. This greater pollen sterility in the hybrid can not be attributed to ir- regular chromosome distribution or to difference in chiasma fre- quency, Kini, LITERATURE CITED a AH (1930). Cytology of Triticum, Secale and Aegilops hybri i mae to pirlanoy. » (Rear studies of the State College of of 65-68 EpGar (1931). Hybrid trees. (Arnold Arb. Bull. Pop. Inform. ‘oonssox, Ep Gar & D.G. Diexi eves Puyeedae ih) the Tradescantia Problem. Jour. Arnold Arb. 13: 213-231.) | 2muNeron, C. D, (1931a). Mei, * (Bic Rey. 6: 221-264) sum 24) egy Sybtey The analysis of chromosome pairing Re : wh Ie A ogia, 3: ES ‘ 3 sh) Hatpane, J. i, 3: : (1931). The cytological basis of genetical interference. _ -AOLLINGSHEAD, Lintian (193 The occurrence of unpaired chro ee yt hybrids between eh sf Triticum ehbr poet (Cytologia, 3: jas j Mapa, Ae : : pollen mother i pee Vicia Faba L. (Mem. Coll. Sei, Kyoto Imp. Univ. Ser- B.5 a Rewer, Atrrep (1927). Manual of cultivated trees and shrubs. T ‘ hehe | tabule a G. (1927). Pflanzliche Chromosomen-Zanien- onderabdruck aus Bd. IV.) 874 EXPLANATION OF PLATE 51. The figures in Plate 51 were all drawn from 62 views of pern of pole mother cells fixed in a modification of Flemmin stained with crystal Tapa Ba Each of one cell at pose metap. The drawings of a camera Magnification x 2600. ; } ta distinct, few terminal, ed orga (a) Two subterminal ata. (b), Stasieh! 9 haets semaines the Hee lord sé Lei decid ) " 20 L eurolepi. i lua). a, a, U : les. b, ¢, cumendalps jack with chiasma ent. Figure 3. Larix Kempferi. Slightly earlier than above. PLATE 51 Jour. ARNoLD ARB. VOL. XIII. SAX, MEIOSIS AND CHIASMA FORMATION IN PASONIA 375 MEIOSIS AND CHIASMA FORMATION IN PHONIA SUFFRUTICOSA Karu Sax With plate 52 and two text figures Tur Generic evidence has shown conclusively that crossing over occurs at the four strand stage and between only two of the chromatids at any one locus (Bridges and Anderson 1925, et al.). An actual physical interchange of chromosome segments has been shown to be correlated with genetic crossing over in Drosophila (Stern, 1931) and in Zea (Creighton and McClintock, - 1931). uch an interchange of chromosome segments should invariably produce an asymmetrical arrangement of the chromatids. If chiasmata are produced by previous crossovers, each chromatid will not always lie in the same quadrant at all loci. A study of : chromatid organization at meiosis should throw some light on the ___ hature of chiasma formation. ie Tn this discussion the term chiasma is used to denote an apparent change of partners among the chromatids at meiosis without Pretation of chiasma formation have been described and by Wenrich (1916), Rob 916), McClung (1927), Wilson 9 (1916), Robertson (1916), 4080), and others. g of paired chro: 1. As McClung has pointed out, tl | ete also show the chrom a | a : eae 376 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Fie. 1. Diagrams or CuiasmMa ForMATION. e pair x ie shenaail 5 Is represented by white rods enyae pees aise a bla The spindle fiber attachment points pre Re s lines. The term town: 1s used to designate the change of pa to among the four chranaatde of a meiotic chromosome without reference Diagram 1. Chiasma formation as interpreted by McClung and pee A chiasma is caused by the alternate o opening out of sister and non diet chromatids in pairs. dig viewed from the end, each chromatid me in its position in the sa e quadrant at all loci. Alternate internodes lie pe at right — ‘te each other. arti : lagrams 2 an a 3. Chias pb Ringe neta rg: Hyyhkae eecaagt p chiasmatypy hypothesis. Each e ist t of a pre between pie of ‘the four chromatids. Two of the chromatids i ae oroutai 1932] SAX, MEIOSIS AND CHIASMA FORMATION IN PZONIA 377 their position in the same Pie at all loci. The pairs of chromatids on both sides of a chiasma should tend to lie in the same plane hiasma formation amodling se eClub interpretation, but with 2 chromatids i in different papebent at different loci, due to a half- twist in one pair of chromatids before pairin Diagram 5. A chiasma similar to iene 1 of diagram 4, but the half twist is caused by a crossover resulting from breaks in two chromatids of 4 previous median chiasma. A bivalent of this type sieht lead to inter- locking of homologous chromosomes at the first meiotic metap , and in respect resembles the configuration shown in diagram 2. ‘ Darlington (1930, figs. 8, 9, 10) and Moffett (1932, diagrams 2A, 2B), have published diagrams of chiasma formation which are in accord with McClung’s interpretation. More recently Darlington and Dark (1932) have shown figures of bivalent chromosomes of Stenobothrus where the relations of the four chromatids can be observed at all loci in some of the chromosomes. The chromatids in the shorter chromosomes usually show a symmetrical arrange- ment, as would be expected if chiasmata are produced by alternate opening out of sister and homologous chromatids at diplotene. In the long bivalents pictured by Darlington and Dark it is difficult to follow the position of the chromatids at all loci, but some of the - chiasmata in these chromosomes are clearly symmetrical, although some bivalents (D, fig. 6) do show asymmetrical arrangements -(?) “compensating” (reciprocal) id never diagonal (equati In Drosophila both types of crossovers occur with about equal frequency. “.oaxemed to Jannsen’s sate ee chiasmatypy hypothesis, + between two of the four 3). This epbeepretation has recently been sponsored oa Belling Hee Darlington (1931), ae others. ies all cases it is assumed that sister oe ids are age an quadrant at al loci, and the two Somat which form the : rel should be adjacent and not diagonal. ae t stages the pairs of chromatids ‘ plane on either side yeas chiasma and not at right angles to L aa As shown in another paper (Sax 1932), it is also necessary Ue sume, on Belling’s hypothesis, that crossovers occur saat _ Sister chromatids which would produce many twists mm pall ‘Sister chromatids between chiasmata. Certaip rtain types of double _ @0ssovers should result in interlocking of homologous ee ig riethee, as would be the case if the spindle fiber were ie : m diagram 2 (Fig. 1). It is possible, of course, that the movemen occur between 378 JOURNAL OF THE ARNOLD ARBORETUM of the chromatids would occasionally result in a symmetrical configuration where each chromatid would lie in the same quadrant at all loci, but such figures should be rare. The most critical studies of chromatid organization and the nature of chiasma formation have been based on preparations of Orthopteran chromosomes. In this group of animals the four chromatids can often be recognized at all loci at the diplotene ft stage. In plant species, however, the chromatids are usually closely paired in spiral chromonemata during late prophase and early metaphase stages so that the individual chromatids can seldom be differentiated until the anaphase stage of meiosis. Newton’s (1926) figures of Tulipa and Fritillaria chromosomes do show the tetrad nature of the meiotic chromosomes to some extent. According to Newton, the hypothesis “which explains the dia- kinetic figures as due to the opening out in two planes at right angles of what are originally four parallel chromatids, is adequate to explain the events of diakinesis and division in Tulipa and Fritillaria.” Taylor’s (1930) figures of Gasteria chromosomes show clearly the alternate opening out of the chromatids in pairs at right angles to each other, with each chromatid in the same quad- rant at all loci. Some of his figures also show the chromatids opening out in pairs at the proximal end, but at the distal end all four chromatids are paired. Such chromosome pairing without chiasma formation can not be reconciled with Darlington’s (1931) theory of chromosome pairing at meiosis. According to Darlington, — ciremosomes are associated only by chiasmata at the first meiotic division. In many species bivalent chromosomes are apparently associated only by chiasmata at the meiotic metaphase numerous species of plants it is clear that alternate internodes — of meiotic chromosomes are oriented in planes at right angles to each other, as is the case in the multiple ring chromosomes Orthopteran species. : CHIASMA FORMATION IN PAONIA Pzonia suffruticosa has five pairs of large chromosomes. The a haploid set of chromosomes at metaphase of the microspore di vision is shown in text-figure 2. Four of the chromosomes et ; an approximately median — spindle fiber constriction, while fifth chromosome has a subterminal fiber attachment with a trabant at the end of the short arm. The somatic perenne” are long and slender, and even at metaphase the paired chroma are often twisted about each other (Figure 1 of pl. 52). If the sister chromatids -of homologous chromosomes are twisted at - [voL. xm ; .. : : | 1932] SAX, MEIOSIS AND CHIASMA FORMATION IN PZONIA 379 time of pairing at meiosis, then the chromatids will have an asym- metrical arrangement at the diplotene stage. Such meiotic chro- mosomes should often show half twists in paired chromatids, even if no crossovers have occurred. The chromosomes at the meiotic prophase were so diffuse and granular in appearance that any critical study of early chiasma formation was impossible. Relatively few chiasmata seem to be present, even at the early prophase. At diakinesis the number of nodes or chiasmata can be clearly observed, but the relation of the chromatids was obscure. Fic. 2. CHromosomes IN PONIA MicroOSPORE. The fi bserv d at the first meiotic our chromatids can usually be 0 e 1 ie iiaaaiile can be determined at all loci. The PRLS are orverage in pairs at the spindle fiber attachment poin other loci they may be well separated. Many of the meiotic chromosomes are P paired without pant mation of chiasmata (Figures 4a, 4d, 7a, 10b of pl. 52). ean a. chromatids appear to be paired throughout their lepé” © 1. ane of partners at any locus. ee < matids have opened out in pairs, but a Oh chromatids seem to be paired. If non-sister chromatids ; ae . figures were to open out in pairs at the distal ends, nape Only _ €Xpect a chiasma to be formed sr petyines eho oe _ Inrare cases do the distal ends of the ma __ the cross-shaped figure which is so typical in ™ rT eed a In many cases, however, there is some separation Geues @, 5, hopes at the distal ends of the erreeeed 380 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm lla of pl. 52). Homologous chromosomes which are paired at both ends usually show the sister chromatids paired only at the spindle fiber attachment and pairing of non-sister threads only at the distal ends (Figures 4b, 6b, 7c, 7d of pl. 52). alf twists in paired chromatids are often observed (Figures 2a, 3d, 4a, 8a, 10a of pl. 52). Such half twists between the fiber attachment and the first chiasma would be expected even if no crossing over occurs. : If crossovers have occurred in the meiotic chromosomes, there should be half twists or an asymmetrical arrangement of the chromatids, but in many cases the chromatids are symmetrical and lie in the same quadrants at all loci (Figures 4c, 4d, 5, 6b, Ta, 7b, 7d of pl. 52). In the chromosomes with a single chiasma it is possible that any earlier asymmetrical relations of the chromatids might be straightened out, in some cases, by the movements of the chromatids during the early stages of division. In chromosomes with a median fiber attachment and two terminal chiasmata, such a change in the relation of the chromatids would be improbable. In some of these chromosomes there are half twists in the chroma- tids (Figures 4b, 7c, and 8a of pl. 52), but other bivalents show the symmetrically arranged chromatids in the same quadrants at all loci (Figures 6b and 7d of pl. 52). Half twists in the chromatids of such chromosomes may be due either to twisting before the homologues pair or to crossing over. No crossovers could have occurred in the symmetrical chromosomes, even though two chiasmata are present in each pair of chromosomes. Although these chromosomes do not have the usual types of chiasmata, quadrant at all loci (Fig. 1, diagram 1). Such figures are very difficult to interpret on the basis of the partial chiasmatypy hy- pothesis. | : There are, however, some chromosome configurations which do support the partial chiasmatypy hypothesis. In these chromo- somes the chiasma cross is formed by chromatids which ver adjacent, rather than diagonal, at the four strand stage (Higart 5 6a and I1b of pl. 52). In these figures the two upper chromatids form the cross and lie above the other two chromatids on each side os of the cross. Such a relation of the chromatids should be character- HM istic if chiasmata are caused by previous crossovers, as shown 12 diagrams 2 and 3 (Fig. 1). | j This type of chiasma is difficult to account for on the hypothesis Coe that chiasma formation precedes crossing over. If no twists — oe 1982] SAX, MEIOSIS AND CHIASMA FORMATION IN PHONIA 381 in pairs of sister chromatids, the chromatids which form the cross will be diagonal on each side of the chiasma (Fig. 1, diagram 1). If a half twist occurs in one pair of sister chromatids, the chromatids which form the cross will be adjacent on one side of the chiasma, but will be diagonal at the distal end, as shown in diagram 4 (Fig. 1). A rotation of the chromatids in the lower left-hand arm of this chromosome would produce a configuration like those found in figures 6a and 11b (PI. 52), but such behavior of the chromatids might be expected only in rare cases. The homologous chromosomes occasionally fail to pair or are separated before the first meiotic metaphase. The univalent _ chromosomes may pass to opposite poles (Figure 4 of pl. 52), or _ Some. Where several chiasmata are form of chromosomes would be expected, and such types of bivalents are E they may pass to the same pole (Figures 8 and 12 of pl. 52). The segregation of both homologous univalents to the same pole will result in genetic “‘non-disjunction”’ if the gametes are viable. (For further discussion, see previous paper, Sax 1932.) Occasionally a bivalent chromosome divides with apparent culty, even when there is no interlocking of paired chromatids. In figure 12 (Pl. 52) a bivalent is shown held together by the attach- stein of two single chromatids. Such figures are common in many Species. : No interlocking of homologous chromosomes Was observed in Pzonia, and none would be expected where only one chiasma 1s formed between the spindle fiber and the distal end of the chromo- ed, some interlocking found in the Orthoptera, and in Lathyrus and Lilium. If chiasmata are formed by previous crossovers, the configuration: shown a 4 diagram 2 should result in interlocked chromosomes if the spi fiber attachment were terminal. On the writer's hypothesis @ crossover between two unbroken chiasmata should also result in o interlocked homologues at metaphase (Fig. 1. diagram 5). PRE- AND POST-REDUCTION Be division, and cs eat: Bini ty: 7 i g Since : — association is often especially clear during interphast served 382 - JOURNAL OF THE ARNOLD ARBORETUM [vou. xm seems probable that sister chromatids are usually held together at the fiber constriction, as is clearly the case in Drosophila. Such an association of chromatids would mean that the first meiotic division is reductional at the fiber constriction and for all loci between the fiber constriction and the first crossover. Between the first cross- over and the distal end of the bivalent chromosome the division is equational. If a second crossover occurs on the same side of the spindle fiber, the loci distal to the second crossover would undergo pre-reduction at the first meiotic division. Wenrich (1916) and Carothers (1931) have described both pre- and post-reduction in heteromorphic chromosomes. According to Wenrich (figure 65) chromosome “C” in Phrynotettic forms only a single chiasma. One of the homologues is shorter than the other, and the two short chromatids are always paired. In about half the cases the first meiotic division is reductional, and in half the cases it is equational. On any theory of chiasma formation it would be impossible to obtain both types of reduction with only a single chiasma and a constant association of short chromatids unless the spindle fiber can be attached more or less at random to either end of the homologous chromosomes. But the spindle fiber attachment point seems to be constant for individual chromo- somes of both plants and animals. In Wenrich’s figure 64, showing the behavior of the unequal chromosome pair B, the segregation is equational, but in order to have an equational division, it would be necessary to assume that non-sister threads are paired at the fiber constriction or that a crossover has occurred in all cases. Similar heteromorphic chromosomes in Trimerotropis have been described by Carothers (1931). The unequal homologues may undergo either pre-reduction or post-reduction, although m the few figures showing chromatid association, only a single chiasma 1 present, and the short chromatids are always paired. If the first meiotic division can be either pre-reductional or post-reductional, ? it is necessary to assume that some crossing over occurs if sister chromatids are always paired at the point of spindle fiber attach- ment. The behavior of unequal homologues should receive further study in connection with the implications involved concerning crossing over and the nature of chiasma formation. ipo< SUMM™M If chiasmata are formed by the alternate pairing of sister and non-sister chromatids, each chromatid may lie in the same quadrant at all loci. An asymmetrical relation of the chromatids can iar only if sister chromatids are twisted before or after synapsis or a crossover has occurred. eee eels a SAX, MEIOSIS AND CHIASMA FORMATION IN PONIA 383 tt chiasmata are the result of previous crossovers, the chromatids "should not lie in the same quadrants at all loci, and symmetrical arrangements of the chromatids should seldom be found. . In Pzonia suffruticosa the homologous chromosomes may pair at meiosis without chiasma formation. When chiasmata are found, ‘the chromatids are often symmetrical in the bivalent chromosomes. The symmetrical relations of the chromatids in bivalents with two chiasmata are very difficult to explain on the partial chiasmatypy hypothesis. These figures are essentially the same as the types found in Orthopteran species. Most of the asymmetrical configurations found in the Pzonia r chromosomes could be attributed to half twists in paired sister chromatids. Some chiasmata are found, however, where the - chromatids forming the cross are adjacent on both sides of the chiasma, Such a relation of the chromatids supports the partial Fe coeelgd hypothesis of chiasma formation. Occasionally one or more homologous chromosomes are not - pained at the first meiotic metaphase, but pass to the poles as —tnivalents. The segregation of two homologues to the same pole _ Would result in genetic “non-disjunction.” __, s1some cases homologous chromosomes are paired without chiasma & and post-reduction divisions are discussed in relation to : nature of chiasma formation. vermpreshegy CITED Be > & E. G. Anperson (1925). Crossing over in t Xe ; i triploid females of ae hila melanogaster. Tiicstas E. E. (1 relation to the : - (1926). The Asarcngy? divisions calg”’ howe i 5 + 1931). The maturation rhgy re and iene of hetero- homologous chromosomes in Acridide eridide (Orthoptera). ns a B. & B. McCuwtock (83), A correlation, of eytologi ¢rossing-over in Zea Ma (Proc. N . Acad. Sei “ f C.D. (1930). A cytological Herth ration of crossing over. 221-264.) | ne (1983), The origin and baer? of U. Stenobothrus parallels se dagset 3: 169-185.) 384 JOURNAL OF THE ARNOLD ARBORETUM iron sae a JANNS igteg F. is (1924). La asaya dans les insectes. (La Cellule, 135-359). Sweetpea (Lathyrus peewee with cre reference to ie cyto basis of crossing over. (Mem. Coll. Sci. Kyoto Imp. Univ. 5: 89-123.) gba s e E. (1927). The chiasmatype theory of Jannsens. (Quart. ‘ at Morrert, A. es ye Studies on the formation of multinuclear giant pollen grains ip netics, 25: 315-337. NEWTON, EW. C C. "Y. (1927). Chromosome studies in Tulipa and some related genera. (Jour. Linn. 7: 339- ROBERTSON, W. R.: B. (1916), Chromosome studies. I. (Jour. Morph. 27: 179-332 Me —————.. (1931). Chromosome studies. II. (Jour. Morph. 51: wfc oS K. (1932). The cytological mechanism of crossing over. (Jour. Arnold . ae . 44, ey C. (1931). ’ Zytologisch-genetische Untersuchungen als Be 0 che Theorie des Faktorenaustausch. Aap: Zbl. 51: 37-87, Tayitor, W. R. (1930). Chromosome structure itosis and meiosis. 270. “hs ( us. . Harv: 36. Wirson, E. B. (1925). The cell in Sevehanenint and heredity. pp. 1282. The Macmillan Co., New York. EXPLANATION OF PLATE 52 Figures of ate chromosomes are fro ra eREApe? of pars mother hia is Pzonia suffruticosa, fixed ‘with. praia ’s solution, an stained wit. h esata igaces iodine. : Figure 1. Soma: ; showing twisted chromatids at meta- Figure - Types of chiasma formation at meiotic metaphase. Figure 3. Early anaphase showing the five pairs of chromosomes. | Figure 4. Metaphase showing chromosome pairing without chiasma * rene 7p be ea igure 5. Symmetrical arrangement of the chromatids. Figure 6. Chromosome 6a has an asymmetrical chiasma of the type expected on Jannsens’ partial chiasmatypy hypothesis. The symmetrical arrangement of the hein i chromo- some 6b can not be reconciled with this hypothesis. id Figure 7. Chromosome ‘without chiasma formation (a) and types of chromatid association (c and d) Figure 8. Twisted chromatids in chromosome a. Four univalents at one pole—genetic “non-disjunction.” Figure 9. A type of chiasma formation tion rarely found in Pzonia. bai Figure 10. Chromosome @ shows a partial twist in sister chromatids. Evidence of Becbiieses of chromosome orcangtal Go oe tae eal (b) chiasmata Figure 11. Symmetrical (a) and asymmetrical (b) ch ‘ stent Figure 12. “Non-disjunction” of one pair of homologues and a persis association of distal ends of svsueresaad gy chromatids. Jour. Annotp Ars., Vou. XIII. PLATE 52 oa\ ae rest) 1 BRSe) 9 Vv bY Mer iS “4 * SIS AND Cutasma FoRMATION IN PHONIA SUFFRUTICOSA. iz JOURNAL . OF THE ARNOLD ARBORETUM -Vourme XIII OCTOBER, 1932 Numer 4 _ BNUMERATION OF THE LIGNEOUS PLANTS COLLECTED BY J. F. ROCK ON THE ARNOLD ARBORETUM EXPEDITION TO NORTHWESTERN CHINA AND NORTHEASTERN TIBET (ADDITIONS AND CONTINUATION)! Atrrep REHDER AND CLARENCE E.. KoBusk! Page 20 of vol. rx. insert before LinIACEAE: : GNETACEAE Determined at the Botanical Museum, Berlin-Dahlem | Ephedra mon C. A. Meyer, Versuch. Monog. Ephedra, i ae » fig. 11. (1847). — Stapf i in Denkschr. Math.-Nat. Cl. Akad. : eel , LVI. pt. 2, p. 73, t. 3, fig. xix. 1-9, xxxi. 3 (Art. Ephedra) ee a ee ee he a a i RY th te Se leg ma eet A OS a et ee e Eastern Tiper. Radja and Yellow River gorges: - ‘mong schist rocky slopes of river valley, west of Radja, alt. 3350 bye og + cae om 27, 1926 (plant 5 cm. high; flowers yellow). oe 20 i insert before Smr_ax TRAcHYPODA Nort.: . Simiex Otdhami ii Miquel in Versl. Med. Kon. Akad. Weten. ser. Notton | (1868); in Ann. Mus. Bot. Lugd.-Bat. m1. 150 (1868).— “orton in Sargent, Pl. Wilson. mr. 9 (1916). RAL Kansu: Lienhoa shan, no. 13664, Oct. 1925 over bushes; fruit purplish black). ies reees does not seem to have been recorded from China Ron Vas ee Deas my to }: O38 us cathayana Rehder in Jour. Arnold Arb. xu. 59 (1981). : Kansu. Lower Tebbu country: banks ie pA (1928) for preceding parts of this Enumeration. rt tn A ip {AW : 386. JOURNAL OF THE ARNOLD ARBORETUM [vou xm of Culungapu above Pezhu, no. 14961, Sept.—Oct. 1996 (pyramidal ; tree with broad crown, 15-18 m., leaf pale green above, grayish — beneath). Populus Simonii Carriére in Rev. Hort. 1867, p. 360.—Rehder in i Jour. Arnold Arb. xr. 63 (1931). jag Populus Przewalskii Maximowiez in Mél. Biol. xi. 321 (1881); in Bull. Acad. Sei. St. Pétersb. xxvir. 540 (1882). 3 SOUTHWEsTERN Kansvu. Tao River basi n: Choni, along — bank of Tao River, no. 12110, May 1925 (tree 15-18 m., pistillate); mountains of Choni, in forests and on banks of streams, alt. 3000 m., no. 12114, May 1925 (tree 20-24 m., staminate, catkins bright red). Both specimens are leafless; the pistillate catkins are 2.5-3.5 cm. long with a short-pilose rhachis and short-pedicelled flowers; the ovary is slightly hairy and bears two stigmas. Photographs representing habit and bark of this species and taken by Rock in January 1926 (no. 12110 according to his notes) are In his collection of photographs. In habit the trees resemble Populus nigra L. . Populus nigra Linnaeus, Spec. Pl. 1464 (1753). : NorTHwestern Kansvu. Richthofen range and adjacent region: Kanchow plain, no. 13325, Noy. 1925 (tree 30 m. or more; trunk white). I refer this specimen with some doubt to P. nigra; the branches — are distinctly angular in their upper part, though perfectly terete in their lower third; the leaves are deltoid, truncate at the base, short-acuminate, and of firm subcoriaceous texture, and the bark of the trunk is white according to Rock. = : Populus nigra L. var. italica Duroi, Harbk. Baumz. u. 141 (1772): OUTHWESTERN Kansu. Lower Tebbu co untry: banks of Peshwekiang, alt. 2000 m., no. 14859, Sept. 14, 1926 (tree 24 m., with ascending branches forming an oval to oblong crown; bark grayish white; leaves dark green above, blue green below). 3 According to Rock’s note the tree has not the columnar habit of the Lombardy Poplar, but the branchlets of the specimen are dis- tinctly adscending and the leaves agree with those of P. nigra vat- 3 ttalica. ‘o Populus szechuanica Schneid. var. Rockii, var. nov. te A typo recedit foliis supra in costa et venis villoso-pilosis et sub- _ tus in costa, venis et venulis satis dense et in mesophyllo sparsé pilosis. SOUTHWESTERN Kansu. Lower Tebbu coun try: Want- sang forests, no. 14846, Sept. 12, 1926 (tree 30 m., trunk 1.25 m- | -REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 387 diam., without branches for 15 or 18 m.; bark drab to grayish brown, longitudinally furrowed; leaf grayish white beneath). _ Inshape and size of the leaf and in the 3-4-valved capsule the ‘ ‘specimen agrees with P. szechuanica, but the under side of the leaf is fairly densely pilose on the midrib, veins and veinlets, and also on = pier side the veins are Lee aue In the latter character it sules and Belewhat narrower EN GREE glabrous above. It is possible, however, that the sterile specimens from Kagoba, gases referred by me to P. Purdomii, belong here. i Hl _ A picture showing the tall slender trunk of this tree is in the oh tion of Rock’s photographs. : Salix L.! ina a Determined by R. Gorrz “Sect. ae Dumortier Salix paraplesia Schneider in Sargent, Pl. Wilson. m1. 40 (1916). es Exstern Tier. Radja and Yellow River gorges, : with Spruces on northern slopes of valley south of Yellow River, _ “bpesite Radja, alt. 3200 m., no. 14111 7, June 10, 1926 (shrub to small tree 4.5~6 m.); snaatheeéat of Radja, alt. $200 m., no. 13977 ?, , 1926 (shrub 2.4-3 m. ); above Picea forest, southwest of sg alt. 3600 m., no. 13971, May 25, 1926 (shrub 1.2-1.5 m.); “Ppesite Radja, alt. 3500 m., no. 13955 <*, May 24, 1926 (shrub 'm., leaf ond oid aig ne ieay mountains of Kulu, alt. 3380 m., no. dune 1929. Shrub 1.5 « eluef difference of this st secs from S. pentandra seems to lie sh color of the under side of the leaves, which, however, conspicuous on young leaves and often nearly disappears, he plant from Szechuan. Other differences indicated by ‘rim his key (in Sargent, Pl. Wilson. m1. 74), as size of ': number of stamens and length of style are acuity less and therefore less important. ROPHYLLAE Schneider ha Schneider in Sargent, Pl. Wilson. mt. 118 (916). ter knew only the pistillate plant the description of mate catkins may be given here: Coaetanea, pedunculo brevi 2-5 mm. longo foliis parvis tous suffulto, ovata, 1-1.5 cm. longa, 0.8-1 cm. tebe during 1928 Piatt es iraareers etcrtehet 388 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm 4 densiflora; squamae late obovatae, semi-nigrae vel antice purpuras- centes, utrinque basi inprimis crispo-villosae, dorso glabrescentes; a nectaria 2, angusta, integra vel 2—pluries anguste lobata; stamina _ 2, filamentis liberis 14 pilosis, antheris ovatis aureis. Eastprn Tiser. Alpine region between Radja — and Jupar Range: alpine meadows of Wajo la, alt. 4200m, — no. 14154 o, June 1926 (shrub 60-90 cm.). Jupar Range: among rocks and alpine meadows, upper Jupar valley slopes, alt. 3600-3900 m., nos. 14294 7, 14295 2, 14296 7, 14297 9, 142989, 14299 o’, 14300 9, 14301 o’, June 1926 (shrub 0.60—1.20 m.); banks of upper Jupar stream, alt. 3600 m., nos. 14352 co”, 14353 9, June 1926 (shrub 1.20-1.80 m.); valley slopes of Kerab, southern slopes of Jupar range, no. 14410 <7, June 1926 (shrub 0.90—1.20 m.). CrntraL Kansu. Lien hoa shan: among rocks, alt. 3500 m. no. 12701 9, July 1925 (shrub 0.30—0.60 m.; leaves rich greet above, glaucous beneath, male); alpine regions among Rhododen- dron scrub, alt. 3300 m., no. 12726 #7, July 1925 (shrub 60-90 cm.; — leaves dark green beneath); summit of mountain, alt. 3800 m., n°. 13421 (fol.), Nov. 1925 (shrub 1.30 m. or less; semi-prostrate; buds . red). | SouTHwesterN Kansu. Tao River basin: back of Adjuan, on ridges with Birch, Rhododendron and Spruce, alt. 3200-3300 m., no. 12655 9, July 5, 1925 (shrub 0.90-1.50 m.); mountains of Adjiian-Toyiiku, Pakeshan, alt. 3000-3300 m., no. 13409 9, Nov. 1925 (shrub 1.20 m., forming dense bushes on limestone ridges: leaves pale yellowish beneath); mountains beyond Adjiian, Min- shan, alt. $450 m., no. 18414 (fol.), Nov. 1925 (shrub 0.90-1.20 ™-; branches thick, reddish; leaves pale, yellowish beneath). ‘ JUTHWESTERN AN (Muli): Mount Siga, northeast of Kulu, alt. 4600 m., no. 17883 ¢, Tune 1929 (shru ub 60-90 em.). WA Salix oritrepha var. tibetica Goerz, var. nov. Frutex humilis ad 1 m. altus, dense ramosus ramulis tenuioribus — brevibus nodosis, foliis julisque minutis. Eastern Tiper. Radja and Yellow River aioe rocky and grassy slopes above Picea forest, northern slopes. Yellow River, southwest of Radja, alt. 3600 m., no. 13969 <", 25, 1926, nos. 13970 9, 18972, May 1926 (shrubs 0.60-1.20 m.): alpine meadows between Howa & Arh’tsa canyon, north of alt. $450 m., no. 14040 #*, May 31, 1926 (shrub 00 em); north Radja, alpine slopes of Arh’tsa bluffs, alt. 3300 m., nos. 14053 o | 14054, May 31, 1925 Login 90 cm.). : Gombs, N Szecuuan (Muli): Mount Mitzuga, west of Muli 75 m., no. 16036 #, June 1928 (semi-prostrate shrub); M : t. 3050-48 (semi : Sis northeast ‘of "Kulu, alt. 4300 m,, no. 17874 9, June vend nei m.). & jg] _REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 389 ie lakes spec. nov. Ther um v. aseig 2 m. altus, ramosus ramulis brevibus nodosis, novellis breviter cinereo-tomentosis, postea glabrescentibus vel pilis brevibus obsitis en omnino sordide subatris; gemmae semiconicae, a ag josae. Folia in petiolo 1-3 mm. longo, juvenilia (matura desunt) mubtus pi ie pian acroscopicis adpressis albo-sericeis tecta, dein oe, a venis tantum pilosis mox glabra, atroviridia, eliptica v. ebovatoliptin, antice rotundata, nervis primariis 3-6, margine mT glan res EGP nticulato; stipulae 0. Amenta (3 tantum ) coaetanea in pedunculo mm. longo, bracteofoliis squamaceis v. opacis foerseens Szecuuan (Muli): Mount enag west of Muli Gomba, rerun no. 16041 ¢% (shrub 2 m.), no. 16079 <7, (shrub 4% m.), Tah giao aso Range, south ‘of "Tatsienlu, alt. 4600 m., no. Edd is near S. one cephe Schneid., but differs chiefly in ne tm the y young leaves. Salix sclerophylla Anders. fro eal I have seen material, a different, more woolly pubescence of fine serration of the leaves lacking in the two other oc ee in Bot. Gaz. ux1v, 137 (1917).—Handel-Mazzetti, Cheapo (Muli): Mount ee ry of Muli Gomba, Sra alt. 4000 1a 16035 9, aaey 1928 (shru 3m. ‘Tas oe of ad bo. 17964 #, no. 17966 9 . eK iio (shrubs 1-1.5 m.). of the young bnaves 3 is the same in all three eaten and stronaly revol, sath ; Ang MOVE SZECHUAN “a Mount Mitzuga, west of “v-4075 m., no. 16074 9, June 1928 (shru bom) t of Muli Gomba, in ie fatkins this hybrid closely perce S. tenella Schneid., which are vid yg aod revolute — the margin w when cmt o gow often to pubescent beneath. Salix luctuosa Lévi. and S. th ot fig Sat can be re doubt that hybrids Schneider in Sargent, Pl. Wilson. m1. 47 (1916).— ettt, Symb. Sin. vir. 7 (1929). Han Nsu. Tao River basin: ridges of Adjiian, alt. 3750 m., no. 12648 9, July 1925 4 o. 12649 o, July 5, 1925 (shrub 90 cm.). Plant also has a dorsal, though very small nectary re freely revolute and py thinly pabopeent 390 2" JOURNAL OF THE ARNOLD ARBORETUM Salix B zetti in Symb. Sin. viz. 74 (1929). SOUTHWESTERN SzEcHUAN (Muli): Mount Mitzuga, west of Muli Gomba, alt. 3050-4875 m., no. 16070 2, June 1928 (tree 5 m.) Salix plocotricha Schneider in Sargent, Pl. Wilson. 111. 49 (1916), ? ee spathulifolia Seemen in Bot. Jahrb. xxv1. Beibl. 82 P. ed 3 pl 905).—Léveillé in Bull. Soe. Bot. France, v1. 304 (1909).—Handel- : Mazzetti, Symb. Sin. viz. 73 (1929) SouTuwestmrn Kansu. Tao River basin: along streams, ‘ near Adjiian, eastern Minshan, alt. 2700 m., nos. 12640 9 , 126402 ?, July 1925 (shrub 1.20-2.40 m.; leaves glaucous beneath); near the head of Maerkhu valley, alt. 3000 m., no. 12960 9, July 25, ; 1925 (tree 4.50-5.40 m.; leaves dark green; petioles and young shoots red; catkins reddish); near Tebbu, Shimen to Drjakana, alt. 3150 m.,no. 13190 9, August 1925 (tree or shrub 1.80-3 m.; catkins long, greenish) ; Toyiiku valley, slopes of Pakeshan, Minshan range, alt. 3000 m., no. 13410 (fol.), Nov. 1925 (shrub 4.50 m., branches slender; leaves dull beneath); beyond Adjiian, in forests along stream, alt. 3000 m., no. 13412 (fol.), Nov. 1925 (shrub or small tree 4.50-6 m., branches straw-colored); along banks of stream below Mt. Kuang kei, Kadjaku, Minshan, alt. 2850 m., no. 13415 (fol.), Nov. 1925 (shrub 1.80-3 m., branches ascending). Upper Teb- bu country: along Kaichow stream not far from its source, south of the Minshan, alt. 3000 m., no. 12486 9, June 1995 (shrub 1.80—-2.40 m.; leaf whitish below). ENTRAL Kansu. Lien hoa shan: in Spruce forests and — outskirts, alt. 3000 m., no. 12729 9, July 1925 (shrub 3 m., leaves glaucous beneath). eae? I was unable to decide whether Seemen’s species is identical with S. plocotricha, since I could not examine the original. A close con- — nection exists doubtless between the three species of the section Eriostachyae enumerated above. Salir Ernesti is distinguished from the other species chiefly by the presence of a second gland in the pistillate flower, a character which is not supposed to be con- stant. The specimens cited above for S. Balfouriana and 8. Ernest are very similar in the development of the leaves which are strongly revolute at first, but differ much in their pubescence, while in re G plocotricha the young leaves are flat and floccose-tomentose. Agal! — : similar to this is 8. Delavayana Hand.-Mazz. which, however, has 4 glabrous ovary. 3 Sect. DenricuLatag Schneider _ Salix denticulata Andersson in Svensk. Akad. Hand. 1850, p. 481 e: [vox : alfouriana Schneider in Bot. Gaz. Lx1v. 137 (1917)—Handel-Ma- (1851).—Klotzsch & Garcke, Bot. Ergeb. Reise Prinz. Waldemar, 119, t. 89 (1862).—Schneider in Sargent, Pl. Wilson. ut. 17 lxest : : y seb REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 391 _ SourH WESTERN Kansu. Tao River basin: along lateral Be aap watershed, alt. 3000 m., no. 12237 o, June 1925 (shrub ~ 1.80-8 m.; flowers yellow); along streams south of Minshan below - Shimen, alt. 3000 m., no. 12484 <7, June 1925 (shrub 1.50-2.40 m.). _ Though this species originally described from the Himalayas has ES tt yet been recorded from China, I have no doubt in its correct ‘identification. The only difference would be the pubescence of the floral bracts which are described by Andersson as “obsolete puberu- ~ he,” while they are on the Kansu specimens rather evenly short- _ pubescent. The two specimens though collected in different lo- _ talities are very similar; in no. 12237 both nectaries are rather deeply Ds eet two narrow lobes. eo LonerrLorar Schneider Salix tibetica Goerz, spec. no 3 Frutex vel arbor ad 4.5 m. nite) ramulis novellis pubescentibus, anniculis omnino glabrescentibus, cortice gemmisque fuscis sub- - inde parum pruinosis. Folia in petiolo puberulo 4-6 mm. longo, : - hovella utrinque glabra, costa supra brevi-tomentosiuscula parce pilosa excepta, adulta ovata vel ovato-elliptica, aa (1 My ) 3-4.5 em. longa, basi rotundata, apice brevi vel triangu- Supra viridia, subtus glauca vel albida, margine sat dense - seatodentata; stipulae 0 vel minutae, ovatae, deciduae. Amenta - Maetanea, in pedunculo 2-6 mm. longo foliolis parvis 2-3 saepe - adueis i imstructo, 7 ovata, 1—-1.5 cm. longa, 0.8 cm. crassa, densi- ma, thachi albo-pilosa; squamae ovatae, semifuscae, utrinque ? > basi crispo-pilosae; nectarium singulum ovale; stamina 2, entis liberis 14 pilosis 4-5 mm. longis, antheris ovalibus aureis. hta fructifera crasse cylindrica, densiflora, ca. 3 cm. longa, 1 cm. a Saas Te 2 “) Thachi squamisque ut in <’; nectarium singulum oblongum ' glabrum 0.3-0.5 mm. longum subaequans; ca capsula glabra, i nm. longa, stylo ca. 1 mm. longo stigmatibus 0.5 mm. longis alls coronato, RN Trer. Jupar Range: Jupar streambed, alt. 3150 San o, June 1926 (tree 3.5-4.5 m.); Jupar valley along alt, t. 8150 m., no. 14284 9, June 1926 (shrub to tree er en ig ea A eee eRe rev ny RD eg oa Fy oe a A 7 sig Ss Waparica Goerz, : altus vel ku alta, ramulis bectbhi! powell tig vel subglabris, demum omnino glabris, anniculis $ vel fuscis, Folia recentissima plana, subtus ab initio 0 sy Renee dense pilosa cito glabrescentia, costa pu 392 JOURNAL OF THE ARNOLD ARBORETUM [vou xm ula excepta, adulta in petiolo puberulo 2-4 mm. longo, elliptica (ca. 1.2) vel apicalia sublanceolata (1:214-3) 3—4 cm. longa, utroque — plus minusve aequaliter angustata, supra atroviridia, subtus glauca _ vel albido-glauca, margine crenulato vel interdum subintegro; stipulae 0 vel minutae, caducae. Amenta coaetanea, in pedunculo 0.4-0.8 cm. longo foliolis parvis vel squamaceis 2-3 instructo, — o& cylindrica, ca. 2 cm. longa, 4-5 mm. crassa, densiflora, rhachi _ pilosa; squamae obovatae, pallidae, brevipilosae, ca. 1 mm. longae; nectarium singulum, oblongum; stamina filamentis 2 liberis pilo- sis ca. 3 mm. longis, antheris subglobosis aureis. Amenta 9 crasse cylindrica, ad 2.5 cm. longa, 9 mm. crassa, densiflora, frue- tifera ad 6 cm. longa, rhachi pilosa; squamae obovatae, semifuscae, puberulae, ca. 1 mm. longae; nectarium singulum oblongum, pedi- cellum 144 mm. longum plus minusve aequans; germen cinereo- pilosum ca. 5 mm. longum in stylum 0.5 mm. longum paulatim attenuatum, stigmatibus 0.5 mm. longis bilobatis plus minusve — conniventibus. Eastern Tiser. Jupar Range: Jupar streambed, alt. 3150 m., no. 14303 9°, June 1926 (tree 3.6-4.5 m.); Jupar valley along streambed, alt. 3150 m., no. 14283 9, June 1926 (tree 4.5 m.). Radja and Yellow River gorges: in streambed at Dachso canyon, north of Radja, alt. 3150 m., no. 14085 , June? 1926 (shrub 3—4.5 m.); northern slopes of river valley mountaims opposite Radja, alt. 3150 m., no. 14001 o*, May 27, 1926 (shrub | 2.40-3 m.). oe : ; The specific differences between S. juparica and S. tibetica are not yet quite clear. The pistillate specimen of S. tibetica differs" : the oval leaves and the glabrous capsules from the pistillate ae juparica. The staminate S. tibetica (no. 14304), however, also Te — sembles S. juparica (no. 14303), together with which it apparently had been collected. It is not impossible that the latter is a hyb Further material is needed to arrive at a clear understanding. 9? Salix hypoleuca Seemen var. kansuensis Goerz, var. nov. a \“"" Differt a typo foliis subtus pallide viridibus (non glaucis), ped¥™ : culis foliis normalibus ceteris paulo minoribus instructo. . SOUTHWESTERN Kansvu. Upper Tebbu coun try: south . of Minshan range, along streams, alt. 2900 m., no. 12440 <7", June — 1925 (shrub or small tree 4.5 m.; catkins yellow); south of Minsha® — range near Tongwa, along streams, alt. 2900 m., no. 12441 9, June — 1925 (shrub 1.80-2.40 m.); southern slopes of Minshan, along bank — of mountain stream, alt. 3200 m., no. 12500 9, June 1925 1.20-1.80 m.). ey 2 AND KOBUSKI, PLANTS COLLECTED BY ROCK 393 1S, sites omnino ahi brunneis saepe subnitenti- gemmae semiconicae, acutae, glabrae, nitidae. Folia juvenilia ibtus pilis longis acroscopicis sparse, supra pilis brevibus obsita, mox glabra, subtus pallida vel subalbida, elliptica, -margine minute denticulata, petiolo 1-2 mm. ; folia matura desunt; stipulae 0. Amenta coae- in lo brevissimo 2 vel 3 foliolis vulgo sat magnis cto, a Fs yNinarica, 2 em. longa, 1 cm. crassa, densiflora, . pilosa; squamae obovatae antice fuscatae, longe albo-bar- -hectarium singulum (vel dorsale minutum) rectangulum; filamentis liberis 14 pilosis. Amenta 9 crasse-cylindrica, om Sins, 8 mm. crassa, rhachi et squamis ut in o; squamae rman nectarium rectangulum pedicellum brevissimum poepste ot mm. longum, eres 04 mm. doticis oe minusve divisis divavieaiin Trser. Radjaand Yellow River gorges: | gomba, on rocky cliffs with Junipers, alt. 3300 m., nos. ‘ and 13929 o'", tag 20, 1926 (shrub 2. 40-3 m. )3 deny of I uns. rH 3450 m., no. 13957 °, May 24, 80 m.); same locality, alt. 3300 m., no. 13959 , 40— 304 JOURNAL OF THE ARNOLD ARBORETUM [vou xm minutum, ovale; stamina 2, filamentis liberis glabris ca. 2 mm. — longis squamas haud multo superantibus, antheris ovalibus aureis. Amenta 9 tantum statu maxime juvenili adsunt, 5 mm. longa, simili figura ut in o videntur; squamae glabrae, pallidae; necta- rium 1, germen sessile, breviconicum, glabrum, stylo stigmatibusque brevibus. SOUTHWESTERN Kansu. Tao River basin: mountains of Choni, in forests and banks of streams, alt. 3000 m., no. 12115 9, May 1925 (shrub 1.50—1.80 m.); mountains of Choni, alt. 2700 m., no. 12123 o, May 1925 (shrub 1.50-1.80 m.). _ Resembles in its habit S. tenella Schneid., but differs in the staminate flowers having only one nectary and in the leaves being glabrous from the beginning. Sect. Drptopicryax Schneider Salix Faxoniana Schneider in Bot. Gaz. uxiv. 143 (1917).—Handel-Maz- zetti in Synb. Si Sin. vir. 82 (1929 SOUTHWESTERN SZECHUAN (Muli): alpine meadows, Mount Mitsaes yt west of Muli Gomba, alt. 4000 m., no. 16054 , June 1928 (shrub 1 m.), moun- tains of Kulu, in gravelly datasibed. alt. 4150 m., no. 17951 o@, June 1929 (prostrate shrub). The plants agree well with the male type of the species (Rock no. 4473) deaetad by Handel-Mazzetti (1. c.). Sect. Berperiro.iar Schneider Salix flabellaris Andersson in Svensk. Vetensk. Akad. Handl. 1850, p. 497 (1851); in Jour. Linn. Soc. rv. 54 (1860); in De Can- dolle, Prodr. xvi. pt. 1. 295 (1868).—Schneider in Sargent, Pl. Wil- son. m1. 142 (1916). SouTHWEsTERN Kansu. Upper Tebbu country: foot of Shimen, alt. 3600 m., no. 13059 9°, July-August 1925 (prostrate shrub covering uldiess and grassy slopes; catkins green). UTHWESTERN SzEcHuAN: Mount Konka, Risonquemba, ba, Konkaling alt. 3690-5335 m., no. 16869 9, June-August 1928 928 (prostrate shrub). Salix flabellaris f. spathulata Andersson, l. c. SOUTHWESTERN Kansvu. Upper Tebbu country: foot ' of Shimen, alt. 3600 m., no. 13058 9, July-August 1925 (prostrate shrub growing perfectly flat, sa kaa boulders and grassy slopes; catkins red). Sect. LINDLEYANAE. began! 3 Salix Lindleyana W. ig (OER in a. Ve Handi 1850, P. pny (S51): ur. Linn. Soe. Iv. er : XV Dp 296 (1 868) .—Schneider in 1916). ( he der in Sargent, Wilson. 11. 62 veetet bereg her ey 1. Wilson. mt. 145 Salix Souliei Semen i in Fedde, Rep. Spee. Nov. m1. 23 (1906). —Sehnei- Wee herve 992] _REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 395 f é SoUTHWESTERN. ogg (Muli): Mount Mitzuga, west of Muli Gomba, ah * 3050-4875 m., no. @, June 1928 (prostrate shrub); Min ya Konka Es Baw Hane, south of TateioniG, alt. 4660 m., no. 17529 7, July 1929 (pros- trate shru i This spseaty very polymorphous species is best to be compared with _ the European S. retusa L. The delimitation of the species thes Ae by Schneider : under the amy Se etl is mostly rather uncertain and apparently _ further reductions will be necessary; shape and size of leaf, its margin and ‘ ons the hen surface are “yore very variable in the same species, as it is case in S. retus Sect. Guaucar Fries ___ Salix opsimantha Schneider in Sargent, Pl. Wilson. 111. 63 (1916). 3 UTHWESTERN SZECHUAN: Chiu-Lung-Hsien Territory, east of the Yalung River, alt. 3080 m., no. 16435 «, May 1929 (tree 3-5 m.; * fis. rich pink with a te plant agrees very well with Schneider’s description. It resembles S. $e. Fazoniana, as already stated by Handel-Mazzetti (Symb. Sin. vit. 82) but r sey distinguished by its tall habit, even tree-like in the specimen cited — X opsimantha Goerz, By hock no RN SzEc : Mou onka, , Risonquomba, Konkaling alt alt. 16833 g, meeny 1928 (shrub 3-4 - ar of Muli Gomba’ alt. 3650-4425 m., no. 16437 9, June 1928 ges 1] 25 : “(he ventral one twice as lo the dorsal one), entire or ihe The two latter seneies P indicate S. Ernesti as the ba ber milk is sti closer to sb since it has ‘ee one long gd Pp sitotsisns Goerz, spec. no S Thing 9.90-1.50 m. altus, ramulis bictlie crassis nodosis, no- iat is ab initio glabris, anniculis fuscis vel sordidis; gemmae semi- Folia ee, adpressae, acutatae, argute carinatae, glaberrimae, fuscae. (Conta in petiolo 3-6 mm. longo glabro, novella utrinque glaberrima “asi ‘sta inclusa), plana, adulta obovata 3-5 cm. longa, apice brevi, ek Totundata, nervis primariis 8-10, reticulo utrinque obsolete 98 Supra atroviridia, subadiposa, subtus albida, margine aD-— denticulato; stipulae 0. Amenta coaetanea, previpedunculata, 8 normalibus parvis suffulta, percrasse cylindrica, ad 3 X 1 (Fructu ad 5 em.), rhachi crassa parce pilosa; squamae magnace & bd mm. ), obovatae, fuseae, venulosae, utrinque brevissime : Pepe dorso glabrescentes; nectarium singulum, © oblongum, : longum; stamina 2, ca. 12 mm. longa, filame ntis liberis 12 antheris ovatis flavidis. Amenta 9 rhachi et squamis ut in "mis ex parte latioribus ($ mm.) antice saepe crenulatis; Singula vel bina, ventrale ampullaceum, ad 0.8 ™m- 9.5 mm. longum; pedicellus brevissimus nectartum | Prete conicum, 3-4 mm. longum, squama occul- Rh ‘i “i ad Dil 396 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm tatum, subrufo-tomentosum, in stylum fuscum plus minusve 15 mm. longum maxima ex parte fissum attenuatum, stigmatibus 0.6 mm. longis bipartitis laciniis filiformibus erecto-patentibus. SOUTHWESTERN Kansu. Tao River basin: Mt. Kuang ke shan, along mountain torrents, alt. 3600 m., no. 12362 9, June 1925 (shrub 1.2—1.5 m., catkins rich yellow); Minshan range, below pass of Mt. Kuang kei, alt. 3600 m., no. 12410 o, June 1925 (shrub 1.2 m. or less; flowers and catkins a rich yellow); Mt. Kuang Kei, trail to Djrakana, Minshan, among Rhododendron, alt. 3750 m., no. 13414 (fol.), Nov. 1925 (shrub 1.2-1.5 m.; branches stiff, thick; leaves pale beneath). CrentTRAL Kansu. Lien hoa shan: alpine regions; alt. 3300 m., no. 12725, July 1925 (shrub 1.5 m., leaves glaucous beneath). This species differs from S. spissa Anderss. found in the Tienshan and Altai Mountains chiefly in the large less pubescent floral bracts and the partly double nectary. It is to be expected that occasionally also in the staminate flower a dorsal nectary may occur. Also the very close serrature and the perfectly glabrous branches of leaves should be noted. Salix Delavayana Handel-Mazzetti in Symb. Sin. vm. 78 (1929). SOUTHWESTERN SzecHUAN (Muli): mountains south of Muli, Mount ie Gib- of Kulu; alt. = ee no. 17953 7965 no. a, Janel 1929 peek 0.60- y g, show be Spay the ie. fest leaves are just expa: h much shictateh S. spodiophylla Hand MARA. ‘Gp. cit. 7), nye is possibly a ihe: es 8S. “Delavayana as one of the pare: Sect. Caprear Dumortier Salix Wallichiana Andersson in Svensk. Vetensk. Akad. Handl. 1850, p. 477 (1851); in Jour. Linn. Soe. rv. 50 (1860); in Svensk. Vetensk. Akad. Handl. rv. 80, tab. 5, fig. 46 (Monog. Salic.) (1867); in De Candolle, Prodr. xvi. pt. 1. 223 (1868).—Schneider in Sargent, Pl. Wilson. i. 64 (1916).—Handel-Mazzetti, Symb. Sin. vu. 87 (1929). Sourarkn Kime: mountains of Pikou and Mosuping, alt. 1200 m., no. 12066 <, April 1925 (shrub 1.5-1.8 m.). ' SOUTHWESTERN Kansu. Tao River basin: anes Choni, west of Tao river, Picea forest, alt. 3000-3150 m., no. 121 Q, no. 12126 o, no. 12128 @, May 1925 (tree 4.50-5.40 m-); Choni, banks of mountain streams west of Taoho, alt. 2850 ™- ath 12135 2, May 1925 (shrub 1.5 m.); mountains west of Choni, 3000 m., no. 12136 <, May 1925 (shrub or small tree 4.5 m.); oe a banks of Choni river, near bridge, alt. 2520 m., no. 12805 Q, June — 1925 (tree 4.5-6 m., often ebiead Lae 1993] REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 397 Western Szecuuan: above Ching chuan mountains along stream, alt. ob pate 12045 9, no. 12047 o, no. 12048 &, April 1925 (tree or shrub m.). \\0 Salix pseudo-Wallichiana Goerz, spec. nov. Frutex excelsus vel arbor ad 6 m. alta, ramulis sat tenuibus no- vellis pubescentibus vel albo-villosulis, anniculis, subglabris brun- ntis vel fuscis; gemmae conicae, brunneae, glabrae. Folia in petiolo puberulo 4-6 mm. longo, utrinque pilis brevibus adpressis sparse obducta, plana, adulta glabra vel costa puberula, late ovato-ellip- tica, apice brevi (1:114-2), apicalia ovato- vel obovato-lanceolata, apice subproducto (ca. 1:3), 3-5 cm. longa, supra opaco-viridia, subtus glauca, nervis primariis 7-10, reticulo vix conspicuo subtus parim elevato, margine integro vel in foliis apicalibus irregulariter dentato, Stipulae 0 vel parvae, semicordatae. Amenta coaetanea vel eoaetanea in pedunculo brevissimo vel in ° ad 7 mm. longo, braeteis squamaceis eaducis suffulta, @ ovata, ca. 2 cm. longa, 1.5 Prec aag densiflora, rhachi pilosa; squamae anguste obovatae, ca. 3 x 18 mm., antice subatrae, longe albo-barbatae; nectarium in mm. longum, stamina filamentis 2 liberis 4 pilosis, SPE aly antheris ovalibus aureis. Amenta 9 crasse cylindrica, ine a Tague 1.5 em. crassa, densiflora; squamae et nectarium stg ay lus nectarium aequans vel triplo superans, pilosus; ave; of conicum, 4.5 mm. longum, stylo 0.4—0.6 mm. longo, x ta 0.6 mm. longis, divisis, laciniis tenuibus, patentibus. ‘ox capsulae post dehiscentiam parum recurvatae. "pe maa Grasslands between Labrang eS. sand w Rive r: in Serchen leading to Yellow River (shrub ea; via and Radja Gomba, no. 13917 &, May 15, 1926 sk ly flowers golden yellow). Jupar Range: of dupar valley, alt. 3150 m., no. 14287 9, June 1926 456m) Radja and Yellow Ri : es cna costae, ellow River gorges: eng ou of spruce forest in Dachso Canyon, Radja, iwi alt. 3150 m., no. 14077 2, June 2, 1926 (shrub 3 m.). ikins : and all parts of the flower are more graceful than in ; epee the filaments are pubescent, the pedicel 4 ay the style longer and the leaves broader. The much the type of S. livida Wahlbg. which, however, ‘s sestpat One could almost take these plants for {1 dllichiana with S. Rockii Goerz, but the former has oe Eastern Tibet by Rock or by any one else. allas, Fl. Ross. I. pt. m. 72, t. 81, fig. 3 (1788).— ent, Pl. Wilson. nr. 154 (1916). 398 JOURNAL OF THE ARNOLD ARBORETUM [vou. xan Eastern Tiper. Radja and Yellow River gorges: valley of Nyavruch north of Radja, lateral valley, along rocky streambed, no. 13932 ¢, May 27, 1926 (shrub 1.8-2.4 m.). SOUTHERN Kansvu: mountains of Motzuping and Pikou, along stream, no. 12071 #, April 1925 (shrub 1.2 m.). SOUTHWESTERN Kansvu. Tao River basin: Minshan range, ravines of Kadjaku, alt. 3150 m., no. 12379 o, June 1925 (shrub 2.4~3 m.): Minshan range, slopes of Kadjaku valley below Kuang ke, alt. 3000 m., no. 12420 o, June 1925 (shrub 3 m.; catkins yellow). CrenTrat Kansu. Lien hoa shan: swampy alpine meadows, alt. 2850-3000 m., no. 12685 9, July 14-20, 1925 (shrub 60-90 em., often 30 cm.; leaves silvery), no. 13420 (fol.), Nov. 1925 (shrub 1.2- 1.8 m.; leaves silvery tomentose on both sides). Saliz sibirica is new for China; the specimens, however, are not all typical. Very characteristic is the pistillate plant from central Kansu which could even be taken for the European S. repens L., if it had longer pedicels. On account of its short pedicels it belongs to var. subsessilis (Regel) Goerz (S. repens var. subsessilis Regel). The two numbers from southwestern Kansu recall forms from the Pamir which constitute a distinct species (S. schugnanica Goerz, adhuc ined.), but are not identical with it. Owing to the lack of _ fully grown leaves a definite conclusion cannot be reached. The filaments are glabrous in all numbers. Salix juparica x sibirica Goerz, hybr. nov. Eastern Tiper. Radja and Yellow River gorges: valley of Nyavruch north of Radja, lateral valley, along rocky streambed, no. 13933 #7, May 1926 (shrub 3-4.5 m. branching from base; flowers yellow). This hybrid was found at the locality cited with both parents. It resembles S. sibirica Pall., but differs in the somewhat slenderer catkins, the slightly pubescent filaments and in the scarcely pubes- cent unfolding leaves. Sect. Viminatzs Bluff & Fingerhuth Salix Rehderiana Schneider in Sargent, Pl. Wilson. ur. 66 (1916)- SouTHERN Kansu: mountains between Motzuping and Pikou, along streambeds, no 12070 9, April 1925 (shrub 1.5—1.8 m.). ‘ The catkins of the specimen are still very young, but according to the floral characters (pedicel and ovary glabrous; style long, stig- species. mas short, pedicel very short) it can hardly be referred to any other 933] _REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK = 399 _ Easrenn Tipper. Grasslands between Labrang and Yellow River: Gochen valley near mouth at Yellow _ River gorge south of Dzang lamassery, alt. 3060 m., (?) no. 18912 9, May 14, 1926 (shrub 1.6-2.4 m.). Radja and Yellow BS River gorges: rocky slopes back of Radja, alt. 3300 m., (?) 3 no. 13925 f, May 20, 1926 (shrub 1.6-2.4 m.); valley of Nyavruch __ north of Radja, lateral valley, along rocky streambeds, no. 13934 9°, i May 97, 1926 (shrub 2.4-3 m.). Jupar Range: Jupar stream- : bed, alt. 3150 m., nos. 14305 9, and 14306 co, June 1926 (tree _ $4.5 m.); along streambed in upper Jupar valley, alt. 3600 m., no. 14350 9, June 1926 (shrub 3-3.6 m.). SOUTHWESTERN Kansu. Upper Tebbu country: along banks of mountain streams, southern slopes of Minshan, alt. 3180 — M., no. 12501 9, June 1925 (shrub 1.8-2.4 m.). Tao River basin: forests of Shiaoku, among Larches, Birches, Spruce, alt. 3000 m., no. 12816 9, July 1925 (shrub 2.4-3.6 m.). Lower : Salix Rehileriana Schneid. var. brevisericea Schneider |. c. 67. . ati et \Nrstenill ot Fear oe (Cat Ney ol Bl ht 4y > hd ke _ ‘ B & ca = Ye ie = meal fe e is") a = - ow P i z 8 ay evidence of a relationship between var. brevisericea and S. ‘Mytillacea, as suggested by Schneider (I. c.). Nevertheless one of — ete forms, the one with glabrous or the one with pubescent ey turn out to be a hybrid. No. 13925 has its buds jt *bening and therefore its identification is not quite certain. “0 .. may possibly be a hybrid of S. Rehdertana and X S. taoensis ag as the thicker very precocious catkins suggest. : - Henrx Dumortier . : eheimsi Marshall von Bieberstein, ¥ 627 (1819) Schneider in Sargent, PI. Wilson. 1. 169. edtschenko, Consp. Fl. Turk. vr. $26. (1916)—Goer ics Kauk. 1. 10 (1930); in Fedde, haat Spec. Nov. gi.) (1930); Salic, Asiat. 1. 16, no. 17 (1931). oe. sy gustifolia Willdenow, Sp. Pl. rv. de) ( 805).—Stechéel6tw iP ull. Soe. Nat. Moscou, xxvut. pt. 1. 196 (1854).—Boissier, Fi. : 400 JOURNAL OF THE ARNOLD ARBORETUM {vou. xm Salix angustifolia var. eriocarpa Ledebour, Fl. Ross. m1. 604 (1850). Salix om were Boissier in Kotschy, Pl. Pers. austr. no. 621: ; Diagn, 1. 99 (1846).—Boissier et Buhse, Herb. Pers. 1. 201, sec. Andersson.— Trautvetter i in Act. Hort rt. Petrop. 1x. 176 (1884). Eastern Tirpet. Radja and Yellow River gorges: valley of Nyavruch, north of Radja, lateral valley, along streambed, nos. 13936 2 and 13937 o&, May 27, 1926 (shrub 1.20 m.); north of Radja, Yellow River valley, alt. 3150 m., nos. 14031 9 and 14032 o&, May 28, 1926 (shrub 1.20-1.80 m.). Ba valley: along stream, alt. 3000 m., no. 14262 92, June 1926. SouTHWESTERN Kansu. Tao River basin: Choni, banks of Tao river, alt. 2460 m., no. 12107 &, May 1925. The specimen from Tibet agrees with the type of the species, only 13936 has somewhat longer styles, possibly due to the influence of S. myrtillacea. The specimen from Kansu is not yet sufficiently developed but seems to belong also to the type. Salix cheilophila Schneider in Sargent, Pl. Wilson. 111. 69 (1916). —Handel- Mazzetti, Symb. Sin. vi. 87 (1929). UTHWESTERN SzecHuAN (Muli): Muli and Litang River valley, near Muli Gomba, alt. 2600-3000 m., no. fait 9, May 1928 (shrub or small tree). Salix myrtillacea Andersson in Jour. Linn. Soc. rv. 51 (1860).— Schneider in Sargent, Pl. Wilson. 111. 71 (1916).—Handel-Mazzett in Symb. Sin. vir. 89 (1929). Ped Salix subpycnostachya Burkill in Jour. Linn, Soc. xxvi. 532 (1899).— Léveillé in Bull. Soc. Bot. 6 Prine, 3 LVI. ore ( Salix squarrosa Schneider in Bot. Gaz. tx1v. 142 (1917). Eastern Tipet. Radja and Yellow River gorge® rocky slopes back of Radja, alt. 3300 m., no. 13927 ¢, May 20, 1927 (shrub 1.80-2.40 m.); rocky and grassy slopes above Picea forest, northern slopes Yellow River southwest of Radja, alt. 3600 m., no. 18967 , May 25, 1926 (shrub 0.60-1 m.); alpine regions south of river, opposite Radja, alt. 3600 m., no. 13997 9, May 27, 1926 (shrub 1-1.2 m.); valley of Nyavruch north of Badin alt. | ; 3300 m., no. 14039 &, May $1, 1926 (shrub 1.20-1.50 m.). SourHwesterN Kansu. Tao River basin: mountains west of Choni, alt. 3000 m., no. 12137 ¢*, May 1925 (shrub 60-90 cm.; flowers red as is the whole plant); Minshan range, south of Cheat on slopes and alpine meadows, alt. 3300 m., no. 12359 ee June 1925 (shrub 0.90—1.20 m.; catkins grey) ; ravines of a . alt. 3150 m., no. 12378 9, June 1925 (shrub 1.50—1.80 m.); along river, near Choni, no. 13402 (fol.), Nov. 1925 (shrub or small tre : 2.10-3 m.); valley of Maerhku, Minshan range, alt. 2700 m., 20- 13407 (fol.), Nov. 1995 (shrub 1.50-2.40 m.; much branched, leaves Hh white beneath), no. 13408 (fol.), Nov. 1925 (shrub 1.20-1. eee much branched; leaves eree bepeems). pens REE and iis] REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 401 meadows, Toyiiku, Minshan, Pakeshan, alt. 3450 m., no. 13411 ~ (fol.), Noy. 1925 (shrub 90 cm., compact). Upper Tebbu country: Djrakana, southwest of Minshan, outskirts of Abies forest, alt. 3000 m., no. 13417 (fol.), Nov. 1925 (shrub 1.80-2.40 m., bronze color; leaves papery, glaucous beneath), no. 13418 4 7 Noy. 1925 (shrub 1,50-1.80 m.; leaves glaucous beneath). th of Muli, Mount Gib- i id ot Be aREE} oe = ee z. =§ _ Mitzuga, west of M , no. _ (shrub 1-2 m.); Minya Konka Snow Range, south of Tsien, alt. 4600 m., e we 187 9, July 1929; Mount Sige, northeast of Kulu, alt. 4300-4450 m., oma ?, June 1929 (shrub 1.50 m .), no. 17889 # (shrub 0,901.20 m.). 3 specimens from Szechuan are generally more vigorous and oz che and thicker catkins than those from Kansu and Eastern ‘species, is seen also in Tibetan specimens (e. g. no. 13927). It is y only a monstrosity and may be due to injuries by grazing : “animals, j insects or climate. Some of the numbers cited show per- 5 ma traces of hybridization, particularly with 8S. Wilhelmsiana, ‘ b but it is hardly possible to give a more exact interpretation of these is _ fms lacking either leaves or flowers. _ Silirmyrtilacea x Wilhelmsiana (S. taoensis) Goer, hybr. nov. ees . Grasslands between Labrang Yello w River: Gochen valley near mouth at Yellow Wer, south of Dzangar lamassery, alt. 3060 m., nos. 13909 <", and wil 2, May 14, 1926 (shrub 3 m.); rocky gorge of Serchen leading llow River gorge, southeast of Radja, alt. 3120 m., no. 13915 w*y 14, 1926 (shrub 2.40-3 m.). Radja and Yellow : ver ©rges: rocky slopes back of Radja Gomba, alt. 3300 (13921 3, May 20, 1926 (shrub 3 m., branching from near Pits valley of Nyavruch north of Radja, lateral valley, along treambed, no. 13930 2, May 27, 1926 (shrub 1.80-2.40 m.); Radja, Yellow river valley, alt. $150 m., no. 14033 °, nae 3-3.6 m.). ERN Kansu. Tao River basin: forest of Ler he ks of Tao river, alt. 8200 m., May 1925, nos. 12103 5 m.), 12104 @ (shrub 1.2 m.) cd aioe ¢ | aber * Mountains of Choni, Picea forest outskirts, alt. 2700 3 9, May 1925 (shrub 1.8-2.4 m.); southern bank, 2265 9, June 1925 (shrub 1.5-1.8 m.). Minshan vines of Kadjaku, alt. 3150 m., no. 12377 9, June m5 catkins small cenit Choni, alt. 2550 m., 402 JOURNAL OF THE ARNOLD ARBORETUM [vou xm Nov. 1925, nos. 13401 (fol.) (shrub 1.5 m., much _ branching), 13403 (ram.) (shrub 1.5—2.4 m., branches red, branchlets pubescent), 13404 (fol.) (shrub 1.5-2.4 m., branches red, pubescent; leaves pubescent on both sides), 13405 (fol.) (shrub 1.2—1.5 m., branches dark red, leaves white beneath) and 13406 (fol.) (shrub 2.4-3 m., branches green); Kadjaku, along stream, alt. 2240 m., no. 13416 . (fol.), Nov. 1925 (shrub 1.5—1.8 m., branches stiff, blackish; leaves _ green beneath). Lower Tebbu country: along stream near Nyiba, alt. 2550 m., no. 14963 (fol.), Oct. 1926 (shrub 3-4. m.; leaves bluish grey beneath). This hybrid is in localities where the two parent species occur, apparently not only very frequent, but also exceedingly polymor- ous. On account of the incompleteness of the material it is not always possible to draw the lines between them and the parents, One may assume that the hybrid is fertile and that by recrossing with the parent the limits become indistinct. Thus e. g. no. 12105 is very near S. Wilhelmsiana and also no. 13911 with its slender Sek elk km i ee PE Le RO RO ee catkins may represent a recrossing with S. Wilhelmsiana, while others might be interpreted as forms of S. myrtillacea. Not rare seems to be a combination of the staminate catkins which resembles those of the European S. purpurea (e. g. nos. 12099 and 12103) or those of its Caucasian variety virescens Anders. (no. 12104). Like- wise the shape of the leaves changes from narrow-oblanceolate 0 — narrow obovate. No. 14963 has leaves resembling those of 8 — tenuijulis Ledeb. which ranges from the Tianshan to Armenia; leaves, however, differ from that species not only slightly in the : serration, owing to the influence of S. Wilhelmsiana, but also completely the stipules which are so characteristic for S. tenwijulls: — Besides the leaves are in October still furnished on both sides with — the long acroscopic hairs of S. Wilhelmsiana. No. 12265 represen's : a somewhat intermediate leaf shape, broad-lanceolate to agarey : late with short point; the serration extends here in contrast 0% — myrtillacea almost to the base and the pubescence is lacking entirely f2 on the mature leaves. ‘The hybrids show also in habit an interme’ — ate character. While S. Wilhelmsiana generally attains only Me height of 1.6 m., S. myrtillacea grows into a shrub up to 4 in. tall i Among the hybrids shrubs up to 3 m. are found. Salix myrtillacea < Rockii Goerz, hybr. nov. Eastern Trsetr. Radja and Yellow River 8 ont ; alt. 3150 m., no. 18958 @, May 25, 1926 (shrub 2.4-3 m-); Spruces on northern slopes of valley south of Yellow River, °PI Radja, alt. 3150 m.,no. 14112 9 ,June 10,1926. Jupar Range? 92), REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK = 408 _ along streambed in upper Jupar valley, alt. 3600 m., no. 14349 9, 1986 (shrub 1.8-2.4 m.). ‘ In the male plant the filaments are mostly connate and besides _ diten villous up-to the apex. Nos. 14112 and 14349 have well developed leaves which are about intermediate between the parents; 3 they are densely denticulate almost to the base. In no. 13112 the -_ longer, and somewhat spreading stigmas indicate the in- fluence of §. Rockit. In no. 14349 perhaps S. Wilhelmsiana is in- ; _ wlyed besides S. Rockii, as suggested by the comparatively small -_fniting catkins. Ov. Szecnuan (Muli): Mount Mitzuga, west of Muli Gomba, no. 16038 <7, June 1928 (shrub 2 m.). resembles S. Delavayana but differs chiefly in the lack of the ; _ Sect. INCERTA 0 Satx attredi Goerz, spec. nov. a Fritex vel arbor ad 4.5 m. alta, ramulis tenuibus, recentissimis : te pilosis citissime glabris, anniculis vetustioribusque brunneis _ M'tuscis, opacis vel rarius nitidulis; gemmae semiconicae, obtusae, Eats » brunneae. Folia in petiolo piloso 3-5 mm. longo, recentis- atts ana, utrinque laxe sericea (subtus pilis longis laxis acroscop!- i. Supra pilis brevioribus), postea utrinque glabrescentia (costa Hem Ls videtur, persistentius puberula), ovato-lanceolata, ca. ; 2%, basi rotundata, apice plus minusve triangulari, supra opace 7 aac subtus pallide vel caesio-glauca, integra; stipulae 0. Amen- ay tantum nota) coaetanea, subsessilia, bracteis minutis caducis 6 Se ‘ 4 rm a, anguste cylindrica, gracilia, 3-4 cm. longa, 2-3 mm. Postea ad 5.5 em. longa, 6 mm. crassa, laxiflora, thachi albo- ree miae obovatae, ca. 0.7 X 1 mm., subbrunneae, utrinque oe breviter albo- vel subrufo-hirsutae; nectarium veneer (0.3-0.4 mm. longum), oblongum, dorsale; pedicellus 8 (brevior quam nectarium) vel nullus; germen ™! ¥s . longum), conicum, acutatum, breviter albo-pilosum, nullo, stigmatibus 0.2 mm. longis plus minusve keuaye el suberectis; capsula ad 2 mm. longa, parcius , iloss. TERN Kansu. Tao River basin: mountains +24 m.), 12149 9 (shrub 2.4-8 m., outskirts of Picea ~Pper Tebbu country: in limestone gorges €? Tebbu land, southern slopes of Minshan, alt. 3300 m., no- “une 1925 (shrub or small bushy compatt tree 4.5 m.). | minutum — of Taoho, alt. 3000 m., May 1925, nos. ‘12147 (fol.) 404 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Nearest to this graceful Willow in habit is perhaps S. hylonoma Schneid., which differs in the denticulate leaves, the dense-flowered catkins, the long narrow nectary, the parted style and, as it seems, also in the larger size of all its parts. The young catkins of 8. Alfredi are also similar in appearance to those of S. heterochroma Seemen, but also these are dense-flowered and have entirely different styles and stigmas and besides longer pedicels. Owing to the lack of staminate flowers, it does not seem possible to refer this new species to any of the sections. The species is named in honor of Alfred Rehder, Curator of the Herbarium of the Arnold Arboretum. Page 38 POLYGONACEAE Polygonum Auberti L. Henry.—Add the following number: SoutHwEsTERN Kansu. Tao River basin: along stream — between Taochow and Kan-ku, alt. 2450 m., no. 13210, Aug. 1925 (huge climber forming dense masses; flowers cream-colored). Page 39 RANUNCULACEAE Clematis aethusifolia Turcz.—Add the following no.: Kansv. Lien hoa shan:alt. 2750 m., no. 12753, — CENTRAL July 14-20, 1925 (climber; flowers yellow, sepals with cream margins). Page 41. Clematis brevicaudata DC.—Add the following no-: Centrat Kansu. Lien hoa shan: along bank and ov? | shrubs, alt. 2750 m., no. 12750, July 14-20, 1925 (climber, flowers | cream-colored). Page 42. following no.: ematis glauca var. akebioides f. phaeantha nasi Centra Kansv. Lien hoa shan: beyond Yitao, ede slopes, alt. 1500 m., no. ‘aces Aug. 1925 (climber; flowers P brown). following no.: SouTHWESTERN KaANsvu. Tin River basin: around alt. 2600 m., no. 12917, July 1925 (forming large etna : bushes, common; flowers brownish ates Clematis tangutica var. . obtusiuscula Rehder & Wilson. nett is] REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 405 aay SAXIFRAGACEAE Page 51. Hydrangea Bretschneideri Dippel.—Add the following no.: _ SourawesterN Kansu. Tao River b asin: below Choni, along shady banks of Tao River, alt. 2500 m., no. 12228, June 1925 __ (flowers white). Page 72. Rubus idaeus var. strigosus Maxim.—Add the following no.: Eastern Trper. Radja and Yellow River gorges: spruce forests, northern slopes of River valley; south of Radja, no. | ROSACEAE : | 13986, May 25, 1926 (flowers white). F Boe ie: LEGUMINOSAE Page 84, let bi ir multijugum Maxim.—Add the following nos.: ex Tier. Radja and Yellow River gorges: _— Brassy » eee banks of river near Radja, alt. 3000 m., no. 14002, % May 27, 1926 (flowers purple). Ba valley: on loess bluffs, alt. - $000 m., no. 14244, June 1926; on loess bluffs and banks of lateral ravines of valley, alt. 2850 m., no. 14363, July 1926. ERICACEAE - Phe, trufum Batal.—Add the following no.: Sourrwesrern Kansu. Tao River basin: Maerkhu 7 mo, north of Minshan, Abies and Picea forest, alt. 3200-3350 m., . 18675, eae 1925 (in fruit; flowers pink). VERBENACEAE i s ‘incana Miquel. . . ay -Mastacanthus ‘Schauer. Ti ileal Maximowicz in Bull. Acad. Sci. St. Pétersb. (1881); in Mél. Biol. x1. 301 (1881). —P’ei in Mem. Sci. +I. no. 3, p. 172 (Verbenac. China) (1932). : collected by Purdom (no. 792) in the Taochow Dis- mgs here. The two numbers of the Rock collection — P’ei as nos. 12709 and 12755 are mistakes for no. 12765. CAPRIFOLIACEAE — Tophylla Willdenow.—Add no. 14070 to the last per ra pertt cf eevee 3350 m.” 406 JOURNAL OF THE ARNOLD ARBORETUM (vous COMPOSITAE Determined by J. Marrre.p ; Aster incisus Fischer in Mém. Soc. Nat. Moscou, 11. 76 (1812)— Hemsley in Jour. Linn. Soc. Bot. xxut. 412 (1888). ] SOUTHWESTERN Kansu. Lower Tebbu count ry:gray- elly rocky banks of Peshwekiang, alt. 2000 m., no. 14556, Aug. 30, | 1926 (shrub 60-90 cm.; flowers white). Aster Limprichtii Diels in Fedde, Rep. Spec. Nov. Beih. xu 508° (Bot. Reis. Hochgeb. Chin. Ost-Tib.) (1922). SOUTHWESTERN Kansu. Lower Tebbu country: Oak forest near Nyipa village, Mayaku, alt. 2300 m., no. 14808, Sept. 9, 1926 (shrub 60-90 cm.; flowers white); dry shale slopes under Oaks at Nyipa in Mayaku, alt. 2400 m., no. 15055, Sept.—Oct. 1926 (shrub 60-90 em.; flowers pinkish white). Aster poliothamnus Diels in Fedde, Rep. Spec. Nov. Beih. xi 503 (Bot. Reis. Hochgeb. Chin. Ost-Tib.) (1922). Eastern Tiger. Radja and Yellow River gorges: among conglomerate boulders in valley near Radja lamassery, alt. ti 3000 m., no. 14201, June 1926 (flowers lavender). y A distinct and handsome species with numerous lavender-colored : flower-heads. is Microglossa salicifolia Diels in Bot. Jahrb. xxrx. 612 (1900). SOUTHWESTERN Kansu. Lower Tebbu country:ga™ elly rocky banks of Peshwekiang, alt. 2050 m., no. 14555, Aug. 30 | 1926 (shrub about 1 m.; flowers white); banks of stream, P eshwe- | : kiang gorge, alt. 2050 m., no. 14800, Sept. 5, 1926 (forming large ss clumps; flowers white). a2 Tanacetum Rockii Mattfeld, spec. nov. eS Frutex parce ramosus; rami vetustiores lignescentes, cortice “aes tenuiter rimoso obtecti, hornotini herbacei, simplices sursum tal tum ramulosi, supra basin foliis delapsis nudi, deinde laxe foliati (internodiis 1-3.5 cm. longis), costati, tenuiter appresse pilosi, .. 3 orsum glabrescentes, sursum griseo-tomentelli. Folia sursum M4 nitudine sensim decrescentia petiolata; petioli tomentelli, 2-5 ™™ _ longi, basi utrinque lobulo herbaceo stipuliformi ornati; laminae ambitu fere rotundatae, 1-2 em. latae, 1-1.7 cm. longae, a truncatae et deinde breviter in petiolum angustatae, marge " a ad tertiam partem sinuatae, supra virides, dense granuloso-pur® tatae, subtus dense appresse incano-tomentosae, palminervi#® Be nervis subtus leviter prominulis e basi 3 rarius 5, venam unica? ie Z - saepius tantum emittentibus, exeuntibus in lobos, lobi 3, inaequy "1982] REHDER AND KOBUSKI, PLANTS COLLECTED BY ROCK 407 i formes, apicali iterum trilobulato, lobulo medio late ovato usque _ subquadrato, interdum utrinque 1-dentato, ca. 3-6 mm. longo, 5-7 mm. basi lato, lateralibus oblique ovatis, lobis 2 lateralibus grosse _ sinuato-tri-crenatis, crenis deorsum magnitudine decrescentibus, lobulis omnibus obtusiusculis breviter mucronulatis. Inflores- : centiae paniculato-corymbosae, axes inferiores ex axillis foliorum _ orientes bene foliati, 20-6 cm. longi, superiores axi primario brevius longiusve adnati, bracteis foliaceis integris oblanceolatis sparse obsiti, 5-2 em. longi, omnes corymbulo parvo glomeruliformi, 1-2.5 em, diametiente, bracteolis parvis lineari-lanceolatis usque fili- formibus obsito terminati, pedunculis propriis tomentellis 0.5-3 mm. longis. Capitula minuta, heterogama; involucri globosi, apice con- _ Sitieti, ca, 2-2.5 mm. diametientes-et alti; squamae 3—4-seriatae, _ @labrae, exteriores ovatae, 1.2-2.2 mm. longae, 0.8-1.2 mm. latae, subobtusae, interiores late ovato-rotundatae, 2-2.5 mm. longae, _ 182mm, latae, latissime scariosae; receptaculum convexum, Vix _ 1mm, diam., nudum, glabrum; flores omnes fertiles, marginales ~ feminei uniseriati, tubulosi, lobis 5 ovato-lanceolatis 0.5 mm. longis inelusis 2 mm. longi; corolla sparse glandulis megacephalis sessilibus obsita ceterum glabra; flores disci tubulosi hermaphroditi, sparse glandulosi, 2.2 mm. longi, tubus 1 mm. longus, subito paullumque ampliatus j in limbum lobis 5 ovatis 0.5 mm. longis inclusis 1.2 mm. 3 Styli rami 0.5 mm. longi, apice truncato barbulati; germen ; em ae obovoideum, glabrum, -vix striatulum, (coctum) FO ee hr ee Dee Peer Cima Bl aN ee ie) Ca a a Lower Tebbu country: banks ‘y Chulungapu near Wantsang, 1980 m., no. 15097, Sept. 1926 (plant 700 cin. ; flowers yellow). 4 Speciem generis hacum foliorum forma camparandam fratra quaesivi, nisi habes folia ceterum longe alieni Chrysanthema “nt aliquo modo pro similia; notabilia etiam achaenia ecostata, ee siniatala, mucilaginosa ita iis Artemisiarum similia. m salicifolium Mattfeld, spec. nov. e basi parce ramosus, 30-40 cm. altus; rami vetustiores pa nudo, lignescentes, cortice griseo, tenuiter rimoso te oliorum fasciculo rosulante, anno sequente ramum fer- 4 malem evolvente terminati; innovationes € ramis vetusti- m defoliatis) hornotini steriles, breves, pool int longi, laxe sursum dense fasciculatim foliati; ;, inflorescentia terminati, gis), fusces- Folia hee simplicia, 408 JOURNAL OF THE ARNOLD ARBORETUM » [vou xa anguste lineari-lanceolata, (2—) 5-7 cm. longa, 2-5 (—10) mm. lata, basin versus sensim longeque petiolatim angustata vel fere rite alato-petiolata et deinde basi dilatata ramo affixa, apicem versus tardius angustata, obtusiuscula et mucronata usque breviter acu- minata, margine integerrimo saepius minute revoluta, supra viridia, primum laxe arachnoidea demum glabrescentia, subtus incano- tomentella, nervo tenui percursa. Inflorescentia corymbosa, semi- — globosa, saepe glomerata, 3-4 (6) cm. diametiens et aequealta; pedunculi communes dense tomentosi, 3—6-capitulati, 1-1.5 (+4) em. longi, pedunculi proprii 1-3 mm. longi; capitula heterogama pluriflora; involucri demiglobosi, 4-6 mm. diametientes; squamae ca. 4-seriatae, Hare herbaceae, laxe arachnoideae, late fusco- vel atrofusco tae, exteriores late ovatae acutae, interiores late obovatae, apice e lacerato late rotundatae obtusissimae; receptaculum nudum, glabrum, ca. 1.5 mm. diam., convexum; flores omnes fertiles, marginales feminei uniseriati; corolla anguste cylindrica, sursum attenuata, vix 3 mm. longa, sub lobis 4-5 ca. 0.6-0.8 mm. longis constricta, glandulis sessilibus obsita; flores disci hermaphroditi cylindrico-tubulosi, sursum paullum ampliati, 3.5 mm. longi, glandulis sessilibus obsiti, lobi 5 ca. 0.5 mm. longi; styli rami breves, apice truncato barbati; germen epapposum, an angulato- obovoideum, glabrum, eglandulosum, costatum, florum ma ium dorso applanatum. CENTRAL Kansu: Lien HG a shan: among limestone rocks on summit, alt. 3450 m., no. 12693 (typus) July 1925 (shrub Mecha CHUAN: Dongrergo chhange mit Spiraea, P Potentilla, , Gebiis: Juniperus, 4000-4200 m., Harry Smith, no. 3529, Aug. 8, 1922; same loeality sonnige Blockhinge bei Huang-lung-ssu, « 4000-4150 m., Harry Smith, 0 3620, July 22, 1922. RN SZECHUAN (Muli Kingdom): mountains between UTHWESTE een Wa- “os Dje and Muli Gomba, in open meadows, 4350 m., no. 16918, Aug. 8, 1 (height 60-180 cm. , flowers yellow jn Species nova foliis integris anguste lanceolatis inter sie species generis eximia. Specimen no. 16918 foliis amplis ad 10 cm. longis id (ig et 1 cm. latis longius acuminatis, inflorescentia paullum altiore, statura elatiore a typo differt. Tanacetum falcatol Petrop. tv. 7 (1923). CENTRAL Kansu. arr ers Wisc basin: rocky gorge gorge of = Hsining beyond Hsiang tang, alt. 2125 m., no. 13246, Aug. 1929 (woody plant, 0.3-0.6 m.; flowers Eastern TIBET. Radja ae Cline River fe 8 1; among rocks, banks of river, alt. 3050 1 m., no. 14193, June ee (shrub 0.6 m.; flowers” bhanty oloba “en H. Krashenikov i in Not. Syst. Herb. | : AND KOBUSKI, PLANTS COLLECTED BY ROCK = 409 Oliver in Hooker’s Icon. xxut. t. 2214 (1892). Kansu. Upper Tebbu country: Spruce Drjakana, overlooking Yiwaku valley, alt. 3200 m., ), Aug. 3, 1925 (shrub 2.5-3 m., with scandent branches; k); among Picea and Abies trees, forest of Drjakana, 14588, Aug. 1926 (shrub 3 m.; flower-heads slender, ome irbamne sna rocky banks: ks: ae 2450 . 1926 (shrub; flower-heads single). Rehder in Jour. Arnold Arb. x. 135 (1929). | Nsu: en route to Lin-hoa-shan from Choni via ‘scrub, no. 12667 (type), July 1925 (shrub Phi; > red) 410 JOURNAL OF THE ARNOLD ARBORETUM — CYTOLOGICAL STUDIES OF CORNUS Hate DerMEN With plate 58 THERE ARE some forty or more species of Cornus, of which some thirty forms are in cultivation at the Arnold Arboretum. This genus has quite a wide distribution all through the northern hemisphere, with one species in the tropical mountains of Africa (Wangerin 1910). According to Berry (1923) over fifty fossil forms have been described, the oldest of these coming from the Upper Cretaceous period; the majority of other forms have been found all through the Tertiary formations. Out of fifty existing forms described by Wangerin (1910) and Rehder (1927) twenty-four are found in central and eastern Asia, two in western Asia, one in western Asia and Europe, ten in Atlantic North America, six in Pacifie North America, three in Central America, one in Alaska and middle western United States, one m Africa and two in the boreal and arctic circumpolar region. Al! but the last two mentioned are woody while the latter forms are perennial herbs. cytological preparations is described in an earlier publication (Dermen 1931). e at A table is given below showing the basis of the taxonomic group- ing of the genus Cornus based on Rehder’s classification (1927), but with slight alterations, and the corresponding chromosome grouping of the species investigated. _ TABLE I. a A, Flowers in ¢ mes or panicles itk tan ii volucre. f ‘21 : . B. Leaves alternate................... ee ea (10 pairs of chromosomes) Cornus alternifolia, C. controversa. Sa : eae BB. Leaves opposite........ = thle eee me eee dS (11 pairs of chromosomes) As alba Rosenthalii, C. _ femmes flaviramea, C. stolonif. era coloradensis, C’. rugosa, C. Amomum, » C. arnoldiana, C. asperifolia, C glabrata, C. racemosa, C. paucinervis, C. coreana, C. sanguinea, C. Breisch- 1932] DERMEN, CYTOLOGICAL STUDIES OF CORNUS 411 AA, ers AA. F in dense umbels with an involucre. B. Flowers yellow with a yellowish involucre not exceeding the flowers and deciduous during anthesis. . . . . gee (9 pairs of chromosomes) Cornus mas, C. officinalis. : BB. Flowers greenish yellow with large white or pink bracts. i oe ee a ee eee eee WIERTS bed oe aie ee been ee (11 pairs of chromosomes) : ornus ida, C. kousa chinensis. ee cae tevpmeeods plant... 0... 2... ee ee eee (22 pairs of chromosomes) Cornus canadensis. From the above table it is seen that chromosome numbers corre- spond to the system of taxonomic grouping. This investigation _ showed four groups of species with basic chromosome numbers 9, 10,11, 22. In Plate 53 are illustrated meiotic and mitotic figures a8 representative of cytological groups. Meiotic chromosomes of €. mas (fig. 1) are considerably larger than the meiotic chromo- somes of C. florida (fig. 2) and much larger than the chromosomes j of C. canadensis (fig. 3). These size differences are also noticeable mn the somatic chromosomes (C'. mas, fig. 4; C. florida, fig. 5). Both _ Meiotic and somatic chromosomes of C. officinalis correspond in oh na and structure to the C. mas chromosomes. Cornus kousa Bi chinensis (fig. 6) was similar to C. florida in all respects; both had eee pair of chromosomes with minute trabants. There were no : trabants observed in any other species. The chromosomes of the i C. alba group were somewhat smaller than the chromosomes of F the C. mas and C. florida groups. In this, as well as in the C. ie, oP A pair of chromosomes were noticeable that were = _nspicuously longer than the others, with double constrictions : (C. stolonifera, fig. 7; C. stolonifera flaviramea, fig. 8). Most of the 2 other chromosomes had subterminal constrictions (C. paucinervis, oe ). Cornus controversa (fig. 10) and C. alternifolia have 20 (C. candidissima) the chromosome count 2 = 8-9 a glabrata n= 11-12. Both these species have 11 pairs 0 mosomes, 1) quite small in size, but showing at least two pairs © ‘Roticeably longer than the others. Several efforts were 412 JOURNAL OFJHE ARNOLD ARBORETUM [vou. xm made to study Helwingia japonica and Davidia involuerata (the only species of these genera available in the Arboretum) but the chromosomes were found too crowded together, thus making an accurate count difficult. The chromosome number for Davidia was estimated 40+ and for Helwingia 80+. Both had small chromosomes like Nyssa. The root-tip cells of Cornus, Nyssa and Davidia were of the same nature; half of them were thick walled showing the presence of a gummy substance. Helwingia did not possess these gum cells. 3 Cornus mas (fig. 4) and C. officinalis have 9 pairs of chromosomes. The somatic chromosomes of these forms indicate the presence of 2 pairs with median or submedian constrictions. These 3 pairs are longer than the other chromosomes of the group. When other forms are compared with these, one finds that all the chromosomes are quite short with only one long pair and with a double constric- tion. From these facts it may be concluded that the two median constricted pairs were segmented, giving rise to the four extra chromosomes of the 11 pair group. The chromosomes of the 10 pair group showed two pairs that were considerably longer than the others. In this case, apparently only one of the nine pairs was segmented to give rise to this additional pair. The 22 pairs of chromosomes of C. canadensis undoubtedly are from the dupli- cation of the 11 pairs of some species like C. florida or some her- baceous diploid form. To shed some light on the origin of this species with a tetraploid number of chromosomes, the author intends to study C. suecica, another herbaceous species, and some other varieties of (. canadensis. From the chromosome counts and structure it is suggested that nine pairs may be taken as the basic number of the genus and that other forms with 10, 11, and 22 pairs are merely alterations of this basic number due to segmentation of some chromosomes é duplication in the case of C. canadensis. There is evidence ™ supporting the hypothesis that fragmentation of chromosomes may give rise to new forms. Stern (1928) gives a case describe by Seiler that clearly demonstrates this point. There were found — two races of butterflies, Phragmatobia fuliginosa, one with 28 and the other with 29 pairs of chromosomes, Seiler finds that the long chromosomes of the 28 pair race are four units long, while " the 29 pair race the long chromosomes are three units long. _ When these races are he finds that long chromosomes pair wit? the fragmented chromosomes. Anderson (1931) has made # comparative study of the chromosomes of the genera Allium and Nothoscordum. The genus Nothoscord scordum is considered closely : 5 ; “1989 DERMEN, CYTOLOGICAL STUDIES OF CORNUS 413 : elated to the genus Allium. In A. stellatum he finds seven pairs of chromosomes, while the characteristic number for the genus jseight pairs. In N. bivalve are found nine pairs. In the nine : chromosomes of the microspore, seven were with median or sub- median constrictions and two with terminal constrictions. In the words of the author, ‘‘These latter are conspicuously marked by t large, deep-staining insertion points. The chromosomes, like those of Allium, are large and ribbon-like. The attachment constrictions in Allium are usually median or sub-median (or at most sub- terminal). It seems quite possible that Nothoscordum may have been derived from an eight-chromosomed parental stock by the division of one of the large median-constricted chromosomes. This is further borne out by the fact. that the combined length of | the two chromosomes with terminal constrictions is only a very little greater than that of the longest chromosome with a median _ Sonstriction.” Thus it may be assumed that when a long chromo- _ Some is fragmented from the point of so-called “ spindle fiber attach- ment point,” then the derived chromosomes build up anew their | own spindle fiber attachment constrictions. ae Tt may be safe to assume that a species with a small number of | chromosomes is the most primitive of its genus. Taking the ee eytological findings based on number, size and structure of chromo- _ Somes, it may be said that a type like (. mas is the most primitive et SAL, we ‘etal ee aa ee ee oe ee | Un SS te e of the genus aad others are derivatives of this, both in respect to cytological characteristics and morphology of inflorescence. fe: ! ? fortunately, due chiefly to difficulties of cultivation, the Arboretum does not have some of the species like C. Volkensii, C. iS cilicia, C. Nuttallii, etc. that could have helped to make this study _ ‘More complete. These species mentioned are of special interest fe in: +a their. geographical distribution and because, in some of their close resemblance to other forms, like C. Nuttallii latter ip i , the former growing in Pacific United States and the Set tea tlantic United States. ected with Europe ia was the land bridge between Europe and Asia. aat C. mas has the least number of chromosomes and in icated above, and that there are many varieties 414 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm is in the middle point of the early geographical formation of the earth, indicate that forms with higher number of chromosomes may have been derived from C. mas and these forms spread to the left through Europe and America and to the right through Asia and down to the present African tropical mountains. Two species, C. alternifolia and C. controversa (alternate leafed), with two additional chromosomes are very similar. The former has moved . to the west occupying Atlantic North America and the latter moved to the present tropical regions of the East Indies, China, Korea and Japan. Other forms with four additional chromosomes must have originated from (. mas, changing morphologically in some respects but retaining the opposite position of leaves. Other species with 11 pairs of chromosomes are most likely derivative ‘orms of a species like (. florida or may very well be direct descen- dants of C. mas, . It was stated above that Nyssa sylvatica had 44 chromosomes ably longer than others. If this observation is correct, then this genus may be considered a derivative form of a Cornus species with 11 pairs of chromosomes. In connection with the chromosome study of Cornus, pollen grains of 25 species were measured and the percentage of pollen grain abnormality was det rmined. : . ‘TABLE Ir. Per Cent of Pollen Measurement _ Name of Species Grain Abnormality in & C. alba Rosenthalii...... 15 215 €. : See yee id 50 260 C. Laat PE WET RESP CELE Ree Q 215 C. ifera flaviramea....... 2 215 C. stolonifera coloradensis. é 260 C. TUNG eet gee 30 var C. Slavkast® 025 OY a 15 or C. Astontwms.52) 2G es ae 30 280 C. dubia® 5 eta eee 65 = C. obliques.) cca ise 4 250 C. arnoldiana? 2 5 40a eae - 80 190 C. aspera, 2 ooo te 25 ane C. Dunbarit® 0 15 215 C. glabrate.. 2... pig EF ae ee3 a C. racemene oo. be, Pe ae 50 180 C. COTEGNG. . on ek ee 10 215 C. Bretschneideri. . 2 2. | 205 CO. florida... i. ea Ea ee con ae 170 C. florida rubra’, Fe 95 160 Cc. Bin Sa PUL a ge ee 70 135 C. kousa chinensis... .......:..,... | 3 125 CL MGS... ie ea ee 3 He C. mas flaws o.oo a ee 3 ard C. fing. pe aa. 110 C. canadenaia. 5 10 ed ee ee ee en ee eee oe ET i eg) ge eed ee et 2 1983), DERMEN, CYTOLOGICAL STUDIES OF CORNUS 415 In the above table are given the species that were studied and recorded. Species with asterisks are hybrids. These are given in Rehder’s Manual of Cultivated Trees and Shrubs, 1927. The Bile «eae. kt measurements in microns are 180-280 for C. alba group, 125-170 for (. florida group, 110 for C. mas group, and 110 for C. canadensis. This record shows most strikingly that chromosome number and size do not control the size of pollen grains, while on the other hand, each group has its characteristic measurement. It was also found that C. mas, C. florida, C. kousa, C. officinalis, and their varieties develop their pollen grains in the fall, while others develop theirs in the spring. In this respect there was found an affinity between the C. mas and C florida groups. Cornus canadensis, while related to C. florida, develops its pollen grains in the spring. This difference may be due to its being an herbaceous form. From all the above facts it is suggested that C. mas or a similar species is the most primitive type; that the C. alba group may be considered as a derivative group from the C. florida type or directly from the C. mas type; and that the C. florida and C. alternifolia groups are parallel derivatives from the C. mas type. Cornus canadensis should be considered a derivative from the C. florida type. At present nothing can be said concerning Davidia, only that the root-tip cells were similar to Cornus and Nyssa, and that cs the chromosome number is estimated to be 40 +. . r ee a eS eee ne) Oe a ee oe Ee eee ee ee er Pee VO ts | ee of Allium stellatum * t > — Anpenson, E (1931). Th | Bon. ‘Ses oF e Chromosome complements and Nothoscordum bivalv (Annals of the Missouri Botanical Garden, an North Ameri d th ‘cc ld. (Rhodora, 33: 0 erica an e ‘orld. odo: Pe ees t, A. (1927). Manual of cultivated trees and shrubs. The Mac- ot = _ mn Co _ New ork, A ae *8). Fortschritte der Chromosomen. Theorie der Vererbung. P oe 9SICAL Laboratory, ARNOLD ARBORETUM, 2 Unrversiry. a a: ich, IV-229.) (1917). The chromosomes, their numbers and general importance- berg, 3 ; , SS Sa a wie oe Fi ese ie \ obs 4 “ 416 JOURNAL OF THE ARNOLD ARBORETUM ——fvonam DESCRIPTION OF PLATE 53 Fig. 1. C. mas. Second metaphase plate showing at one pole 9 chromo- somes Fig. 2. ye florida. First aed phe plate Slee n= 1 chromosomes. Fig. 3. C. canadensis. First metaphase pla Fig. 4. C. mas. Metaphase plate from sip packs with 2n = 18 chromosomes Fig. 5. C, : Metagniass plate from root-tip section with 2n = 2 chromosomes. Fig. 6. anes pram: Metaphase plate from root-tip section with 22 chromosomes. Fig: ° 7. °C. Gilgen r Motatanes plate from root-tip section with 2n = 22 chromosomes Fig. 8. C. stolonifera flavir ramea, Pigepiions plate from root-tip section with 2n = 22 chromosom Fig. 9. C. paucinervis. Metaphase he from root-tip section with 2n = 22 chromosomes. Fig. 10. C. ‘pe hoxmipa Metaphase plate from root-tip section with 2n = 20 chrom Fig. 11. Nyssa ipbvadicg. Metaphase plate from root-tip section with 2 eihb choies ba Jovr. ArNotp Ars. Vou. XIII. PLATE 53 & x) i yo" AWS “Nfs AR aN ) 9 = 8 _Vs wa WAYS wey) Anis (A i ELE CYTOLOGICAL STUDIES OF CorNUS ee : 1982] PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK 417 LEAVES FROM A COLLECTOR’S NOTE BOOK Ernest J. PALMER With one text figure Gainesville, Fla., April 6, 1931. THE UPLANDS about Ocala, where we began collecting today, are covered for the most part with a deposit of fine sand overlying the soft porous limestone, which latter deposit is quarried exten- sively here for road building. ‘These sandy uplands support a mixed growth of Pine (Pinus echinata and P. caribaea) and of deciduous species, amongst which are Carya alba, Quercus lauri- folia, Q. cinerea, Q. rubra, Q. Chapmanii, Q. stellata var. Mar- garetta, Diospyros virginiana, Crataegus constans, C. amica, C; _— inopina, Rubus cuneifolius, Prunus umbellata, Rhus quercifolia, Ceanothus microphyllus, Asimina speciosa, and Viburnum rufidulum. In dryer places this gives place to a more stunted growth, largely of shrubby species, locally known as scrub, in which Quercus Catesbaei, Q. myrtifolia, and Q. cinerea are often common, with . Xolisma ferruginea, X. lucida, Osmanthus americana and some- times Pinus clausa. Many interesting herbaceous plants grow here also, of which Lupinus diffusus is one of the most common and conspi eee s.r ee Fla 2 ee _ Most of the Crataegus of the open upland woods, including the m ecies mentioned above, belongs to the very distinct Flavae i soup, and the trees have a strange appearance with their generally = ba Pe rved, branches, slender drooping branchlets, and trunks . GNered with thick black bark, that is deeply fissured and divided : the : a horizontally, having much the appearance of that of | gy Owering Dogwood. . ee out in the forenoon to the Ocala National Forest, : PhLy) Ol aiaaee the lowlands and flood plains along the Oklawaha “Wer, which we crossed near Silver Springs, it was interesting to en change in the character of the forest. Cabbage growing amongst the Pines, just before we reached the the river valley is stiff and black and largely calcareous. — appear and give place to Tazodium distichum, Quercus floridana, U. alata, Celtis laevigata, Crataegus viridis — ‘Terinus profunda var. Ashei,! with Rhus Toxicodendron type we Profunda var. Ashei, var. nov. eee 4 Pema tfoliolis glabris vel'raro secus nervos parce tomentosis sepalis minori- ider : 10-1 ie ; 53 i aa 4 i beak ough Sp Para Mca cities seonaas bre sn patling, L5-85. 418 JOURNAL OF THE ARNOLD ARBORETUM and Sabal glabra amongst the undergrowth. Some of the dwart — Palmettos here were nearly two meters tall. A few trees of Quercus ctprins were seen growing just above the swampy flood plains, he afternoon we stopped just north of the boundary of © ‘aks County, near the village of Micanopy, to examine the Crataegus, of which several species were in bloom. sides Cratae- [vou. xm gus egens and (. Brittonii, another species was abundant, which | from the characters of the flowers and leaves apparently belongs — to the Parvifoliae or Uniflorae group, but differs in habit from most of them, as it becomes a small spiny tree 3—4 m. tall. The flowers are single or two or three together and have 20 stamens, yellow anthers and 5 styles, as in C. uniflora, and also have the large conspicuously eee Tapeh calyx lobes of that group. dm. lo: usually 7, ovate-lanceolate or lance-elliptic, cuneate or rarely rounded at base, redo at apex, usually entire, petiolules of the — leaflets 5-15 mm. long, those of the terminal leaflet 2-5 cm. long, glabrous t the: in which the mature leaves are glabrous, and other mediate, and for this reason I thin it better to regard the glabrous or sparsely Pu = i ie F. ; : rathe I Sri pecie: was first called to this tree by Mr. W. W ‘Ashe, for whom the variety is named, and on whose notes and collections I have en to supplement my 0% made this season in in the region where it grows. Maryla and: Dorchester Co., J. A. Cope, i= pease eastern shore ; Pot an, Sept. 927. Virginia: Alexander Island, Alexandria W. Ashe, nos. 1, 6, and specimens appear to be inter 1 Co., W. i , i: Sept. 18, 1924; near Alexandria, Jos. H. Painter, no. 912, Aug. 9, 1904. eet Carolina: On Brogaw River, near Northeast C Fear River, Pender Co. Ashe, Oct. $0, 1928; Pender Co., E. J. Palmer, no, 38256 et : ; Sept. 1 : af River, Alachua Co., ater, W. W. Ashe, June 4, 1892; near Hirt — W. W. Ashe (type) May 18 1 1929; Between Ft. White and Bildreath, W. W. 40 June 4, 1929; Oklawaha River swamps, Marion County, W. W. Ashe, April 26, oe: between Ocala and Sulphur Marion County, W. W. Ashe, ee Lg Palner, ver Junction, without date, W. W. Ashe; Campbell, B. F. Bush, nos. 436, "Au Harvey’s Canal, New Orleans, R. i 5 1982] PALMER, LEAVES FROM A COLLECTOR’S NOTE BOOK 419 Tallahassee, Fla., April 7th. Our first stop this morning was a few miles south of Gainesville, _ where Crataegus was abundant and several species were in bloom, growing in deep fine sands. The species collected in the thickets and along the border of open woods here were C. integra, C. impar, (. fortis, C. adusta and C. gregalis ?, as well as the Parvifoliae species, _ like the one we saw at Micanopy yesterday. Here also it becomes small pyramidal tree, 4-5 m. tall, with intricate spiny branches. _ Near Bronson Xolisma lucida was growing in a small swamp | filed with the Pond Cypress, and some of the specimens were $m. tall. _ In sandy upland woods near Chiefland, Levy County, Crataegus _ amica and C’. egens were in bloom, and in burned over ground there _ Were large patches of Castanea alnifolia, both sterile and fruiting, tone of them more than 3-5 dm. tall. About noon we crossed the Suwannee River at Old Town and _ Stopped on the west side for lunch and to work the collections. Betula nigra is growing on the river banks here. The leaves ap- _ Pear to be small and unusually thick for the species, but it can _ Sateely be more than a geographical form. Viburnum obovatum | as also collected here in young fruit, but with a few clusters of ae stillremaining. It is a small tree here, the largest specimens % being 67 m, tall. Quercus lyrata, Q. stellata var. Margaretta, Acer _'™’rum var. Drummondii and the curious little Cycad, Zamia . § foridana, were also collected, and I photographed a fine clump of villosus in sandy soil along the river bank. cen qer ae April 10th. fis morning, accompanied by Dr. Harper and a local ornitholo- ot measured, must have been 9 or 10 meters. A little ; dant in low wet ’ is abundant in the low wet woods. This tree related to Ulmus americana, a ae ae ae cae ie 420 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm pearance on account of the conspicuously buttressed bases of the trunks and its small leaves. Some of the trees had a curious ap- pearance from the markings caused by woodpeckers, which for some reason that I am unable to explain encircled the trunks in bands at rather regular intervals, giving them a sort of jointed appearance. We had lunch here at the picnic grounds and after- wards explored the woods, collecting Juniperus lucayana and Vi- burnum scabrellum, and in shallow muddy ponds the little Quillwort, Isoetes flaccida. Some of the native trees, as well as the Pecans cultivated about the town, were well loaded with great clusters of Mistletoe (Phoradendron flavescens), and I secured a photograph of one of these. A Crataegus of the Crus-galli group was abundant in the opel parts of the swampy woods, and I examined many of them here as well as others later in the day about Wakulla and St. Marks. ‘The type of Sargent’s Crataegus limnophila! came from St. Marks, and it is described as having flowers on slightly villous corymbs and 15-20 anthers of dark rose-color. Some of the trees observed here quite agree with the description in these and other respects, but others growing with them, quite identical in habit, foliage, bark fruit and other characters, have the flowering corymbs in some cases quite glabrous, and there appeared to be a complete gradation . from these to others in which the branchlets and petioles as well as the corymbs and hypanthiums are copiously villous. The flowers of all the trees examined here have 15-20, mostly 20 stamens and red anthers, but specimens collected a few days later at Chattahor bs chee have only 10 stamens, and slightly villous corymbs. PY or paring these specimens with the glabrous species described from Florida by Beadle as Crataegus pyracanthoides' said to have 7 stamens and red anthers, it seems evident that they are both as of one species, as there is a very clear identity in the foliage: flowers * and fruit and all other characters except pubescence and variation in the number of stamens. It seems clear, therefore that set should be regarded as only varieties of one species, and the gla sbre = form with corymbs and sometimes foliage and branchlets ar : less villous becomes C.. pyracanthoides var. limnophila (Sarg-)" nov. ; This also shows how impracticable it is to maintain the grou? distinction made by Beadle between Crus-galli and Berberifol™ based merely on pubescence. 1 Jour. Arnold Arb. ur. 3 (1922). — ie eee eae 2 Biltmore Bot. Studies, 1.136 (1902), eke q ios 1992) PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK 421 About the old town of St. Marks, lower down the river, both the _ pubescent and the typical varieties of this Crus-galli species are even more abundant, and in addition I collected here Crataegus integra, C. amica, and C. assimilis. The Honey Locust (Gleditsia tm ) is growing in the swampy woods here and some of the frees examined have unusual bark for this genus, that on the old trunks being dark, thick, and very rough, with fissures and ridges - tivided into short blocks, much as in Cornus florida or Diospyros _ tirginiana, I photographed the trunk of one tree with this sort of bark, and I think that no one familiar with the typical appearance _ of this tree farther north would recognize it. However, I can find no differences in the fruit, flowers or foliage to justify regarding it as Bs Ee April 11th. we The Apalachicola River, which crosses the western part of Florida s 's formed by the junction of the Chattahoochee and Flint Rivers, _ just north of the state line, in Georgia. In this part of Florida it -_ hascut its channel deeply through beds of soft Tertiary limestone, developing in places distinct bluffs and ecological and soil condi- = tions quite different from those of other parts of the state, which . ee for the very interesting and distinct flora. We drove out . today with Dr. R. M. Harper and a party of geologists from the _ State Geological Survey, our objective being Allen Bluff in Liberty . This is one of the highest bluffs of the coastal plain and it ethaps the highest in Florida. The precipitous part of the cliff = extends for perhaps half a mile along the river, is formed of strata of limestone and marl. In places it is quite per- but on the slopes there is an abundant growth of trees . ‘The soil on the highest levels above the bluff is of deep ravines cutting through this have penetrated the calcare- % and these support a rich growth of both woody and her- ‘cous plants. Several small trees of the Tumion (Torreya taxi- seen here, and amongst other woody plants were Hama- ls sum, besides several sorts of Crataegus, amongst © C. integra, C. condigna and C. armentalis, besides a small. 499 JOURNAL OF THE ARNOLD ARBORETUM [von On the return trip our car separated from the rest of the party and took a more northerly route, where following a weatherbeaten and misleading sign, we came out into a piece of deeply rutted sandy country road at places almost impassable and where some of dilapidated wooden bridges across the streams looked quite dan- gerous. Before venturing across two or three of the worst of these we stopped to reinforce them by laying loose planks lengthwise for running boards. In the deep sands by the roadside after getting over the worst of this we stopped to collect a little shrub that was in full bloom and very showy with its profusion of violet or purplish flowers. This proved to be Conradina canescens, of the Mint family, — or a related form somewhat intermediate between that species and — C. puberula. April 12th. This morning we drove out several miles north of Tallahassee, on the invitation of Mr. Goode, manager of the Horseshoe Planta- tion, to look at a Crataegus tree there. The tree standing neat the manager’s house is a large symmetrical specimen, about }? meters tall at a rough estimate. The low conical crown 1s formed of numerous slender wide-spreading branches, and slightly 28 zag branchlets, unarmed or with a very few slender spines. Most of the flowers were gone, but I secured a few belated ones in which the anthers were still unopened. There was an abundance of last season’s fruit under the tree. From a study of this material and other specimens in the herbarium of the Arnold Arboretum — _ it appears to belong to the Ignavae group of Beadle, which closely related to if not a part of the large Flavae pated And it seems to be quite distinct from any described species 4 A description of this species is given below che the name — Crataegus leonensis,' sp. nov. As contrasted with Crataegus igna™ 1 Crataegus leonensis, sp. Pe Arbor ad 10-12 m. alta; truncus interdum 8-4 dm. diam., cortice profunde nme fusco-cint vel fere nigro; ramuli gr Osi, sparse i in riers: is armati Folia ovata, rhombica vel obovata, 2-4.5 cm. longa, 2-4 Ns ngum um at 0 uw fe lorum ovata vel sub-rotunda ad 6 em. . lataque, basi obtusa, rotunda ve , dorso sulcat A tree 10-18 meters with depressed round t 5 aid wide-opreedia ading intscst® i wp £55 do deere i with thick, ridgy, dark gray nearly black bark. Hrasehlets ieahe zig-2ag, s armed \ sapere . 3-5 em. long. Leaves cre — for a few short nese along midrib me 92) +PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK —428 it is a larger tree, much less spiny, with larger, thinner leaves which are more coarsely serrate on the margins and often broader ; bracts numerous, linear it subgl 9-12 se’ utlets £ I surface and with prominent nd western Florida, growing in sandy open upland woods. 7 95, 30, Plantation, near Re fed ent, _ 16, 1919, nos. 5645, 5646, April 6, 1920, - os (type), April 3, 1923, nos. 6181, 6182, Sept. 27, 12, 1931. . 2° SO, cd A Sai rity dee 424 - JOURNAL OF THE ARNOLD ARBORETUM [vou. xu similar to those of some of the Rotundifoliae species. The rough dark bark and conspicuously glandular petioles, as well as its geographical range, indicate, however, that it properly iy: with the southern Flavae group. Mr. Goode later drove us about the plantation, and I saw several other specimens of this proposed species growing in the sandy open upland woods. Other species of Crataegus also found here were C. adinca, C. assimilis and C. consanguinea. In the afternoon Mr. McDougall and I drove out to Wakulla Springs, where I was surprised to find the Washington Thorn (Crataegus Phaenopyrum) growing in the low woods. This species has not previously been known as far south as Florida, so far as! know. In the sandy woods, on higher ground in this vicinity we also collected Crataegus abstrusa, C. clara and C. consanguinet, and two or three other as yet unidentified species. De Funiak Springs, Fla., April 13th. Leaving Tallahassee this morning, after taking leave of - friends at the State Geological Survey, where we have been our headquarters, and accompanied by Dr. R. M. Harper, who plans to accompany us as far as Birmingham, Alabama, we turned west and made our first stop at Chattahoochee, just south of the Georgia boundary The geological hvniition here is a rather pure Tertiary limestone, which forms bluffs and steep hillsides along the river and deep wooded ravines leading down from the uplands. The rich lime- stone soil and humus accumulated along the slopes and in the valleys supports a rich and varied flora, in places quite suggestive of iui woods; but with a mingling of southern plants. The beautiful and rare Tumion (Torreya taxifolia) attains . best development here. This remarkable tree, a relic of the distribution it had in an earlier geological period, is repre resented it ve America by this species, confined to a narrow belt in Fi Florida and . Georgia, and another (Torreya californica) on the Pacific coast In the Old World there are four species in eastern Asia. Te here were growing on steep hillsides and in the deep ravines. a largest specimen seen was perhaps 10 metres tall, with numerous spreading and ascending branches dividing about a meter ae : the ground. The trunk below ie she branches had a diameter” — 3.or 4 decimeters. Along the river bluffs cee ravines were also growing such familie northern species as Quercus alba, Ulmus fulva, Celtis laevigat™» te Nyssa sylvatica, and the low ee Viburnum affine va H | 1932] +=PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK 425 pomolacum. The Columbine (Aquilegia canadensis) grows on diffs and rocky ledges, and Phlox pilosa is abundant in the open woods. The most interesting discovery was the northern Prickly Ash (Zanthorylum americanum), which I do not think has been reported from Florida before. Acer leucoderme is also abundant on the bluffs and Crataegus spathulata and C. pyracanthoides var. limnophila were found in the low woods along the river. ee} Elba, Ala., April 14th. We left the camp at De Funiak Springs early this morning and drove out to the south and east, making our first stop at a crossing _ also an abundance of Mountain Laurel (Kalmia latifolia). This isthe second locality in Florida in which I have seen both of these ® plants growing together, the other being along the Ocklocknee River in Gadsden County. Cornus florida was in bloom in the low woods, the flowers being the largest that I have ever seen, some of them having a breadth across the bracts of more than 12 em. (4 inches). | ; A little farther south we got into a poor looking wooded country i any hills and ridges dissected by deep ravines. We stopped to iY eaplore one of these where a spring flowed out producing a rank & iceidend of ferns and flowering plants, with a great variety of trees me ns shrubs on the banks and along the little stream. Amongst iH Pssigs ‘were the Christmas Fern (Polystichum acrostichoides), _ *eech Fern (Thelypteris hexagonoptera) and Cinnamon Fern (Os- “= . a cinnamomea). The Beech (Fagus grandifolia var. caroliniana), he Nene: pyramidata, Tilia floridana, and Amelanchier canadensis, rowing on the ridges and banks, and farther down I collected uum densiflorum, Symplocos tinctoria, Stuartia malacodendron, it grandifolia and Cornus alternifolia. ‘Teturning to camp and loading up our paraphernalia, west to Crestview, where we turned south and were ° and thick, with the ridges curiously cross-fissured, into short blocks, as is common in this group. Some 426 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm of the trees attain a size of 8-9 meters, with trunk diameter of 2 dm. or more. 3 We were in search of a rare Oak described from this locality a few years ago by Mr. Ashe, as Quercus caput-rivuli, and later referred by him to Quercus arkansana var. caput-rivuli. We stopped for lunch at an abandoned ranger station in a clearing of the Pine woods. The pretty little Phlox Hentzii and Lithospermum Gmelini were growing in the sands here. After finishing lunch and working the collections I set out to explore some ravines just back of the house. The fine sand here is underlaid by harder loam and clay, eroding rapidly into these deep ravines with steep or precipitous sides, that work their way back into the uplands. In one of these I soon came upon several speci- mens of the Oak I was looking for, some of them having acorns. The trees growing here, as well as some seen later in the day, some miles to the north, have straight ascending branches, forming slender pyramidal crowns, and with smooth pale bark. This gives them a rather different appearance from Quercus arkansana as It grows in Arkansas and Alabama, and although there is little in the foliage or fruit to distinguish them, it is perhaps best to regard the Florida trees as a variety. : Along these ravines I also collected Illicium floridanum, Clino- podium coccineum and Prunus alabamensis. The last has not, so far as I know, been found in Florida before, and has only been know? from limited areas in Alabama and Georgia. Between Niceville and Laurel Hill, not far from the Alabama line, I again saw Quercus arkansana var. caput-rivuli, growing © similar situations to the other station and of quite similar habit. — Crataegus lacrimata was also abundant here, and the little legume Lupinus Westianus, was growing in the sandy woods. Along @ little creek north of Laurel Hill and just before we crossed the state line, — Cliftonia monophylla was abundant and in full bloom. Here 3 I collected Ilex decidua and a curious form of Liriodendron Tulipe fera, the small leaves of which have short, rounded lobes. yee s Troy, Ala., April 15th. Our first stop this morning was along a creek near Elba, where We stayed last night, and on the banks of the stream we found the i orange-flowered Azalea (Rhododendron austrinum) in bloom. Ths — seems to be a northward extension of the known range for ex species, which has only been known previously from a few localities — in Florida, I believe. The Mountain Laurel and shrubby PawP®™ (Asimina parviflora) were also growing here. Three miles north oF be e 199] PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK ~ 427 New Brockton we collected Crataegus senta, growing in open sandy woods. This was a good-sized tree, perhaps 7 meters tall, with the thick rough bark and recurved branches, characteristic of the Flanae group. Crataegus calva, growing near, has a somewhat similar habit, but was here a smaller tree, and it has larger, more _ showy flowers. After passing through Enterprise, and near the west edge of Dale County, we stopped to collect in sandy open woods, and here were found Crataegus lacrimata, C. lenis, C. attrita and C. gilva. Quercus Catesbaei and Q. stellata var. Margaretta were also abundant here, and alow Opuntia was growing in the sand and nearby, along a little creek we found Cornus alternifolia and Rhododendron canescens. Near Ozark we added Crataegus incilis, and C. opima as well as _ Bumelia lanuginosa to our collection, and near Brundidge Crataegus segnis and C. uniflora. About noon we reached the “pocosin,” a locality that I had for many years been anxious to visit. After turning off the highway we found ourselves on a poor country road, and began plowing our way through the deeply rutted soft sand, which threatened to stall our heavily loaded car. After some maneuvering and a good deal of _ Pushing we managed to get on and stopped for lunch at a primitive _ little school house on the edge of the woods. ie, The pocosin embraces an area of perhaps a hundred acres in the ae uplands, and is traversed by several ravines and small a a which have their sources here. Although the surface deposit are nearly pure sand looks quite similar to that of the surround- 3 rigs which originally were occupied largely by Pine forest, it ee distinguished and its boundaries clearly defined by its rich _ —), Peculiar flora of small deciduous trees and shrubs, many of 8 this are only found here in this part of the state. The cause of __,'S marked difference in flora has not been explained, but es a me ~ SSIS Suggests itself that it may be due to a difference In sub- : a the water content of the porous sand to be retaine —e “Man in the surrounding areas, and giving rise to s€ as aie = ae , ee ed a short, interesting account of the locality and a of the plants in 1914.! ae make it was not Quercus laurifolia ng along the ravines. Dr. Harper, who was with us and acting ~ shrubs, but the — Quercus stellata var. araneosa 428 JOURNAL OF THE ARNOLD ARBORETUM [vow. xm Quercus velutina Crataegus bisulcata Quercus Catesbaei Crataegus macilenta Quercus marilandica Rhus canadensis Quercus cinerea ! Aesculus discolor Quercus rubra Prunus umbellata Quercus arkansana Acer floridanum Rhododendron canescens Tilia floridana Crataegus Sargentii Osmanthus americana Crataegus contrita Viburnum rufidulum Quercus arkansana, which was collected here by Dr. Mohr many years ago, but which was not recognized as a distinct species until found by Bush on Red River, in southwestern Arkansas, and described by Sargent in 1911! was one of the particular objects of our quest. It is not at all rare here, although not so abundant as m the Arkansas station. I was impressed with the absolute and un- mistakable identity of the species in these two widely separated localities as well as the close similarity of soil and ecological con- ditions under which they grow. There can be no reasonable doubt as to the distinctness of this species, which is one of rarest and most interesting of the American Oaks. Selma, Ala., April 16th. We passed through a section of hilly country in Wilcox County, this forenoon, where Magnolia macrophylla is quite abundant open woods along the streams. In Marengo County we crossed section where a Cretaceous limestone comes to the surface, and the influence of the calcareous soil was plainly seen in the change of flora. Pines were absent here and Juniperus virginiana was the only Conifer on the uplands. Quercus Muhlenbergii and Q. Durandii are characteristic trees in the glades and on the hills. Crataegi Ashei was in bloom in the glades. As it grows here it is a Very spiny shrub about 3-4 meters tall, with large, handsome flowers: Along the bluffs of Pine Barren Creek, a little farther on Dr. Harpet showed us Dirca palustris growing along a north-facing. bluff. © Near Berlin, Dallas County, we crossed another limestone are* . and again found Crataegus Ashei as well as C. concinna and Rhamnus lanceolata in the glades and thickets. Crataegus insidiosa and ©. frugiferens as well as an unknown species of the Pruinosae grouP were growing along small streams. __ | | | Birmingham, Ala., April 18th, Dr. H. A. Wheeler, Curator of the Birmingham Museum, ¥™ — has been our host here, guided us out today to the mt 1 Trees & Shrubs, m. 121 (1911).—See also Palmer, E. J. in Jour. Arnold Arb. % as 1982] PALMER, LEAVES FROM A COLLECTOR’S NOTE BOOK 429 ity on Shades Mountain, several miles from Birmingham. [had visited this place with him previously, but was glad of an opportunity to explore it more fully. There is a scenic road which we followed up the mountain along a line of sandstone cliffs. Pinus virginiana, Quercus marilandica, Castanea dentata, Celtis pumila var. georgiana, Amelanchier cana- densis and Crataegus regalis were amongst the trees noted here. There are also many interesting herbaceous plants which we did not have time to collect. Silene Wherryi and Senecio plattensis were conspicuous in the rocky woods. Several ferns are found along the cliffs, one of the most interesting of which is Asplenium pinnatifidum, growing in shaded clefts, although it did not appear tobe abundant. Besides the Pennsylvanian sandstone, which is found at the lower levels, igneous rocks occur in many places. At one point where we stopped and spent some time in exploring and collecting, a little stream, Lost Creek, flows across an outcrop of granite and other igneous rocks, which form glades or barrens ! several acres in extent, and have a distinct and most interesting flora. Over the more exposed parts the vegetation is sparse and consists of mosses, lichens and herbaceous plants, with shrubs and stunted trees that have established themselves in clefts and broken places as well as more abundantly along the rocky margins of the stream. The rare and pretty little Dimorpha cymosa grows in large : — on thin soil of shallow depressions in the granite. Cheilan- & ‘Nes lanosa, Arenaria brevifolia and Coreopsis crassifolia also grow i ln indantly in the open places. Amongst the woody plants found EB ytnes barrens and along the creek are Quercus Boyntoni, Q. geor- «ana, Celtis pumila var. georgiana, Xanthorrhiza apiifolia, Phila- _ “Sphus hirsutus, Amelanchier canadensis, Aronia arbutifolia, nus alabamensis, Malus bracteata, Crataegus ignava, C. venusia, irifoliata var. pubescens, Acer rubrum var. tridens, Oxyden- arboreum, Vaccinium sericeum,! V. vacillans, V. tenellum, idron arborescens, R. canescens and R. alabamensis. | cus georgiana, which is locally abundant, has not before wn from Alabama, and only from the type ocality at untain, Georgia, and vicinity. It is usually a stout shrub 2-4 meters tall as it grows here, but in better © edge of open woods it sometimes becomes a small tree meters tall, and with a trunk covered with dark, ridgy, nelan : ) ig pee Raw Mohr in Contrib. U. S. Nat. Herb. v1- B (PI. Life Ala.) (1901). | um sericeum C. B. Robinson in Torr. Bull. xxrv. 570 (1912). 430 JOURNAL OF THE ARNOLD ARBORETUM [vou xm bark. Boynton’s Oak (Quercus Boyntoni Beadle, or Q. stellata var. Boyntoni Sargent) is also fairly abundant in the barrens and it is also a shrub 1-3 meters tall. Vacciniwm sericeum is conspicuous when in bloom on account of its very large flowers, perhaps the largest of any species of the genus. Prunus alabamensis is a shrub up to 2-3 meters tall, and it seems to be confined to the rocky glades and creek banks here. Kalmia is abundant along the rocky ledges and margins of the creek, and Rhododendron arborescens, not yet in bloom, and Xanthorrhiza apiifolia were found lower down the creek, where Cheilanthes tomentosa, Thelypteris asplenifolia and Tradescantia hirsuticaulis were also collected. Rhododendron canescens and R. alabamense were growing abundantly on banks a little above the creek and glades. The latter is rather a rate species and is distinguished by its glabrous winter buds. Over most of the mountain and surrounding region the forest is of larger growth and of mixed stands of Pine (Pinus echinata, P. taeda, P. palustris and P. virginiana) and deciduous species, it which Oaks, Hickories, Maple and Ash predominate. Tupelo, Miss., April 21st. We came into the little county seat town of Hamilton in a ste rain yesterday evening, and chanced to stop for lodging at a house nearly opposite the district Agricultural High School. After having had supper at the restaurant I talked to our host, a Mr. Love, and explained to him the object of our trip. He proved to be an interesting man, a native of Cape Cod, who had wandered over a good part of the world, and was fond of nature and out-of doors life. He was much interested in our work and offered - guide us the following morning to some places that he knew, where he thought we would find good collecting. Accordingly we set out early and after a short drive turned 7“ . the highway, near the site of the old town of Pikeville, which flourished in the days before the coming of the railroads. After travelling as far as we could over a rough mountain road we left the car and walked over to Dugan Creek, a small stream thst 2 has cut its channel through sandstone beds, which form dlifis — and overhanging ledges along its course. On the rocky banks s and along the cliffs I noted Fagus grandifolia, Magnolia acuminata, — M. macrophylla, Rhododendron alabamense. Corylus americana a eB T'suga canadenis, a curious mixing of northern and southern speci® ae The Hemlock is locally abundant here and some of the trees aI i of a large size. It has been recorded from several other ce in Alabama by Dr. Harper and others. Ferns were abundant, the we > i ‘ 1982] PALMER, LEAVES FROM A COLLECTOR’S NOTE BOOK 431 following species were noted and collected here: Osmunda Clay- toniana, Thelypteris noveboracensis, Adiantum pedatum, Asplenium pinnatifidum and Trichomanes Boscianum. Several colonies of the Trichomanes were found under the overhanging ledges of sandstone where the surface was kept permanently wet by seeping water and where direct sunshine penetrated only for a brief time each day. The fern-like moss, Fissidens polypodioides, was growing very luxuriantly with the fern, as well as several other mosses and Holly Springs, Miss., April 22nd. We drove back several miles this morning to investigate some thickets of Crataegus that we noticed yesterday evening, but which we did not then stop to examine on account of the rain. The hills here are calcareous, underlaid with a soft Cretaceous limestone, and Crataegus israther abundant. We stopped at two localities between this place and the little town of Mooresville and found what I take to be C. tersa, C. macra, C. frugiferens, C. amnicola, C. biltmoreana = he apiifolia, the last growing in low ground along a creek. A species of the Crus-galli group was also abundant. It is a small, moderately spiny tree, with slightly villous branchlets and corymbs. oe oe an full bloom and I secured good specimens, but do not rec- _ ‘SBnize it and suspect that it may be an undescribed species. Malus _ 4ngustifolia was also common and in full bloom in the thickets, wed I also found here the little Adder’s tongue Ophioglossum es pacman. Along the banks of the creek near Mooreville the ue = Was also abundant, and I found Castanea pumila, C’. dentata, y Writing ' canadensis and Rhododendron canescens Var. subgla- Bt oe along the same stream. a bviea back through Tupelo we turned aside to visit a Negro com p aa School, near Okolona, in which the father of my travelling i ai: me + had long been interested. We found the wife of the a: ‘After y: tin charge, apparently a very capable and energetic woman. : ; and *tenand several of the buildings, including class rooms, shops 4g. President’s house and seeing the students at lunch in the ae = eal, we took a hasty departure, favorably impressed by the Po "Pathetic, though heroic effort being made to carry be the : and in + kaha Negro education here with inadequate equipment oe face of local indifference and hostility. e258 i n Tupelo and Okolona there is some botanically interest- hie The soft marly limestone comes to the surface . Po ag and large fossil oyster shells are a undant. Crataegus es ftequent in the thickets as well as the unknow» Crus-9e ‘een near Mooreville. -A patch of Cercis canadensis, 0 432 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm which the plants were all shrubby and not more than 2-3 m, tall, had the leaves more than half grown but still retained clusters of flowers on many of the branches. Sikeston, Mo., April 28rd. After crossing the Mississippi at Memphis this morning we fol- lowed the highway which is built through the wide alluvial valley, — making our first stop north of the town of Marion, Arkansas. — Crataegus trees in full bloom in the low partly cleared woods at- tracted out attention. On investigation they all proved to be ¢. viridis, which is one of the commonest species of the valley, ranging from the Gulf to northeastern Missouri. I also collected here prinus, Ulmus crassifolia and Ilex decidua, which last was in Bisbis and with leaves nearly fully grown but retaining a full crop of bright red berries from the previous year. I remember having _ found it similarly retaining the fruit a few years ago, near Salisaw,. Oklahoma, and from specimens in the herbarium of the Arboretum it would seem that this is not unusual in the southern part of its range. In the afternoon we made a stop in the lowlands of Pemiscot County, Missouri, near Hayti. Such typical species of the swamps and lowlands as Tazodium distichum, Arundinaria macr Populus heterophylla, Planera aquatica, Catalpa speciosa, F rarinus profunda, Forestiera acuminata and Bumelia lycioides were growing here. Catalpa is fairly abundant and is native here. The southem Bumelia is rare so far north and I have only seen it at one other station in Missouri, near raed Ee Butler County. Cape Girardeau, Mo., April 24th. This morning we crossed Crowley’s Ridge, just west of the town of Campbell, Mo., and stopped for a few minutes to examine flora. This remarkable bit of relief, surrounded on all sides by the lowlands of the valley, extends along the great river, which at eat! time is supposed to have flowed to the west of it, from County, Mo., to Helena, Ark., some fifty miles below Memphis. In places it rises to a height of more than 100 meters above surrounding lowlands. The formation consists largely of UC — solidated beds of sand, gravel and clay, with loess becoming ™™ — common towards the southern end. It was originally and still is A many places covered with a forest of deciduous trees, inel ad - several eastern species which reach their western limit here, . with a flora, rich both in woody and herbaceous species, et : different floristically from that of the surrounding country 2 fers and Ericaceae are Sat ot tas , LEAVES FROM A COLLECTOR’S NOTE BOOK 433 ee as of Pinus echinata that had invaded the gravelly - hills near Jonesboro, Ark., and the Tree Huckleberry, Vaccinium _ wboreum, is found rarely in the same vicinity. Several species of aks Hickories Black and Sweet Gum, Beech, Tulip-tree, Elms, Maple, L Linden and Ash are the commonest forest trees, with several pecies of Prunus, Crataegus, Malus, heaepaie Hydrangea and shrubs ‘as undergrowth or in more open place only thing of especial interest found at this place today was that may be a hybrid between C. viridis and a species : ot ie Tg group. date in : oto we Bees 3 in the outskirts of Poplar Bluff to coccinea var. tubercadtate, which with the ed with coarse gravel, and on account of a eae while ae 434 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm carrying quantities of flat rocks from a hillside to build up as we pried the wheels out, since it was impossible to use the jack. This took so long that by the time we had the car out and loaded again it was too late to go over to the interesting hills, which here form the extreme northern end of Crowley’s Ridge, and which I had. hoped to photograph. However, we climbed one of the nearest and examined the flora. This is close to the little town of Perkins, Scott County. The hills, rising abruptly from the surrounding lowlands here have a remarkable appearance, made more striking by the fact that while the bottoms have been mostly cleared for cultivation, these ridges are still heavily wooded. Many of the Beech, Oak, Hickory and other trees attain a large size. I photographed one large specimen of Beech and a King-nut tree, the trunks of which had grown together, apparently forming two halves of a composite trunk, for several feet above the base. This is the only part of Crowley’s Ridge where I have seen outcrops of stratified rock, the usual gravel and clay deposits here having covered some rem- nants of the hard Ordovician dolomite, which is being excavated by - erosion and is beginning to appear as cliffs and ledges in & few places. Hugo, Okla., May 26th. Our camp last night was in a beautiful and rugged part of the Ozarks, a few miles from the little town of Talihina, Oklahoma. After breakfast this morning I started out for a walk of several miles over the steep dividing ridges and deep hollows that, viewed from a high point, extend as far as can be seen in all directions. The prevailing geological formation is Pennsylvanian sandstone, — varying greatly in stratification and density at different ae ; Most of the country is still heavily wooded with a mixed f Pine (Pinus echinata) and deciduous species of Oaks and Hickories — prevailing on the ridges, and many other trees and shrubs along escarpments and ravines and in the deep narrow valleys. Some the species found in such places are Juniperus virginiana, Castone® oe ozarkensis, Quercus borealis var. maxima, Q. Shumardii var. Schnee Ulmus alata, Nyssa sylvatica, Amelanchier canadensis, Acer rubrum : A. saccharum, Cornus obliqua, Vitis aestivalis and V. rupetrs — Along the rocky margins of a mountain creek I also : Carpinus caroliniana, Rubus flagellaris, Ceanothus ovatus, Hypere™ i slightly pubescent foliage and branches, which I do not recognize: : in the In places along moderate slopes hard beds of the come to the surface, forming glades or barrens, conspicuous ’ : - 435 f forest for the absence of trees. I stopped to examine and to take _ some photographs in one of these, which was several acres in extent. - Conspicuous amongst the large number of herbaceous species here were Cheilanthes lanosa, Camassia esculenta, Delphinium azureum, Talinum teretifolium, Tephrosia virginiana, Tragia nepetifolia, Ptilim- nium Nuttallii, Spermolepis echinata, Opuntia humifusa, Hedeoma his- 7 pida, Kneiffia linifolia, Pentstemon arkansanus, Ruellia ciliosa, Specu- | laria leptostachya, Coreopsis grandiflora (not yet in bloom), Thele- sperma trifidum, Echinacea angustifolia and Krigva occidentalis. In the afternoon I stopped along a little creek near the village of Finley to photograph some fine blooming specimens of Yueca | arkansana. A little further on, along the bluffs and banks of Mill § Creek, I collected Tilia floridana, Chionanthus virginica, Castanea - Kaimichi, along the bed of a little mountain creek, I found Salix _ petiolaris, Andrachne phyllanthoides, and a curious and unknown form of Witch Hazel. This seems to be quite an extension south- F ward of the known range of Salix petiolaris, which was previously known from Canada to northeastern Missouri. The Hamamelis is evidently closely related to the spring-blooming species, H. vernalis, which is so abundant along the rocky streams of the Ozark region. Although the plants here are smaller than the average for Hamamelis vernalis, they have the characteristic : ‘omentulose branches and leaves more or less pubescent along Alfred Rehder as a eis vernalis forma tomentella }. The type specimen “ or ee from Potean, Oklahoma, which is in the same gem j a 5 == as Kaimichi, and it is also in cultivation ape rahe _ hen However, the Kaimichi plants are so clets® _ tly from the typical form of Hamamelis vernalis but also pen ‘Rima iomeniella that they can scarcely be reg * apy : Peay referred to the latter. The leaves are thickly perv sibs ot @ persistent close felty pubescence, often tawny oF red hai ah _ the etioles and veins, and with scattered grayish stellal® Wy na also copiously villous: vernalis! Professor rgent .cryetet - ‘Sat on both surfaces, and specimens that he has noted as kee q ™ his type material scarcely differ from the type specimen ©" : “Sergent, Trees and Shrubs, m. 137 (1911). 436 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm forma tomentella. The foliage of the plants found near Talihina is so different that they appear more distinct than any American species that has been distinguished since the publication of Hama _ melis virginiana. But since a name has already been given toa tomentulose form of H. vernalis, to which this plant is clearly related and since it is likely that other intermediates will be found it is perhaps best to regard all of the tomentose forms as one variety, _ and I am therefore calling the pubescent plants Hamamelis vernalis var. tomentella. About Antlers, Oklahoma, the soil of the uplands is extremely sandy. Quercus cinerea grows in such soil, and I found it near here several years ago, but did not see it today. Crataegus pilifera is growing here as a small tree 4-5 m. tall, and C. uniflora asa slender shrub scarcely a meter tall, but in abundant fruit. Jat- ropha terana, Lithospermum Gmelini and Pentstemon paucifiorus were also collected in the sandy open woods. As we approached Hugo there was a marked change in the | _ character of the country and in the flora. Cretaceous limestone — comes to the surface in many places and the soil resulting from it is a stiff black gumbo, which is quite fertile, as shown by t better quality of farm improvements and other evidences of pros- perity. Woods are confined largely to the vicinity of the streams — with broad stretches of upland prairie, and glades are are frequent — on limestone outcrops. Maclura pomifera is native here, and the little wild Rose Rosa foliolosa is abundant. There is also 4 : profusion of wild flowers in rainy seasons. Pentstemon Cobaea, — and Petalostemon albidus are conspicuous and I also maser little suffrutescent Phyllantio polygonoides. . Stillwater, Okla., May zat - We sitchind: our tent last night along a pretty little stream, Pennington Creek, near Tishomingo, the county seat of Johnston ] ; County and formerly the capitol of the Chickisaw Indian 1" There is an outcrop of granite in this section, and the creek bs cut its channel through the hard igneous rocks, great masses ° which lie scattered about, piled up and eroded into the ast 1 Hamamelis vernalis var. tomentella (Rehder), comb. a Hamamelis vernalis ¢. t peutic Racer te as Arnsth Sob: 1. 256 59 8 20) a ucescen dense = seme er juve A typo differt fois -stellato-pubescentibus, infra lato-pubescentibus; ictinke aecvicbe fulvo-pubescentib' vo-stellato-pul tes. ial Near * * hi, wy. Zz sz... gt . 7 . Kaimic Faas The plant here | differs from being densely felty-tomentose on the bp oedrn of the leaves i i932] PALMER, LEAVES FROM A COLLECTOR'S NOTE BOOK 437 fantastic shapes. ‘The locality bears the rather sinister name of the “Devil’s Den,” and it is much frequented as a summer resort and by campers and fishermen. There is considerable small timber along the creek and in the more rugged or more protected places amongst the granite, but over considerable areas where unbroken beds of the rock come to the surface, trees and shrubs are absent and only small colonies of peculiar herbaceous plants grow in the slight depressions where thin deposits of soil have accumulated and where water stands for some time after rains. In some of the larger depressions there are permanent pools and small ponds. Juniperus virginiana, Juglans nigra, Carya Buckleyi var. arkansana, Quercus stellata, Q. Muhlenbergii, Q. macrocarpa, Q. velutina, Q. Schneckii, Q. marilandica, Ulmus americana, U. alata, Celtis laevi- gata var. texana, C. reticulata, Maclura pomifera, Prunus lanata, Cercis canadensis, Sapindus Drummondii and Bumelia lanuginosa are some of the commonest trees, with Yucca arkansana, Rubus ostryifolius, Rhus glabra, R. copallina, R. Toxicodendron, R. trilo- bata, Ilex decidua, Lonicera albiflora var. dumosa and a few other shrubby species occupying most of the area. Vines are abundant in the woods and amongst the protecting rocks. The following climbing or trailing species were seen: Smilax hispida, S. Bona-noz, Cebatha carolina, Aristolochia tomentosa, Vitis cordiformis, V- lpina, Cissus incisa and Melothria pendula. : Confined to the margins of the stream are Salix nigra, S. longupes Nar. Wardii, Populus deltoides, Alnus maritima, Platanus oceiten’ ss is iDandast in the sewhat heavier soil Gaillardia pulchella is very abundant. 438 JOURNAL OF THE ARNOLD ARBORETUM [vou mm PARASITISM OF MYXOMYCETE PLASMODIA ON FUNGOUS MYCELIA Frank L. Howarp! anp Mary E. Currie With plate 54 To FrurTHER the general thesis that the Myxomycetes play a rile in the consumption of fungi which cause wood decay, the digestion and assimilation of the mycelia of chiefly lignicolous fungi by plas- modia was studied after the parasitism of the plasmodial stage of several Myxomycetes upon mushrooms and polypores had been demonstrated (5). Enlightenment upon the problem was sought by observation of the habits of plasmodia in the forest and also by : : laboratory tests of the feeding habits of plasmodia on pure cultures of numerous fungi. This study has brought to light the mycopha- gous habit in several species of Myxomycetes, in addition to those already reported (5), and has disclosed a wide range of fungous mycelia capable of being digested by plasmodia. RELEVANT LITERATURE The digestion of the mycelia of fungi by slime molds has, with but one or two exceptions, been reported only incidentally by in- : vestigators in connection with their study of some other phase of the biology of the group. The work of A. Lister (6) gives us ee excellent notes on the behavior of the plasmodium of utricularis, but only the reaction of the plasmodium to the a lium of a chance, unidentified fungus was observed. cultivated B. utricularis upon moist bread and he observed that the 5 hyphae of various species of Aspergillus and Penicillium, whi developed on the bread, were dissolved and absorbed by the plas | modium. Physarum nutans has been credited by Elliott and Elliott ash with the absorption and destruction of the mycelia of Bulgaria polymorpha and Coryne sarcoides within the wood of an oak — Sanderson (7) frequently encountered Physarum 4 Cooke upon the mycelium of Sphaeronema fimbriatum which of tapped surfaces of Hevea; Physarum reniforme Lister 0D the cortex of Hevea attacked by Ustulina zonata; and Arcyria | Wettst. associated with Sphaerostilbe repens in Malaya on fuel es of split timber and on small diseased roots of Hevea. One read Sanderson’s paper without feeling that the plasmodia of aye : mycetes may parasitise fungi responsible for the decay of wo _1 Fellow of the National Research Council, 1930-1932. _ 19323] HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES = 489 Skupienski (9) describes the plasmodium of Didymium difforme feeding on Aspergillus glaucus, Sterigmatocystis sp., Penicillium sp., - Stysanus sp., and various other molds, yeasts, and bacteria. He also found that the plasmodium of Didymium nigripes would digest _ the hyphae of Penicillium (8). On the other hand, Celakovsky (1) working with Chondrioderma (Didymium) difforme Rost. reported plasmodia as non-mycophagous, since he observed plasmodia en- velop spores and hyphae of Penicillium glaucum, Mucor stolonifer and Phycomyces nitens, and later relinquish them unharmed. MATERIALS AND METHODS The plasmodia used in these studies were collected in the forest and brought into the laboratory for cultivation on media, or were cultivated directly from spores (4). The cultures of fungi employed _ Were either isolated by the authors or obtained through the gener- alg other workers, especially Dr. Irene Mounce and Dr. C. L. In the earlier trials (Tables 1 and 2), in order to test the parasitism of plasmodia, petri plates of nutrient agar were inoculated in the center with the fungus being tested, which was allowed to grow e until the colony covered about one-half the diameter of the dish ‘ before a bit of plasmodium, previously freed from contaminating ‘4 fungi, Was transferred to the culture. The various standard nu- _ Went agar media used for the culture of the fungi included rolled ei oat, ground corn, ground corn and dextrose, malt extract, and potato er trials (Table 3), the senior author devised the ib : a3 ar . ick about one centimeter : similar disk of nutrient agar upon which the a : When the fungous hyphae grow from 144 a plain agar, a piece of plasmodium is introd Be ‘uaa the hyphae observed. This ch nutrient agar disk and its employed and where the plasmodium very slowly hee? a, i di rs: i whether the growth of us? eete was due to digestion of the fungus or t0 8”. it directly from the agar. In spite of repented ae eon: sible to free certain of the plasmodia eae baer ob- ) Was an advantage, as it helped to inh Ot 4 the » ars in the later trials the H-ion concentration of fe y into account and was adjusted t pe ah 440 JOURNAL OF THE ARNOLD ARBORETUM [vou. xu THE PARASITIC HABIT AND THE INFLUENCING FACTORS In order first, to determine if the digestion of fungous hyphae by Myxomyecetes is affected by the nutrients in different culture media, as has been reported for the consumption of some bacteria by plas- modia, and second, to find a favorable medium, the junior author made a few preliminary experiments in which fungi were grown on five different agar media, namely, potato dextrose, linseed, Vic TABLE 1.—PARASITISM OF PLASMODIA er FUNGI GROWING ON DIFFERENT MEDIA 0 a a t= a4 a 4 & 8 5s ¢§ a o¢ Bi 3 rt S-¢8 @ a@ go 3 3 e 8 am 8 Bee 88 8 28 of - 259 8 2 € 3 8: 5'e ere g a. Se 2 a a 8 eB a & i Ble 2423234465 : 4 a| 8 i 2a 8 hea ae ae 4 SSR EERE EE eB Bee Funeus AaneaeaertHAa Ae am me Corn Dextrose AGAR 1. Collybia velutipes |VP — — P P — VP P P — VPP PF 2. Pleurotus ostreatus | VP P — P P — — — P — 77 — ¥ heer f op Pe t. P ie 4. Polyporus resinosus Pn — NPS ee PO Rouen Oat AcaR 1. Pleurotus serotmus | VP P — — — — — — VP Sree 2. Polyporus resi Bee tect ee eee Potato Dextrose AGAR 1. Fomes applanatus ee et ee Ba ee ie < 2. Pleurotus ostreatus |— N VP Merce ko VP eee. 8. “e 3 Oh. ie oN LP Meare ae at Re iin VP V8 7 = 4. Polyporus resinous | P N — — — — — — — Ps Eee | LinsEep AGAR Ea 1. Collybia velutip 5 he 2a be i eee 2. Fomes applanatus ee Ie ee I ee ee ee ee jas ee : Pbwessciok ae Seow ae ea cake | Victa Fapa Acar | take 1. Fomes applanatus pees GR See Pe ee VP: thor. 2 se fi f . N SF NES all Sap ee A ee me ere WMO EE ay Pha ee és ss eee Benen Svedpege bre ee 0 he N eager eeet ; Polyporus resinpius © | NU es ee ee. et 5. Polystictus versicolor | VP — — — — — — — — => Faba, rolled oat, and corn dextrose. Pasmada were tra f thei ; 0 = the plates in which the fungi were growing and the cenit | parasitism on the mycelia are presented in Table 1. The 4 terms used in Table 1 and in subsequent tables to describe the =a - of parasitism of the mederaidiiy Lesey tere the criterion ‘ * | 4992} HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES = 441 tate of digestion of the fungous mycelium, are: ‘very parasitic,” “parasitic” and “‘not parasitic.” “ Very parasitic” denotes a case in which the mycelium of the fungus was rapidly and usually com- pletely consumed, as for example, the case of Badhamia magna on - Collybia velutipes. Where “parasitic” is recorded, the mycelium was more slowly and usually not entirely consumed. “Not para- sitie” is recorded when the mycelium was not digested, even though the plasmodium may have passed over it. : The incomplete data recorded in Table 1 indicate that a plas- modium cannot always attack the same mycelium when grown on adifferent medium, for example, Badhamia rubiginosa attac ed and consumed the mycelium of Pleurotus ostreatus growing on corn dex- trose agar but avoided the same fungus growing on potato dextrose agar. This would seem to indicate that the nature of the medium plays an important part in rendering the mycelium susceptible to attack by a plasmodium. It should be noted, however, that some species of fungi even on the same favorable medium and exposed to plasmodia capable of vigorous parasitism on other fungi, invariably rman unattacked. For example, Physarum cinereum avoided the mycelium of Fomes pinicola growing on Vicia Faba agar but did attack Fomes applanatus hyphae growing on the same agar Ex- amples further substantiating this principle are brought out m later experiments (Tables 2 and 3). a Table 2 shows species of plasmodia which consumed mycelia oor on corn dextrose agar. Petri plates containing this agar heey : rete with different fungi and when a considerable mycelium ny Pi tte gig four or five cultures of each fungus were 10° wait eee B: pears piece of plasmodium. The plates were kaacurenst re ahyy This work brings out many idiosyncrasies 9 Tids _ He plasmodia to digest different species of fungi. Some slime mé Pe Fare! es be generally parasitic upon fungous hyphae while peter _®eotly selectively so. The length of life of # plasmodiwi® tar : figs. 1-10). Sometimes 4 celial culture, digests the my celium with eet 2 my hyphae, as does and leaves a path freed from fungous | etimes ‘the plas- ol wm on Collybia velutipes ge Li deni spreads out in all directions, exposing ® 42 JOURNAL OF THE ARNOLD ARBORETUM [vou xm circle of bare agar as it consumes the mycelium; such is true of Trichia decipiens Macbr. and Lindbladia effusa Rost. on Daedalea _ confragosa. And sometimes, if the mycelium is in a tough appressed layer, the plasmodium removes only the superficial hyphae, as for example, Lycogala epidendrum Fr. on Lenzites betulina. TABLE Pett eo OF PLASMODIA UPON FUNGI GROWING CORN DEXTROSE AGAR. _ N—not parasitic, ethaxzer VE very parasitic. | 3 = gg a g g 5 £24 2 Ble £ s S @ @ 3d 8° 6 @ ele ee SESE ER ERE aie 3. 3° 8 2 2&3 8 8 ooo Bloat ease feeaaai - oot —_ -_ Biggs Pepa 46 Fb beg 3 © 3 §$ Ss 2? € 3 8 Ss 5 sepa ER ERE EB BEE Funeus Se Re ee RETR Ree BE 1. Collybia velutipes VP — P+ P PP op FR ee Daedalea confragosa VP. PF —N PP PF — — VP SS 3. Fomes p __ applanatus re: F — Pp PP PP — — he 4. Fomes fomentarius EE PoP. PoP PP Pe ee 5. Fomes P igniarius VP — — P— P P PP - - 3-7 6. Fomes ‘ pini 2 ee Se ee ee Noo se tee 7. Lentinus pe aber é lepideus eo Noe pp pl SS Pe ae betulina ee ee PP — — 2 Ree ostreatus veo Pe, £ FP Pe PP VERA 4 Pau y VR Ne Pee oP POP VP VE AS sind g q - plang? ts resinos PF. NN ONON: N i ERS Gwe GY gc : 12. Polystictus nigro| ‘ ¥! Pp. Ss -margi Se er Pp PH — Pee Slices —-—— — — — — — — P- Tp be Polytictus sp Poe) ee oe Pe Pr Pp i NN PR 2 enw Ny PS POF ee 16. Trametes pini FNS Pie ne PK UP Ree asmodiz ‘on fungi growing on rolled oat agi 4 was tried by the senior wats with the results: Leocarp™ — fragilis slowly consumed hyphae of Monilia (Neurospora) wees and of a Penicillium sp. Physarum polycephalum rapidly a the mycelium of Alternaria sp., Aspergillus sp., axtri \Blaazlilliltilalalittitttlaaz Aa Oa Il\alaazzbazabbazazazeaa| |zmaazczacaa| alazzetalazzeaeuas | laa l|aaae | Bl zaz|PeabazzaZnas | Z2eAzAZnAaAe alma | ahaa ZzAaZ | AMAzAAZAAA aanzazee||azzlaazzil |laazezaeeae ‘llzzzllllllaallallaala2alae mwlatlilililazazllellllalazma Blzb1 || Beez lable |GESEEES alzzz | |ameaczzzzzze | ZAZAZZA& | o Mmiizlilititlitlatlitilaatal Flay ly ay Ay ay hag ng hag ay ay ay lg BRRRR | | | lambs | SRaSSSS ESSE > PN ga EP oat Re le ie a og Ch Ct am Flt Pl ey Fy Ge yy le ys walat2l | |Baazaal lal |2a2e24442) So — es ee - - rrr rr wrerrrrr wizezzbee ree ee | am Amne mpg |lizzziiiiilazlitl lal laa) uiny | zazbazhn | zzmuceZZ | MEAZARLAzA [Baebes |znnennn | MPaaMeeee Je |Razene |azezian | Ameen ZeZe| N—not parasitic, P—parasitic, VP—very parasitic. £8 | ooo: Bee ab ek esa » 22 g Sese 4 e8kbS29s8 Gboee ooees «|e BeSEba225 S528! 2283582 Ete: 3 a ott I 2 o Ta 6 5 gee }| eS PRES ES gTE pepicat sss 5 £| Sbs238 P SS bi22;2222252 = . % ae a i 2s mm el q o = wm. : Sr Nu Be -¢ BESS Seas eze 5, \aa6868 ¢ Seiissteace oe) 444 JOURNAL OF THE ARNOLD ARBORETUM [vou. xin Exidia glandulosa, Merulius americanus, Monilia (Neurospora) crassa, Nidularia pulvinata, Penicillium sp., Tremella mesenterica, and Tremella sp.; but did not digest the hyphae of Beauveria globu- lifera, Guepinia spathularia, and Mucor sp. Physarum viride plas- modia rapidly digested the mycelium of Chaetomidium fimeti; and more slowly digested hyphae of Aspergillus sp., Beauveria bassiana, Monilia (Neurospora) crassa, and a species of Penicillium. A wide range of species of fungous mycelia capable of being digested by various plasmodia was tested by the disk insertion method (Plate 54, fig. 1), and the results are tabulated in Table 3. Occasionally cultures exhibited a puzzling variation in the digestion of the mycelium on oat agar disks and to obviate this, duplicate cul- tures were prepared. Indeed, in some instances to give added cer- tainty, if the duplicate cultures showed any discrepancy, three be: four additional cultures were made. In such doubtful cases, it is the average behavior of three to five plasmodia that is recorded in Table 3. The variation, however, was generally a question of the rate of digestion of the mycelium by the plasmodium. For mn stance, Arcyria occidentalis slowly consumed the mycelium of Fomes igniarius in two trials while in a third trial the plasmodium com- pletely digested the mycelium, but in Table 3 this plasmodium 1s listed merely as “parasitic” instead of “‘very parasitic.”’ Likewise, out of five trials of Physarum tenerum on Fomes fomentarius, three gave evidence of parasitism while the other two did not, yet the -plasmodium is listed as “parasitic.” As previously stated, variation was commonly one of the degree of parasitism with but one or two exceptions. One exception occurred in the case of Fuligo septica digesting Lentinus lepideus, in which the plasmodium died on the mycelium in two cultures, avoided the mycelium 1 third, but completely consumed it in a fourth. : The fungus listed as “mycorrhizal fungus” (see Table 3), 5 # subculture of Mycelium radicis atrovirens isolated in Sweden by E. Melin and obtained through the courtesy of Mr. A. B. Hatch. This non-sporulating fungus produces a dark submerged mycelium and ! a lighter aerial mycelium which was consumed by Physarum poly- cephalum, Physarum tenerum, and slightly by plasmodium T. RE The behavior of the plasmodia in digesting fungous myclit — seems unquestionably to vary with the medium upon the fungi are growing. The plasmodia of Physarum tenerum and P her ¢ rum polycephalum more rapidly consumed the mycelium of ‘Scho phyllum commune from disks of malt extract agar than from disks of rolled oat agar. Another, more striking example, of the effect of the medium upon the ability of a plasmodium to digest eunttk 1932] HOWARD AND CURRIE, PARASITISM OF MYXOMYCETES 445 | | mycelium was demonstrated by Hypozylon coccineum. This fungus isvery slowly digested on oat agar by the plasmodium of Physarum polycephalum, while in repeated trials the same fungus growing on ) bean pod decoction agar is rapidly consumed by the same myxomy- | cete. It is of interest to note that on the first medium a dark green pigment is produced by the fungus while on the second it is absent, but whether the presence or absence of this pigment is in any way connected with susceptibility to parasitism remains a question. Similarly, a yellow undetermined species (Plasmodium T) digested the white aerial hyphae of Hypoylon coccineum from a disk of bean dium may flourish on the mycelium of one fungus and not attempt to attack that of another species. Plasmodia of the same genus but of different species of slime mold may show as great a variation In parasitism as those of two different genera. In general, it may be id that some Myxomycetes are restricted in choice of host while others seem to be generally mycophagous. SUMMARY Laboratory and field observations have disclosed the poe gous habit of several plasmodia of Myxomycetes other than fifteen species that the authors (5) recently reported digesting ey eles, which furthers the hypothesis that paqsagre ae 7 digest. the mycelia of a wide variety of fungi responsible cetes found to consume it were: 1, Arcyrt@ -— weidentalis, 2, Badhami piginosa, 4, B. atriculariy _ adentalis, 2, Badhamia magna, 3, B. epee se “a clavat 8 H. (T@parium, 9, Leocar Fis, 10, Lindbladia effusa, 11, Lycdd® ™ 2 ciaweke pus fragilis, 10, wee 4 - a ‘pidendrum, 12, Physarum cinereum, 13, P. flavicomum, iget ti 17, P. viride, 18, Stemom “fusca, 19, Trichia decipiens, 20, T. persimilis, ane cae oe Retods for testing the parasitic habit were rel) hb | Growing (Plate 54, figs. 1-2). The ail! aie be iely wood-inhabiting fungi were tested 2 ot ee ‘din varying degrees by plasmodia- == her thanks to closing, the junior author wishes borer’ : 446 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Professor J. H. Faull for the direction and the advice given during _ the progress of her work which was carried on at the Univenae Toronto during 1919-1921. The senior author takes this oppor- tunity to thank Professor J. H. Faull for making possible the joint 4 publication of the observations made independently by the authors, — Professor W. H. Weston, Jr., for his interest and counsel given, a the Board of Fellowships in the Biological Sciences for a grant which made this work possible. LaBoraTORIES OF CryproGamic BoTANY Harvarp UNIVERSITY LITERATURE CITED . Cetaxovsky, L., Jn. Ueber die Aufnahme lebender und todter ver daulicher Kérper in die Plasmodien der Myxomyceten. (Flora 76: 1892.) _ 182-244. Exuiorr, W. T. & J.S. Etziorr. The sequence of fungi and mycetonoa. (Jour. Bot. 58: — 1 Hiuron, A. E. tes on the cultivation of plasmodia of Badhamia ularis. done Quekett Microse. Club ser. 2, 12: 381-384. 1914.) _ How waRD, F. L. Laboratory Spgs of Myxomycete plasmodia. — (Amer. Jour. re 18: 624-628. 1931.) ————— & M. E. Curriz. Parasitism of myxomycete plasmodia on the sporophores of Hymenomyecetes. (Jour. Arnold Arb, 13: 270° 48-49. 1932.) Se oe Te ISTER, A. Notes on the plasmodium of oe utricularis and Bre feldia maxima. (Ann. Bot. 2: 1-24, pls. 1 1888 be - R. Notes on Malayan Abate (Trans. Bate askew Soc. 7: 239-2 256. 1922. Se Oo - NN GQ Ng oni ed eee nad Didymium ditfome ; Duby, ezese pierwsza. yea paytclogie Poloniae, 5: 255-336, P wie 12-18, 1998.) i : EXPLANATION OF PLATE 54 oe Fig. 1. Plasmodium of Physarum polycephalum leaving the transferred piece of agar and m ey! ati ag ke penne of Trametes pin. growing on an oat eat tik which has been inserted ge ge plate of plain agar Six hours later the fungous hyphae ‘ completely conmumed, x ‘ Fig. 2. Plasmodium of P. polycephalum co consuming the white ey ; Fomes igniarius fro m an oat agar disk and leaving some older dark-colored hyphae. nodium Fig. 3. big wr of a virescens sere the hyphae * i mes Daedalean 4) Fig. 5. Pi pepe of F way pasty aoe Pl , fomentarius hyphae dextrose agar. Fig. 6. Plasmodium of Brefeldia maxima reseed hyphae . colt velutipes. Jour. ARNoLpD Ars. Vou. XIII. PLATE 54 uncus MyYcELIA Parasirism oF MYxoMYCETE PLasMODIA ON F NOTES 447 fig. 7. Plasmodium of Brefeldia maxima parasitizing the mycelium of . aedalea confragosa on corn dextrose agar. Fig 8. Plasmodium of Lycogala epidendrum attacking the hyphae of Dae- ea. confragosa. Fig.9. Plasmodium of T'richia decipiens removing the mycelium of Dae- dalea confragosa from corn dextrose agar. Fig. 10. Plasmodium of Badhamia magna consuming the hyphae of Fomes applanatus. NOTES The Arnold Arboretum during the Fiscal year ended June 30, 1932. Cumaticatty the year 1931-32 proved generally favorable to the growth and well-being of the collections. On February 4th, however, there was a heavy fall of snow that adhered to the trunks and limbs of trees and burdened shrubs and coniferous foliage with a thick mantle of white. Although the landscape effect was ex- . ceptionally beautiful in the Arboretum, much damage was done a and several weeks had to be devoted to the pruning necessary to femove injured branches. In the spring the rainfall was plentiful and adequately supplied the needs of the plants for moisture. The Cherries and Rhododendrons produced an abundance of lowers and were exceptionally fine. Now and again, usually through the unintentional carelessness . . visitors, fires break out in the arboretum and threaten injury Eh destruction to valuable specimens. In the spring of 1932, “veral serious fires, a few doubtless of vandalistic origin, caused es tstressing damage. One of these menaced Hemlock Hill on April Fe ee bet before it was finally checked, had destroyed a plantation - ofTapanese Yews. Tt is evident that fire is an evil that must be : revamp by the most efficient means if the work of years 1 not | ie Sacone in a few minutes and losses incurred that can never = han Continued our exchanges of plants, cuttings. 8f afts and yr . = year there were sent out 833 cuttings and oer: other ine its and 941 packets of seeds. There were received a ge Sea ‘institutions, 1666 plants, cuttings and grafts and 294 | et a alee four hundred periodicals, bulletins <02) ae " Cotte ee he Library from all parts of the world a f 448 * JOURNAL OF THE ARNOLD ARBORETUM [vou xm by Grant Dooks Darker. The second number of the Contributions is in the Press and the third number is under preparation and should be issued in 1933. One thousand, one hundred and sixty-eight visitors registered at the Administration Building. They came from thirty-one of the United States, from Canada, China, Hawaii, Ireland, England, _ Egypt, and West Africa. | The “Harvard Experiment Station” situated near the city of — Cienfuegos in Cuba, was originally associated with the Botanic Garden of the University under the supervision of the Director. It was founded by Mr. E. F. Atkins of Boston for the purpose of carrying on investigations looking toward the improvement of the varieties of sugar-cane and with the intention of assembling there a representative collection of tropical plants. As the enter- prise became more and more closely associated with applied biology, it was transferred to the Bussey Institution of Applied Biology. Early in 1932, the Corporation voted to transfer the Station to the Arnold Arboretum with the title, ‘Atkins Institution of the Arnold Arboretum.” In this regard it is worthy of note andr that in 1926, Professor Sargent had looked favorably on a proposal to place the Cuban Station under the general control and manage ment of the Arnold Arboretum and had assigned to Professor Jack the task of making a representative collection of the woody = plants that grow in the vicinity of Cienfuegos. In recent yea — Professor Jack has visited the Institution frequently and bes given close attention to the introduction of Cuban woody plants to ' the living collections, while diligently working on the formation of j an herbarium comprising the flora of the surrounding pe At the end of this report there is appended a ge. covering the publications of the staff and their ari for the year ending June 30th, 1932. This bibliography indicates veY — clearly the wide extent and varied nature of the investigation’ — that are receiving the attention of the Arboretum.—0. A. oe) Pathological Laboratory.—The Laboratory in Plant Pathology submits its report for 1931-2, making reference to improved a facilities for investigation, extension service and research emer a Pathological collections constitute an important eqn our laboratory because they afford both subject matter for ‘ae? and specimens for reference. Therefore, a consistent t ete F been made to add to the materials required for a more vege! : hensive solution of our immediate research problems, and a may be useful by way of Atrio of an wide oer of plant teronee oe ee NOTES 449 Ey contributions have been received from correspondents, eer aa been obtained by exchange and many have been ‘collected by members of the staff. Of the last named particular - mention should be be made of extensive accessions from Europe and America through G. D. Darker and from the western States and ‘England through J. H. Faull; the former are rich in ‘needle fungi, the latter in coniferous rusts. Pertinent to this sub- leased to report a closer contact with the Farlow Library rium, recently initiated by the appointment of Arbore- -tmnpesetation on the Administration Board of that institution. A second step forward in reference to improved facilities is the coon accommodation made by the Corporation of . - departments. We now happily record ap- the — peo yisien made for them in the new study ws the “needle eat’ fangs the et annialp the form 450 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm of a handsomely illustrated monograph, have been issued as No, 1 of “Contributions from the Arnold Arboretum of Harvard University.” Twenty-four species are described as new, 10 om Firs, 3 on Spruces, and 11 on Pines; also the pathogenicity of 5 species, of a limited number tested, has been established by direct experimentation. Distinct progress has been made on the beech disease problem in which the sequent agencies of an insect and a fungus are involved (John Ehrlich: The Occurrence in the United States of Cryptococcus Fagi Dougl., the Insect Factor in a Menacing Disease of Beech. Journ. Arnold Arb. 13:75-80. 1931). Success- ful control measures have been tested through the codperation of the Boston Parks Department. Also field work in the Maritime Provinces, where the trouble has been devastating, is now in course of completion, an undertaking made possible through a liberal grant made by the National Research Council of Canada. Quoting Mr. Richard J. Hayden, Superintendent of Parks, Boston, Mass, — from a recent article in Horticulture—‘This insect (factor) is new and has not had time to become widespread here. Hence its eat y eradication may prevent the establishment of a very serious pest. Mention also should be made of the completion of Dr. K.S. Chesters studieson Phytophthora Blight of Lilacs, and of his rectification of - error in the technique, fundamentally important, used by certa!? — investigators in the uncharted domain of plant immunology %* — means of recognizing acquired immunity in plants. Dr. . will continue his studies during 1932-3 in Europe. ae A list of other topics under investigation includes: rusts of Cont . fers; trunk diseases of Conifers; wilt diseases of Elms; spermogoni® ot of rusts (L. M. Hunter); “‘cedar apple” diseases of Red Cedi Apples, Hawthorns, ete. (I. H. Crowell and J. D. MacLachlan); ; mycorrhiza of trees (A. B. Hatch; Mr. Hatch’s present field work in the Black Rock Forest, New York State, is fully supported eh | a contribution from an unnamed donor through the rector © the Harvard Forest).—J. H. F. ar Cytology Laboratory.—During the past year additional © logical studies have been continued to determine the Te" between chromosome numbers and taxonomic grouping in differen’ species and genera of woody plants. The plants invesv™” include species and genera in the Ulmaceae, Tiliaceae, and beer aceae, as well as a number of rare or monotypic genera. 4 St sist : chromosome behavior in several genera of conifers by eee has provided information which may account for the morpholoe™ stability of this group of plants. A comparison of the chrome” of Yucea and Agave, in collaboration with Mrs. McKelvey, “that these genera are closely related, even though they have been placed i in different families by most taxonomists. Cytological investigations of hybrids have thrown some light on ~ the relationships of geographically distinct species, and have indicated the probable origin of the Pomoideae. Several studies on the mechanism of chromosome pairing and - vision have been completed during the year. This work is a part oof a general project involving an extensive study of the mechanism . - otheredity. Collaborators in this work include Dr. Anderson, Dr. ~ Hally J. Sax, Mr. Dermen, Mr. King, and Mr. O’Mara. During the past season more than 350 crosses have been made between different species and varieties of trees and shrubs. Rel- - atively few crosses between distinct species are successful, but he a hybrids obtained should be of considerable interest. g work has resulted in species hybrids of Syringa, oe Malus, Philadelphus, Ulmus, Lonicera, Ribes, and Rosa. Numerous crosses were made between Rosa rugosa and other ‘ _ species, many of which were successful. This work was aided by _ "special contribution from a friend of the Arboretum. —K. S. The Herbarium.— The Herbarium contains 358503 specimens, ; » having been added between July 1, 1931 and June 30, 1932. p the accessions approximately 2750 came from the United Sates and Canada, 1670 from Central and South America in- heey? of Mexico, $90 from Europe and Western and Central we Moo from Eastern Asia and 375 from Southern Asia and be ted 260 from Africa, 1250 from Australasia and 1200 rep- hae cultivated plants. oad More important collections received during the year g: more than 1000 numbers with many duplicates a Sitio the Pikaanniine’ Museum, 750 Chinese Seer, m Kweichou by Y. Tsiang and 425 Chinese Plants ct d ao F. nee ae Ward, 600 numbers with dupli- . Bangham, ! 160 : by J. Burtt Da’ * Se ane i avy: G. Klug in #in western China by J. F. Rock, 400 plants from northern Tlie ' collecte < 452 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm S. D. McKelvey, and with flowers and fruits in formaldehyde, cytological material, also insects pollinating the Yucca flowers. The fruit collection numbers now 7439 specimens, 188 having been added during the year. To the wood collection 83 specimens were added bringing the number up to 2365. The collection of negatives of types and of other herbarium specimens consists now of 1857 negatives, 91 having been added during the year. The installation of 16 new cases and 29 half-cases necessitated a rearrangement of the whole herbarium leaving space for accessions for a number of years. Besides constantly using the herbarium in the determination of plants sent in for identification and of some large collections chiefly from Eastern Asia and North America members of the staff have been engaged in special work; Dr. C. E Kobuski has finished a revision of the Chinese species of Jasminum and Dr. Eva M. F. Roush is engaged i in a similar revision of the genus Eurya, Dr. I. M. Johnston is continuing his work on the Bora- , Mr. E. J. Palmer on the genus Crataegus and Mr. 4. Fadiden | is continuing the revision of the ligneous plants by H. Léveillé from Eastern Asia and the identification of collee- | tions of Chinese plants. Among visitors who have consulted | herbarium may be mentioned Dr. F. P. Metcalf of Lingnan Uni 7 versity, Canton, China, Dr. E. D. Merrill of the New York Bo : Garden, Dr. §. F. Blake of the Department of Agriculture and Dr. R. E. Woodson of the Missouri Botanical Garden. ; For study outside the Arboretum 931 specimens were sent O° loan to institutions and individuals in this country and in Europe — There have been distributed 22037 specimens to 40 institutions in the United States, Canada, Europe, Asia, Africa and Botanical exploration by members of the staff or by expeditis partly financed by the Arnold Arboretum has been carried 0” ‘both Americas, Eastern Asia and Australia. é Mrs. Susan Delano McKelvey travelled from the middle ‘ March to the end of May about 11000 miles in Texas, New Mexico Arizona, California and Oklahoma for the purpose © 7 and collecting Yucca; she obtained in addition to 350 herbaria specimens and 250 specimens of flowers and fruits in formalde solution, about 100 numbers of chromosome material and . numbers of pollen smears, 125 numbers of entomological specimens ¥ including nearly 2000 Yucca moths and 100 excellent photogra? Professor J. G. Jack spent the month of August 1931 and bs months of February and sees 1982 at He Harvard es a) NOTES : 458 Garden at Soledad, Cuba, and collected in the Garden and in the grounding country about 4500 sheets of herbarium material lesides wood specimens and seeds. Mr, E. J. Palmer with Dr. Edgar Anderson collected during the ~ month of April 1932 in the Atlantic coast region from New Jersey _ toGeorgia, Mr. Palmer paying special attention to Crataegus. Dr. H. M. Raup and Mr. E. C. Abbe started in June 1932 on a tour of botanical exploration of the Peace River region in the _ provinces of Alberta and British Columbia. — From March 1930 to May 1932 Mr. S. F. Kajewski has made extensive collections in the Solomon Islands and has collected more than 1000 numbers with numerous duplicates. Professor Albert N. a Steward and Professor C. Y. Chiao of the University of Nanking és undertook an expedition to the province of Kweichou from the tad of June to November 1931 which proved very successful; f ‘i expedition to Kwangsi planned for 1932 had to be abandoned _ account of the Sino-Japanese conflict in the spring of this year; * ‘twill probably take place next year. Professor H. H. Hu of the Tin Memorial Institute of Biology in Peiping sent his collector, 5 Mr -F. T. Wang, to Szechuan during the summer of 1931 but owmg @ ‘9 troubled conditions in that province the expedition was not « erent ul as was expected, though about 1000 numbers vs duplicates were collected; this year the botanical exploration Boxes will be continued and probably extended into — by Goerz returned in August 1931 from his collecting tour Chile “heastern Asia Minor. G. Looser, 4 resident of Santiago, ne > made a collecting tour in December 1931 to southern Co- : hich is botanically almost unexplored.—A. R. ! .—Additions to the Library during the past ted 878 volumes, 205 pamphlets and 321 photographs, making : phlets and 16,786 photo” 200 taken in the Ar- y Mr. Herbert W. in the South and — ge colored print of the Bok Singing Tower, the gift aa > fe ‘ 454 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm accuracy the many varieties of color and shading in Azaleas and “4 Rhododendrons, by Mr. C. H. L. Gebfert. ; Cards filed during the year include 1,200 in the Catalogue of books — in the Library, 350 in the Catalogue of photographs, 5,446 in the — ‘‘Card-index of New Genera, Species and Varieties published by the Gray Herbarium,” and 3,827 in the manuscript “Index of IIlustra- : tions and of New Genera, Species and Varieties of Ligneous Plants : published since 1915” prepared at the Arboretum. 4 In addition to the cards filed nearly 4,000 slips have been prepared — and filed for the printed “Catalogue of the Library,” which is now — making satisfactory progress after a long delay due to unforseen circumstances. : Five hundred and seventy volumes, including periodicals, have been bound, while about 100 smaller books and pamphlets have — , been put into pamphlet binders. " The growth of the library and the increase in the number of pe — sons using it have made necessary a large undertaking considered im former years unnecessary and disfiguring—placing book numbers on the back of the bindings. Much of this work has been done — during the past year and has proved a great saving of time and 3 effort as well as a means of locating books out of place, misplaced 4 books being inevitable with shelves open to staff and visitors. y During the year a number of research workers have made eee: the unusual facilities which the library affords, especially in Chine® literature. Dr. Franklin P. Metcalf of Lingnan University, Canton, ff China, left about the middle of September after nearly two yeas — study in the library and herbarium, preparing a Flora of Fukien, d Dr. S. F. Blake of the: United States Department of Agricultut spent some days examining the books on the floras of the :. Among other visitors using the library were Mr. Ahmed Filmy, : under the auspices of the Egyptian Legation, research workers” from Arthur D. Little, Inc., and the United Fruit Company: The number of new periodicals received during the yeaT is rather bd larger than usual, many coming in exchange for the “Journal of BE Arnold Arboretum,” the “‘Arnold Arboretum Bulletin of Misee! laneous Information,” “Contributions from the Arnold A of Harvard University,” and for herbarium specimens, some by f% and a number by purchase. They are: __ sae Acta phaenologica. Deel i, afl. 1—> ’s-Gravenhage. 1931 - ‘ Acta phytogeographica suecica. 1—> Uppsala. 1929 Acta phytotaxonomica et geobotanica. Vol. i,no- 17 1932 — oe oe ANNALES sabarienses: folia musealia. i. Szombathely. 1932 NOTES 455 1932] Arcatvos de botanica do estado de S. Paulo. Vol. i, fase. 1 — Sao Paulo. 1925 — : AUCKLAND INSTITUTE AND MUSEUM. Records. Vol. i, no. 2. [Auckland.] 1931. Brack Rock rorest. Bulletin. No. 1— Cornwall-on-the-Hud- P=. om, 1980 — __Burenos Aires—Universidad. Revista de la Facultad de agro- ; nomia y veterinaria. Tomo vii, entrega 1 — Buenos Aires. 1930 — i (ly ir ee es wee x _ Burayray, Brazil—Instituto soroterapico. Anexos das Memiérias : do Instituto de Butantan, Secgao de botanica. Vol. i, fase. be 16. Sado Paulo. 1921-22. wy Cavantmuiesta. Vol. i, fasc. 1 Barcinone. 192 Cracow, Poland—Universytet jagiellonskt. le % stituti botanici. Nr. 1-8. Cracovie, etc. 19 ; bai Diresa delle piante contro le malattie ed i cnraiitt Anno 1x, tiones In- _ Granp Canyon nature notes. Vol. v, no. 6, §-10, 12; vi, 1 > [Grand Canyon.] 1931-32 > ARVARD UNIVERSITY—Botanical muse Wea leaflets. No. 1 > Cambridge. 1932 — we Hone Kone naturalist. Vol. i, no. 1 Hong Kong. 1930 — e goa of western botany. Vol. i, no. 1 > San Francisco. : " Lrons—Soeidt botanique. Nouveau bulletin. _ Lyon. 19138. um. Botanical museum OL et Année 1, no. 14. ae VerpeE notes. Vol. ii, no. 1 > [Mancos, Colo. 1931] > Narvre. Vol. 129, no. 3244 — London. 1952 ~ Pm Tomo i — Buenos Aires. 1912 YY a Soctepape BroreRIANA. Memorias. Vol.1~ Coimbra. a 3 SOcIETA INTERNAZIONALE DI MICROBIOLOGIA—Se10né wahians, Wiles | Bclicttino. Vol. iv, lax. 1 —t Mise OP Sours Arrica—Botanical survey. Memoir 1 Pretoria. 191 ne Perse Arrigabieral and forestry college. LAE botanicae Steal Contributions KU IMPERIAL teaver 1 [Taihoku] 1930 > 5 pga, ib inite universiteit—Botanisch m museum @ en herbarium Mededeelingen. No. 1— Amsterdam. neatges on other i important accessions are: 10F, 4. H. Botanica in originali. deburgicae. 1758(37}-64 iat 12 cent. ‘Ga 6 vol.)- ee = sae on ae a 456 JOURNAL OF THE ARNOLD ARBORETUM [voL. xm GrasowskI, H. E. Flora von Ober-Schlesien. Breslau. 1843, — [Riocreux, Alfred. Collection of 87 original drawings. 1860.}— Beautifully tooled green leather binding. Purchased from the — income of the Mary Robeson Sargent fund. : TuunBere, C. P. Nova genera cgugnyh ve Upsaliae. [1781- 1801.] Hv, H. H. Enumeration of plants in Chekiang. [Peiping. 1930?]—Gift of the author. Levupers, P. E. Etwas vom coffee. [Glucksburg. 1784.] ALLET, Pierre. Le jardin du roy tres chrestien Henry IV my de France et de Navare. [Paris.] 1608. 73 plates before num- bers.—Unbound, in half green morocco book-box. Ammann, Paul. Character plantarum naturalis, 4 fine ultimo videlicet fructificatione desumtus, ac preemisso fundamento methodi genuine cognoscendi plantas, per canones et exempla digestus. - Francofurti; Lipsie, apud Nicol. Scipionem. 1685. Batscn, A. J. G. K. Versuch einer anleitung zur kenntniss t geschichte der pflanzen fiir academische vorlesungen cuituerien und mit den néthigsten abbildungen versehen. 2 theile. Halle. 1787-88. 11 plates. M’Manon, Bernard. A catalogue of garden, grass, herb, flower, tree & shrub-seeds, flower-roots, & sold by Bernard M’Mahon, seedsman. [Philadelphia? 1800?]—A photostatic copy of one E of the oldest American nursery catalogues. Gift of Mr. E. C. Vick. Ol @c.— oe 2 “ee [Ames, Oakes. Original manuscript of new species, “New com — binations and additions to the orchid flora of the Philippines prepared for fasc. v of his “Orchidaceae” in 1914-15. ais, : of the author. 4 Mvuetier, Johann Sebastian (afterwards John Mutter). tratio systematis sexualis Linnxi. London. 1777 noth : f°. 104 plates, and i engraved: title-page. (4 colored bend & inserted.) 4 —_—— copy sak the 104 plates, colored by hand, with a out title or text. ‘2 Linx, H. F. Icones plantarum selectarum Horti regii botani 3 berolinensis. Auctoribus H. F. Link et F. Otto. Fase * (in 1 vol.). Berolini. 1820-28. 60 colored plates. a3 Moricanp, Stephano. Plantae americans rariores. Gate x 1830. 10plates. - Scuteicuer, J. C. Cotslouies plantarum in Helvetia ¢ cet ay transalpina sponte nascentium. Bex. [1800.] pre The same. |Bex.] 1807. NOTES 457 Vatmont pE Bomare. Dictionnaire raisonné universel d’histoire naturelle. 5 tom. Paris. 1764. | . Rercnensacu, H. G. L. Repertorium herbarii. Dresdae et Lipsiae. 1841. (Der deutscher botaniker, 1.) Fucus, Remacleus. Plantarum omnium, quarum hodie apud pharmacopolas usus est magis frequens nomenclature juxta grecorum, latinorii, gallo., italorti, hispa & germa. sententiam. Parisiis. 1541.—Very rare, first edition of the first botanical work by the first Belgic writer. Purchased from the income of | the Mary Robeson Sargent fund. Rinaupr, Giovanni de. Il mostruosissimo mostro, nel secundo si tratta dell’herbe, & fiori—Pritzel 7634: “Liber rarissimus.”” Apparently first and unique edition. The second part of this peculiar work is of botanical interest and treats of the symbolic a signification of flowers in old literature. An early dictionary g of language of flowers—a most curious and uncommon tract, __alsoof folkloristical interest. Thislittle work isvery scarce. Gift of Mrs. Sarah C. Sears. Exrnart, Balthasar. Oekonomische pflanzenhistoire. 2 ver- besserte aufl. 12 teile (in 4 vol.). Ulm und Mémmingen. 1756, °53-62. | Mottennorrr, P. G. & O. F. Manual of Chinese bibliography; e : Vang a list of works and essays relating to China. Shanghai. 253 a ATG, _ Corrfia, Pio. Diccionario das plantas uteis do Brasil e das = | _ oticas cultivadas. Vol. i, ii. Rio de Janeiro. 1926-31. Illustr. i [BRUNFELS, Otto.] In Dioscoridis historiam herbarum certissima ___ adaptatio. Argentorati. 1548. Wacts.—Extremely rare ne } __ tion. A contemporary ownership note on the fly-leaf 98) that this belonged to Giustiniani Paoluccelli, that it was bousht he ment herbarium. | Watson, H. W., company. Union list of serials in libraries of the : +. Jan. 1925-June, t out as snter-library loans to oe nene, nited States Department of Agricu'S’ ; | 3 Ms Horticultural ] Society, Gray Herbarium, Lowthorpe ay ool, 458 JOURNAL OF THE ARNOLD ARBORETUM {vou. xm Ohio State University, Laselle Seminary, Smith College, Olmsted Brothers, University of New Hampshire, Massachusetts Agri- cultural College, Canada Department of Mines, Harvard Museum — of Comparative Zoology and other institutions. Sixty seven photo- graphs were loaned to the Royal Horticultural Society for the Coni- s fer conference of 1931. A few books were borrowed. i The library has been fortunate in being able to make up fromodd — numbers and oversheets, with the aid of a few photostatic repro- ductions, several copies of the “Silva of North America” by C. s. ) Sargent. With the exception of one copy these have been sold. At — d the same time several complete sets of the plates were made it Paris, and may be purchased from the library. They would make valuable addition to any botanical library or herbarium. The library has also sold 1,972 photographs taken by the late Dr. E. H. Wilson in China, Japan, Australasia, India and Africa, and many other photographs.—E. M. T Bibliography of the published writings of the staff and July 1, 1931-June 30, 1932 Asse, Ernst C. *hromosome numbers and the anatomy of the eee: xylem in oe Le Oleaceae. [By] Karl Sax and Ernst C. Abbe. (In ournal of | Arnold arboretum, 1932, xiii, 37-48.) Ames, Oakes The Arnold ease A; [report 1930-31]. (In Journal of the sa : arboretum, 193 193 297. ANDERSON, Et eaee The. a of ible! By Edgar Anderson and W. H. Judd. ( arboretum bulletin of aa teye' ‘rman 1931, v. 49-52. NOTES 459 The chromosome complements of Allium stellatum and dercragees bivalve. (In Annals of the Missouri botanical garden, 1931, 465-468.) Hybrid trees. (In Arnold arboretum bulletin of popular information, 1931, v, 65-68.) Species hybrids in Aquilegia. By ped Anderson and Brenhilda Schafer. (In Annals of reeeh | 1931, xlv, 63! — ivy. (In Arnold arboretum bulletin of popular information, 1932, 16, Pharapter recombination in Drosophila. oe Proceedings of the National academy of sciences, 1932, xviii, 427-42 Gea Gibiitions to omy Yai lable dec By Edgar Anderson and .Diehl. (In Journal of the Arnold te ld 1932, ~~ 213-231.) Fire! (In Arnold arboretum bulletin of popular v information 1932, vi, How big is a cell? (In Science, 1 te rai 359.) Rosa rugosa and its hybrids. By Edgar Anderson and W. H. Judd (In Arnold erecaue cpagans of ver information, 1932, vi, 29-35.) BENNETT, Elizabeth Des CuEsTER, Kameth § Lilac gattblight ae its relation to water deficiency in Aas plants. (In Proceedings of the National shade tree conference, 1931, pp- pais } A comparative study of three Phyto diseases of ‘lilac and o their pathogens. (In Journal of the Arnold arboretum, 1932, xiii, 2 Studies on the po reaction in plants. i, il. (In Journal of the ( iGoekie ee dar pre etum, 1982, xiii, 52-74, 285-296.) . Mary q Parasitism of Myxomycete plasmodia on the sporophores of Hymeno- mycetes. By F. L. Howard and Mary E. Currie. Ge Journal of the Ds oo beret, We, xiii, 270-284.) VARKER, Gran The H Hypodermataceae of conifers. (In Contributions uaa a Arnold 5 arboretum of Harvard university, 1932, no- 1.) 7 A Vem era of Berberidaceae: ee y of the chromosome number in two gen si, 28) (In Journal of the A ‘Arnold arboretum, 1931, . Chromosome umbers in the genus Tilia. (in Journal of the Arnold Drm ert 1932, xiii, 49-51.) | "Sine Toda tien, BPA Sara < Foret patho logy in Nova Scotia. (in : Renett he Deptt ln : Pigteathctony. 1022 cotton. apo ory a” ror Ehrlich and ¥. A. Wolf. (In pence toe vaca sora Pav, Jobeph Hi crane , xiii, 75-80. —-Amnold Ee laboratory x; lreport ee) @ > varies sad the Ateele arboretum, 1931, xii, 297 460 JOURNAL OF THE ARNOLD ARBORETUM Ivo. xan Milesina — on Aspidium Braunii Spenner. (In Journal of the Arnold ; arboretum, 1931, xii, 218-219.) ; An outline of forest pathology [by E. E. Hubert. Review]. (In Phylo- 1932, xxii, 393-395 “Winter browning” = pander (In Arnold arboretum bulletin of popular information, 8.) ' Howarp, F: L. . : at arasitism of Myxomycete plasmodia on the sporophores of Hymeno- — mycetes. [By] F. L. Howard and Mary E. Currie. (In Tournal of he “exgr hem etum, 1932, xiii, 270-284 84.) JACK, Bag tee Geo : Cherries and peaches. (In Arnold arboretum bulletin of popular infora- ; $e (In Arnold arboretum bulletin of popular information, 1932, _ i, 25-28. ‘ Junsberries. (In Arnold arboretum bulletin of popular information, 1932, — Some canly flowering trees. (In Arnold arboretum bulletin of popular j ey few vi, 4.) roy Jun, aaay 2 ‘eben wig Journal of the Kio guild, 1931, v, 70-81) (In Gardeners’ chronicle, 1931, xc, 469.) rnold Hebinia Uasteigs, Cn By Edgar A Anderson and W. H. Judd. (In & | arboretum bulletin of popular i repeawr brags v, 49-52.) 1982, 4 (In or Ea 1.) . Gardeners? hi icle, 1932, xci, 20 : oe at oar eR ig Gardener rm Gardeners’ chronicle, 1932, ; Hoax pees ant brids. Anderson and W. H, Judd. Arnold sievaia ride, By Eagar information, 1932, Vi, 29-85) Konusxr, Chreuas Baaaes rnold | opsis of the pre species of Jasminum. (In fant of the A 4 arboretum, 1932, xiii, 145-179.) : O’Manra, Joseph. . tt) Chromosome pairing in Yucca flaccida, (In Cytologia, 1931, i, 00-1 be Pater, Ernest Jesse | A conspectus of the ie genus Amorpha, (in Journal of the Arnold arbre 1931, xi, 157-1 18, : Notes o Spins Engelmanni. (In sesnenavabieytti : 3 Bas av : : 7 & ee minutiflorum. da Rhododendron nay mes 1 ay 3 oy oV ii Bhs is cere ie: Sax NOTES 461 ‘wold arboretum Herbarium; Fg oe 1930-31]. (In Journal of the Arnold arboretum, 1931, xu Notes on the ligneous plants pap ae ‘by Léveillé from eastern Asia. In Journal of the Arnold ingtonen ds est ‘xii, phe: 281.) i Viecteeral study of Abies pinsa [Revie of A. Barbey’s “A travers les foréts de Pinsa pe dtAndeladel ae (In J Journal of the Arnold arboretum, 1931, xii, 519-220. Zur geschichte der gehdlzeinfihrungen nach Nordamerika. (In M engender Deutschen dendrologischen gesellschaft, 1932, nr. 44, i Kalmiopeis is, a new genus of Ericaceae from northwest America. (In : aakarg of th the Arnold arboretum, 1932, xiii, 30-34.) Nomenelatorial notes on plants growing in the botani cal garden of the’ tkins institution of the Arnold arboretum at Soledad, Cienfuegos, Cuba. By F. T. Hubbard and Alfred Rehder. (In Harvard university, ical museum leaflets, 1932, no. 1, pp. 1-11. , Karl. ‘Amold arboretum Cytology laboratory; [report 1930-31]. (In Journal oo hed eamet 1931, xii, 299. umbers in the ligneous Saxifragaceae. (In J ournal of the S Rineld pilioretuon, 1931, xii, 198-206 a ring formation in Rhoeo discolor. (In Cytologia, 1931, iii, a id and mutation. (In Proceedings of the National academy of 1, xvii, 601-603.) ears technique in plant cytology. (In Stain technology, 1931, vi, Chromosome Bambers and the anatomy of the secondary xylem in a Oleaceae. [By] Karl Sax and ity C. Abbe. (In Journal of Maria _ Amold arboretum Library fre paper 1930-31]. (In Journal of the Arnold Pe ask. coe s waters oben | ‘_ EK. Staff of the Arnold Arboretum, 1932-33 Oakes Aura, A.M., Arnold Professor of Botany, Supervisor. Joux Gronce Jac, Assistant Professor of of Dendrology. ‘urator of the Herbarium. Josurn H. Favis, Pb.D., Profesor of Forest Pathology. . &.D., Profesor of Plant Anatomy. of Cytology. | , Herbarium. _ Eanust J. Paiaren, Colle Reearch Bra M. Fuso Rousn, Ph.D., Assistant in the Herbarium. Kexwern 8. Con D.. Research Assistant, Phytopathology. : Comeren, Ph.D. inant D. Danknn, Ph.D), Assistant in Phytopathology. Lite Denwex, M ERRATA AND ADDENDA Page 31, line 22 for relict read relic “Tine 18 from below for relicts read relics 75, line 8 for Sporadically read Sporadical 83, line 3 from below for Hedycarya read Hedycaria 4 112, line 2 from below for CHAMBEYRENIA read CHAMBEYRONIA : “156, line 17 insert under Jasminum lanceolarium Roxb.: gee ae A Pritzel in Bot. Jahrb. xxrx. 422 (1900)— ; Zant a oldie allay, Chung in Mem, Sci. Soe. China, 4 1. Szechuan: Wine ‘nahn. Bock & Rosthorn, — no. 19, July 1891 (Herb. Oslo and Berlin; photo. inA. A). — 4 “157, line 1 for 6954 read 6964 . 158, after line 11 from Below insert under Jasminum lanceo- one var. puberulum Panpeuults . a Teta Gore Bok Prt : vie 406 (1910) Syn. no. Hupeh: Monte “Triora,” alt. 1950 m., C. St ; nos. 1221 and 1221a, Sept. 1907 (Herb. Biondi, Florence; x photo. in A. A.). ac “195, line 10 from below for In read If 302, line 6 for 1928 read 1898 $29, line 5 from below for 1351 read 1531 i ae Pat reer er wees es and represented by a cba se Supegne typied ~ * INDEX Synonyms are printed in italics; new names in bold-face type. BE Asse, Ernst C. and Sax, Kart, _ Chromosome numbers and the anat- omy of the secondary xylem in the : Achyranthes aspera, 81 AMES » Oakes, A new genus of the ida from the New Hebrides, os | Contribution tothe fora ofthe ft pe ihe % . Pz one 1929, 127 Poe. secondary xylem in eng Lae poeme numbers of t ao te 1931-1932, 458 Arnold Arboretum during the fiscal year ended June 30, 1932, The, 447 Arnold Arboret e ition to North- western China and Northeastern Tibet, Enumeration of the ligneous plants collected by J. F. Rock on the, Arnold Arboretum, New species, varie- ties and combinations from the her- barium and the collections of the, Arnold Arboretum, 1932-33, The Staff — nidus, 1 466 Beaumontia grandiflora, Diaxylary la- ticiferous cells of, 35, pl. 41 Beilschmiedia sp., 84 Bibliography of the published writings of the staff and students of the Arnold Arboretum, July 1, 1931- June 30, 1932, 458 Bischofia javanica, 91 BiakeLy, W. F., A previously unde- scribed variety of Loranthus vitel- linus F. v. Muell. from North Queensland, 298 Blechnum gibbum, 122 refeldia Bulbophyllum sp., Calanthe sane eb 139 Calpidia excelsa, 81 ae hag Milnei, 8 oe Sospeus lonchophorum, soe ampsis — paniculata, 95 yeulatus, 116 — Ganatince latifolia, 117 Ceratostylis kaniensis, 138 — subulata, 109, 138 Cerbera manghas, 18 Chambeyronia sp., 113, fig. a2 Cheilanthes hirsuta, 122 Chenopodium ? suffruticosum, 338 JOURNAL OF THE ARNOLD ARBORETUM [vou. xm Cuester, Kennetu S., A comparative study of three Phytophthora di- seases of Lilac and of their pathogens, _ 232, diag., pl. 46, 47 ——.,, St peat on the precipitin Te selon in plants. I. The of the phan precipitin reactnn, 52, pl. 42 , Studies on a precipitin re- 3 aition in plants. Preliminary — report on the pats fe the “‘ Normal : precipitin reaction,’ 28 Chinese plants, Notes on some, 297 Chinese species of Jasminum, Sree of, 145 Chomelia banksiana, 4 Shaeastes numbers and the anat- — omy of the secondary xylem in the — Cleidion angustifolium, 93 ° — Vieillardii acutifolium, 93° Clematis aethusifolia, 404 — brevicaudata, 404 — glauca akebioides pcanth sot tica obtusiuscula, 404 46, 4 Cinco wo he tm INDEX 467 Kajewski in 1928 and 1929, 1, 81, Curr, Mary E. and Howarp, 3 figs., pl. 43 Frank L., — of myxomycete - Contribution to the flora of the New plasmodia on the sporophores of Hebrides and Santa Cruz Islands, eyincnon ail 270, 2 figs. pl. Orehids collected by S. F. Kajewski in 1928 and 1929, 127 Cyathea lacininta, 118 Contributions to the Tradescantia — leucolepis, 119 en peabilem, 213, 3 figs., pl. 45 — medullaris, 119 Ls , E. > Pteridophyta, 118 Cycas circinalis, 118 — Cordia Myxa, Cyclophorus acrostichoides, 124 _ —subcordata Cypholophus macrocep! Cornus, Bees studies of, 410, Cyphophoenix sp. nov.?, 113, fig. 1: pl. 53 3,4 _ Corymborchis veratrifolia, 109, 127 Cyrtandra aneiteensis, 25 Cotoneaster Bodinieri, 309 ilk — Kajewskii, 25 Cytological laboratory, The Arnol Arboretum during the fiseal year ended June 30, hie La . »3 Cytological mechanism of crossing _ Couthovia neo-ebudica, 23 over, 180, 11 diag., fig., pl. 44 _ Crataegus problem, The, 342 Cytological studies of Coenen 410, gyt pl. 53 Dalbergia Cavaleriei, 330 — macrophyllum, 133 — separatum, 133 — yanikorense, 134 at (§ spew) sp.; _ Reda eee 8 fagi (Baer) Dougl., the — Cavaleriei, 323 _ r in a menacing disease 3, 327 — cinerascens, the oceurrence in the concinnum, 328 _— Esquirolii, 327 _— gangeticum, 328 ary E. and Howarv, — Griffithianum, 327 L, Parasitism of myxomy- — gyroides, 327 2 on fungous mycelia, — 468 Dianella ensifolia, 112 Diaxylary laticiferous cells of Beau- montia grandiflora, 35, pl. 41 Dreat, D. G. and ANpERson, Eve Contributions to the Tradescantia problem, 213, 3 figs., pl. 45 Dioscorea bulbifera, 111 — nummularia, 111 Diospyros Chaffanjoni, 302 Dipteris conjugata, 1: Docynia Delavayi, 309 — docynioides, 310 — rufifolia, 310 Dolicholobium aneityense, 2 Donax arundastrum, 110 Elatostema macrophyllum, 106 Elytranthe ; Endiandra ineateaiiita. 84 Enumeration of the ligneous plants collected by J. F. Rock on the Ar- i Kajewskii, 109, 135 » 135 Ericaceae from iri ‘America, -@ new genus of, 30, pl. 40 Eriobotrya Brackloi crichophl, 307 — Cavaleriei, 307 — Brackloi, 308 —_ — pseudo-Raphiolepi, 308. Euphorbia obliqua, 90 Excoecaria Agallocha, 95 Fagara volubilis, 463 JOURNAL OF THE ARNOLD ARBORETUM — sp. nov.?, 83 Hedysarum multij ‘emign Delaneh 99 Fagara volubilis pubescens, 463 Fagraea sp., 22 Faradaya neo-ebudica, 28 Ficus acrorrhyncha, 98 — aspera, 101 — austro-caledonica, 99 — casearia, 106 — ciliata, 101 — copiosa, 102 00 oo Kaiewskil, 103 — Moseleyana, 106 — neo-ebudarum, 97 — obliqua, 101 —— philippinensts sessilis, 96 0) 0 Geblera suffruticosa, 338 Geitonoplesium cymosum, i Geniostoma rupestris, 22. Gleichenia Brackenridgei, 124 collected by S. ¥ ae and 1929, 1, $1, 3 figs. Vissksisae vernalis tomentella, Hedycaria neo-ebudica, 83 ugum, 405 tans, 26 , clavata, 279 iain The Arnold Arboretum during the fiscal year ended June 30, ze Gramineae, 116 » 94 : : cete plasmodia on fungous mycelia, & ae, pl. 54 : PEM OT Parasitism of myxomycete , 270, 2 figs., pl. 48, INDEX 469 Jasminum, Synopsis of the Chinese 68 —Beesianum X officinale grandi- florum, 162 — cee od 171 — bignoniaceum, 150 — Blinii, 163 — Bodiniert, 159 | a pewter 165 470 Jasminum odoratum, 171 — officinale, 160 grandifiorum, 161 —_—— Pee hey 161 iflorum < Beesianum, 162 — pachyphyll um, min — quadrifolium, 172 — quinqueflorum, 172 64 — S. F., Contribution to ‘te ora of the New Hebrides, plants z hei arasek in Bee and 1929, 1, 81, 3 figs., pl. 43 and 1929 by, 127 Kajewskia aneityensis, 113, fig. 2 _ JOURNAL OF THE ARNOLD ARBORETUM i Contribution to the flora of the New Hebrides and Santa Cruz Islands, Orchids collected in 1928 Kalmiopsis, a new genus of Bricacese from northwest America, 30, pl.