JOURNAL
OE THE ARNOLD ARBORETUM HARVARD UNIVERSITY
ALFRED REHDER EDITOR
JOSEPH H. FAULL anp CLARENCE E. KOBUSKI
ASSOCIATE EDITORS
VOLUME XV
JAMAICA PLAIN, MASS. 193-4.
Reprinted with the permission of the Arnold Arboretum of Harvard University
KRAUS REPRINT CORPORATION York
1968
DATES OF ISSUE
(pp. 1-89, pl. 82-86) issued January 24, 1934.
(pp. 91-161, pl. 87-90) issued April 20, 1934.
(pp. 163-266, pl. 91-109) issued July 16, 1934.
(pp. 267-389, pl. 110-118) issued October 25, 1934.
Printed in U.S.A.
TABLE OF CONTENTS
NOTES ON THE Licneous PLANTS DESCRIBED BY LEVEILLE FROM
EASTERN ASIA. By Alfred Render a2 seas scaaenas dw iaes sass 1 SPECIATION IN UvuLaria. With plates 82 and 83 and 5 text figures.
By Edgar Anderson and Thomas W. Whitaker ............00005 28 A QUANTITATIVE COMPARISON OF SPECIFIC AND GENERIC 2
NCES IN THE BETULACEAE. With one text figure. By Edga
Anderson ta! eel gy a OO: 3): a a Sie ae |. cone 43 THE Bio.ocy oF MiLEsIAN Rusts. With plates 84-86. By J. H. Faull 50 A REMARKABLE SPRUCE Rust, PERIDERMIUM PARKSIANUM, nl. sp
By Th Sony ra ee a re eee eer ears sarees | ae eee 86 Notes: THe BEACH PLUM IN MicHIGAN. By Ernest J. siege —
QUERCUS ELLIPSOIDALIS IN Missouri. By Ernest J. ‘Po CMON. Soc 88 Notes ON THE LIGNEOUS PLANTS DESCRIBED BY Lalenit FROM
TOAST ERM isiA: oY IL Ped Chder 3) vccs eae tee abee eas late 91 STUDIES IN THE GENUS FRAXINUS. A PRELIMINARY KEY TO WINTER
TWIiGs FOR THE SECTIONS MELIOIDES AND BUMELIOIDES. With
plates 87-89 and 2 text figures. By C. M. Whelden .............. 118
ES » PLANTS OF OKLAHOMA. By Er J. Palme 127
CHROMOSOME CONSTITUTIO RTAIN MonocotyLepcns. Witl
text figures. By Thomas W. Whit seid CTE OTe | eer ee ee 135 THE OCCURRENCE OF TUMORS ON CERTAIN NICOTIANA H’7prips. With
plate 90 and 1 text figure. By as W. PP RUURET C4) cwecnea cen 144 Notes: BListerR Rust oF PINUS eee age (Review). By J.
Faull, — Wrumeyver’s “THe Genus D1aportHe NifscHKE AND
ITS SEGREGATES” ” (Review). By J. H Paes ere bas eA 154 Tue Hosts, Lire History AND CONTROL OF THE CEDAR: APPLE RUST
FUNGUS, GYMNOSPORANGIUM JUNIPERI-VIRGINIANAE SCHW. With
two text figures ae Be 91-98. By Ivan H. - SOOO rasicsewas 163 THE CAMBIUM AND ITS DERIVATIVE TISSUES . 1X. STRUCTURAL
VARIABILITY IN site REpWwoop, ee. ee AND ITS
SIGNIFICANCE IN E IDENTIFICATION OF Foss1L Woops. With
one text figure aad cee 99-106. By I. W. Bailey ind Anna F.
DH ain ee Pe Ae ick ee EE OR EAS ER ee eee 250 Care ouke OF ie Manet With plates 107 anid 108. By
Reate Sak and JM 2 Beal or occurs Vee a oo as SA RO ERO 255 THE SEVENTH one OF THE RELIQUIAE FARLOWLANAE. DIs-
TRIBUTED BY THE FARLOW HERBARIUM OF HARVARD UNIVERSITY.
With plate 109. By David H. Limder ......00c 0c ccc ducnvcacsees 259 NOMENCLATURAL PRIORITY IN THE UREDINALES. By J. C. Arthur .. 263 Notes: TREES OF THE SOUTHEASTERN STATES. (Review.) By E. J
| iS ee eee eo tae: a ar eres | Mr tee ae errr 266
OTES ON T LicgNEous PLANTS DESCRIBED By LE\VEILLE FROM
EASTERN ASIA: BY Al[ree RONGEe 666s ees ede tas odes es vanes 267 THE CAMBIUM AND ITS DERIVATIVE TissuEs. No. X. STRUCTURE,
OpTICAL PROPERTIES AND CHEMICAL COMPOSITION OF
CALLED MippLe LAMELLA. With two text figures ancl plates 110-
113. By Thomas Kerr and I. W. Batley ..... 0.0. cece cece eens 327
iv TABLE OF CONTENTS
A VERIFICATION OF THE OCCURRENCE OF YUCCA WHIPPLEI IN ARI-
zONA. With plates 114-117. By Susan D. McKelvey ............ 350 A KARyYO-SYSTEMATIC STUDY OF penne With one text figure. By
Thomas 1 AW RUGROY 660 6665.6 ce Fi a RS8 ON 44 READ ea ROR TAREE TENA 353 A CULTURE CHAMBER FOR THE STUDY OF MycorruHizAE. With two
text cell and plate 118. By Me Eh. GUGREM GaGa Vices ea eet eee es 358 THE ARNOLD ARBORETUM DURING THE FISCAL YEAR ENDED JUNE i
934; The Arboretum; The Pathological Laboratory; The Cyt
genetic Laboratory ; e Herbarium; e Libr te “Bibliography
of the published writings of the staff and erie July 1
TAI I FPO 8s, G0 aad 5 LAS REDE S ee DLO RORY ORT E es 366 STAFF OF THE ARNOLD ARBORETUM 1933-1934 ....... 0.0. c eee eee ee 376 ERGATA: ANG: ADUEWDA. 6.6.0559:.0 20004 Kee 6rd es Cw tere Eee 377
TNDEX: ee eege epee be eee eaters Gisee Few Shae Ga ee a arte eee ae aie are
JOURNAL
OF THE
ARNOLD ARBORETUM
VOLUME XV. JANUARY, 1934 NuMBER 1
NOTES ON THE LIGNEOUS PLANTS DESCRIBED BY LEVEILLE FROM EASTERN ASIA:
ALFRED REHDER
CELASTRACEAE (Supplement)
Tripterygium hypoglaucum (Lévl.) Hutchinson in Kew Bull. Misc, Inform. 1917, p. 101. — Léveillé, China Rev. Ann. 1916, p. 23 (Msc.). — Loesener in Ber. Deutsch, Bot. Ges. La, p, 10 (1932).
Add as synonym:
steal Franchetiana Léveillé in Bull. Géo;r. Bot. xxv. 26
1 xcl. descript. fructus; Cat. Pl. Yun-Nan 227 (1917). — Synon. nov.
CuHInA. Kweichou: Brousses des montagnes a Siao-ou-long, 2600 m., E. E. Maire, June 1912, “arbuste grimpant i: long rameaux, fl. blanches” (holotype of Microrhamnus Franchetian1, excl. fruiting branch; photo. and merotype in A. A.).
In revising Léveillé’s Rhamnaceae I found that his Microrhamnus Franchetiana belongs only partly to the Rhamnaceae, while the flower- ing branch upon which the description chiefly is based, is to be referred to Tripterygium hypoglaucum. The leafless fruiting bi'anch represents Berchemia Giraldiana Schneid.
STAPHYLEACEAE
Staphylea holocarpa Hemsley in Kew Bull. Misc. Inform. 1895, p. 15. — Léveillé, Cat. Pl. Yun-Nan, 267 (1917).
Xanthoceras enkianthiflora Léveillé in Fedde, Rep. Spec. Nov. x11. 534 (1913); Fl. Kouy-Tchéou, 384 (1915) “Zanthoceras.” — Synon. nov.
Tecoma Cavaleriei Léveillé, Fl. Kouy-Tchéou, 50 (1914). — Synon. nov.
1Continued from Vol. x1v. 252; for preceding parts see Vol. x. 108-132, 184-196; Vol. xm. 275-281, and Vol. XII. 299-332.
2 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
Cutna. Kweichou: Tou-chan, E. Bodinier, April 1900 (ex Léveillé; holotype of Xanthoceras enkianthiflora) ;/ Gan-chouen, CF Cavalerie, no. 3913, April 1912 (sub Zanthoceras enkianthiflora in F. Kouy-Tchéou, I. c.; duplicate in A. A.); nord de Gan-chouen, J. Cava- lerie, no. 3913, Nay 1912 (holotype of Tecoma Cavaleriei; merotype in
The type of Xanthoceras enkianthiflora is missing from the Léveillé herbarium, but I have before me the additional specimen cited in Flore de Kouy-Tchéou and an unnumbered specimen collected by Cavalerie with a label not in Léveillé’s handwriting, giving a different locality but the same date. The flowers, however, instead of being 3-4 times larger than Nanthoceras sorbifolia as stated in Léveillé’s descrip- tion, are really smaller, and I cannot reconcile this part of the descrip- tion with the specimens before me. The type of Tecoma Cavaleriei is in fruit and Léveillé apparently took the thin capsule, splitting at the apex into several lobes, for a flower.
Euscaphis japonica (Thbg.) Dippel, Laubholzk. 480 (1892).
Evodia Chaffanjoni Léveillé in Fedde, Rep. Spec. Nov. xurr. 265 (1914); Fl. Kouy-Tchéou, 375 (1915). — aren Ov. Euscaphis staphyleoides Siebold & Zuccarini, a 1. 125, t. 67 835). — Léveillé, Cat. Pl. Yun-Nan, 267 ae
Cu1Ina. Kweichou: environs de Kouy-yang, fn du Collége, J. Chaffanjon in herb. Bodinier, no. 2263, April 15, 1898 (holotype of Evodia Chaffanjoni; merotype in A. A.).
ICACINACEAE
Iodes ovalis Blume, Bijdr. 30 (1825). — Baillon in De Candolle,
Prodr. xvir. 22 (1873 lodes wvitiginea (Hee.) Hemsley in Jour. Linn. Soc. xxii. 115 (1886).
Vitis chrysobotrys oo be Vaniot in Fedde, Rep. Spec. Nov. 111 350 (1907). — Synon. Vitis fae Léveillé, Fl. "Kouy- Tchéou, 28 (1914), pro parte, quoad Esquirol, nos. 508 and 868; non Léveillé (1907). Sabia edulis Léveillé, Fl. Kouy- gui 379 Seagal nomen; pro parte, quoad Cavalerie, nos. 2033, 3904. — Syn Hernandia sinensis ee FI. aes Tchéou, "379. (1915), pro synon. Sabie edu CHINA. Kweic he ou: without precise locality, J. Esquirol, no. 508, June 1905, “fleurs verdatres” (holotype of Vitis chrysobotrys; photo. in A. A.) ; prés fleuve Houa-kiang, J. Cavalerie, no. 2033, June 6, 1904 (syntype of Sabia edulis; photo. in A. A.); without precise lo- cality, J. Cavalerie, no. 3904, in 1912 (syntype of Sabia edulis; photo.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 3
in A. A.); without precise locality, J. Esquirol, no. 868 (sub “Sabia edulis | Vitis chrysobotrys Lévl. & Vant.]” in herb. Léveillé; sub Vitis Seguini in Fl. Kouy-Tchéou; dupl. in A. A.).
Sabia edulis was published as ‘‘comb. nov.” without description, only with the citation Hernandia sinensis Lévl. as a synonym, but that name is only a manuscript name and appears on the sheet o! Cavalerie’s no. 2033 with a Latin description and a pencil-note in Léveillé’s hand- writing “non publ.” The name Sadia edulis is not found on this sheet nor on the sheets of the other two numbers. Vitis chi'ysobotrys is re- ferred by Léveillé to Vitis Seguini as a synonym in his Flore de Kouy- Tchéou (p. 28), but in his herbarium it is given under Esquirol, no. 868, as a synonym of Sabia edulis.
I am unable to separate J. vitiginea clearly from J. ovalis Bl. which is apparently a very variable species of wide distribut on. By Gagne- pain (in Lecomte, Fl. Gén. Indochine, 1. 845. 1911) J. witiginea is made a variety of J. ovalis, but the distinguishing characters given—plant less pubescent, leaves very pale beneath—do not seein to distinguish the plant from the type which is described by Baillon (1. c.), as with “foliis . . . parce villosis’” in contrast to the tomentose var. Miquelit. I have seen no spontaneous specimen from Java except one from a plant cultivated in Buitenzorg (Hochreutiner, Pl. Bcegor. Exsicc. no. “27) which seems referable to var. Miquelii. I prefer, therefore, for the present to accept J. ovalis Bl. in the wide sense of most modern authors.
As the author of J. vitiginea usually Hance is cited, but he did not publish the combination. In 1873 (Jour. Bot. x1. 2€6) he described Erythrostaphyle vitiginea as a new genus and new species of Verbena- ceae. In the following year he published a note (Jou. Bot. xu, 184) saying that Professor Oliver had pointed out to him that the plant described under the above name is a species of Jodes closely allied to the Sumatran J. tomentella Miq. Hance did not make a new combina- tion and leaves us in doubt, whether he considered it referable to J. tomentella or a new species. Hemsley seems to have been the first to publish the combination /. vitiginea crediting it to Hence; it does not appear in Index Kewensis except as a reference from Erythrostaphyle vitiginea, but under Jodes there is no /. vitiginea.
Iodes Seguini (Lévl.), comb. nov.
Vitis Seguini Léveillé in Fedde, Rep. Spec. Nov. iv. 331 (1907); FI. Kouy- ee 28 (1914), pro parte, synon. et speciminibus Esqui- rolii exclusis.
Sabia edulis Léveillé, Fl. revedge nc ee eta pro parte, quoad specim. Cavalerie, no. 3932. — Syn
4 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
~ het hai Gagnepain in Lecomte, Not. Syst. 1. 201 (1910). — eillé, Fl. Kouy-Tchéou, 291 (1914). — Synon. nov. ee ee ae se levitestis Handel-Mazzetti in Anz. Akad. Wiss. Wien, sien p.°149 (Pl. Nov. Sin. Forts. 12, p. 6) (1921). — Synon.
CHINA. Ky weichou: environs de Hoang-ko-chan, rocailles de la cascade, J. Seguin in herb. Bodinier, no. 2614, March 29, 1899, “liane s’attachant aux rochers comme un lierre” (holotype of Vitis Seguini [Hb. Léveillé; photo. in A. A.| and of Jodes rugosa |Hb. Paris]; route de Tchen-fong a Hin-y-fou, rochers, J. Cavalerie, no. 3932, June 1912, “liane aux fruits roses mangeables”; (syntype of Sabia edulis; photo. in A. A.); ad viam Tschenning—Huang-tsauba in convalle flum. Hoadjiau-ho infra vicum Tingdaoyin, Handel-Mazzetti, no. 10374, June 20, 1917, “fr. rubri, sapore dulci sed paululum amygdalos amaros admonente” (syntype of J. vitiginea var. levitestis; isotype in A. A.).
Vitis Seguini and lodes rugosa are based on duplicates of the same collection, but Léveillé cites “Seguin, no. 2614” while Gagnepain cites “no. 2614 (Bodinier),’ and in his Flore de Kouy-Tchéou Léveillé enu- merates both species and cites the same number with different collectors under each. Cavalerie no. 3932 agrees with J. Seguini, but the leaves are not deeply cordate and not as strongly veined though distinctly reticulate above which distinguishes them from J. ovalis. Handel- Mazzetti’s no. 10374 is similar, but the stems are only sparingly verru- culose and the leaves are as in Cavalerie’s specimen rounded or sub- cordate at base, not deeply cordate as in the type of J. Seguini. The fruits of both specimens are identical and are described as edible and rose-colored or red; they are oblong, 1.8-2 cm. long and about 12 mm. wide, the putamen with a thin and smooth testa. All the specimens cited above agree in the more or less verruculose branches, in the leaves being more strongly veined and reticulate above and in the short in- florescence branched near or at the base, characters by which this species can be readily distinguished from /. ovalis.
ACERACEAE
Dipteronia sinensis Oliver in Hooker, Ic. Pl. xxrx. t. 1898 (1889). Acer Dielsit Léveillé in Fedde, Rep. ae Nov. x. 432 (1912); FI Kouy-Tchéou, 383 (1915). — Synon. Cuina. Kweichou: Pin-fa, J. Cavalerie, no. 1097, June 1903 (holotype of Acer Dielsii; fragments in A. A.). Acer Miyabei Maximowicz in Bull. Acad. Sci. St. Pétersb. xxxm. 485 (1888); in Mél. Biol. x11. 725 (1888). — Koidzumi in Jour. Coll. Sci. Tokyo, xxxit, art. 1, p. 57, t. 31 (1911).
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 5 Acer Hayatae Léveillé & Vaniot in Bull. Soc. Bot. France, L111. 590 9
06).
Japan. Hokkaido: near Nanai, U. Faurie, no. 6086, June 1904 (holotype; photo. and isotype in A. A.).
Acer Hayatae has already been identified with A. Miyabei by Koid- zumi (I. c.).
Acer pictum Thunberg, Fl. Jap. 161 (1784).
Acer Hayatae var. yap aaa & Vaniot in Bull. Soc. Bot. France, Lit. 590 (1906). — on. ;
Japan. Hondo: near sd U. Faurie, no. 5020; May 20, 1902 (holotype; photo. in A. A.)
Acer pictum f. tricuspis Rehder in Mitteil. Deutsch. Dendr. Ges. xxi. 258 (1913).
Acer Bodinieri Léveillé in Fedde, Rep. Spec. Nov. x. 433 (1912) ; Fl. Kouy-Tchéou, 382 (1915). — Synon. nov.
Cuina. Kweichou: Tsin-gay, Kao-po, J. Cavalerie, no. 1162, July 1903 (holotype; photo. in A. A.).
The specimen differs from the type of this form in the larger leaves with rounded base and in the wings of the fruit spreading at a right or even acute angle; in the rounded base of the leaves it comes near Harry Smith, no. 6551, from Shansi, but that plant has much smaller leaves and wings spreading at a wide angle. In shape the leaves re- semble those of A. Wilsoni Rehd., but this has a paniculate inflorescence and 3-nerved leaves with different venation.
Acer caudatum Wall. var. ukurunduense (Maxim.) Rehder in Sargent, Trees & Shrubs, 1. 164, t. 82 (1905); 11. 26 (1907).
Acer lasiocarpum Léveillé & Vaniot in Bull. Soc. Bot. France, L111. 591 (1906). Acer spicatum wee var. eine ig Maximowicz, in Mém. Div. , Pétersb. 1x. 65 (Prim. Fl. Amur.) (1859). — Koidzumi in cae Coll. Sei Tokyo, xxxir. art. 1, p. 31 C1911).
Japan. Hondo: in sylvis Takayu, U. Faurie, no. 6102, June 1904 (syntype of A. lasiocarpum, isotype in A. A.); in Jizogatake, U. Faurie, no. 5467, July 1903 (syntype of A. lasiocarpum; photo. in A. A.); in Hakkoda, U. Faurie, no. 6100, Aug. 1904 (syntype of A. lasiocarpum, isotype in A. A.); in sylvis Ochiai, U. Faurie, no. 6101, Sept. 1904 (syntype of A. lasiocarpum; isotype in A. A.).
Acer lasiocarpum was first identified with A. caudatum var. ukurun- duense by the writer (op. cit. 11. 26).
Acer oblongum Wallich apud De Candolle, Prodr. 1. 593 (1824).
Acer Paxii Franch. var. integrifolia Léveillé, Fl. Kouy-Tchéou, 383 (1915), nomen. — Synon. nov
6 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
CuIna. Kweichou: route de Pin-yue 4 Kouy-yang, L. Mar- tin in hb. Bodinier, no. 2618, May 18, 1899, “arbre” (holotype of A. Paxti var. integrifolia; photo. in A. A.).
I have not been able to find an earlier publication of Léveillé’s name except that in Flore de Kouy-Tchéou where it appears as a nomen nudum.
Acer oblongum var. biauritum W. W. Smith in Notes Bot. Gard. Edinb. vii. 329 (1915). — Metcalf in Lingnan Sci. Jour. x1. 195 (1932).
Acer oblongum var. erythrocarpum Léveillé, Cat. Pl. Yun-Nan, 252 1917).
Cuina. Yunnan: valleys in the mountains of the Chungtien plateau, alt. 10000 ft., G. Forrest, no. 10644, July 1913 (holotype of A. oblongum var. biauritum; isotype in A. A.); brousse de Mahong, alt. 2800 m., FE. E. Maire, May |1910-12|, “petite arbre, fruit rougeatre” (holotype of A. oblongum var. erythrocarpum, merotype in
The type of A. oblongum var. biauritum has mature fruits of light grayish yellow color, while the young fruits of Maire’s specimen are purplish and so are the fruits of Simeon Ten, no. 471, the only other specimen with young fruits in this herbarium.
Acer coriaceifolium Léveillé in Fedde, Rep. Spec. Nov. x. 433 (1912); Fl. Kouy-Tchéou, 383 (1915); “coriacetfolia” — Metcalf in Lingnan Sci. Jour. x1. 198 (1932).
Cu1naA. Kweichou: Pin-fa, bois, J. Cavalerie, no. 3100, July 2, 1907 (holotype; merotype in A. A.).
By Fang (in Contrib. Biol. Lab. Sci. Soc. China, vi. 167, 1932) the species was referred as a synonym to A. oblongum Wall., but it is easily distinguished by the green tomentulose under side of the oblong- lanceolate leaves. It is, as Metcalf points out more closely related to A, lanceolatum Molliard than to A. oblongum.
Acer Fabri Hance var. rubrocarpum Metcalf in Lingnan Sci. Jour. XI. 206 (1932).
Acer Fargesti Franchet apud Rehder in Sargent, Trees & Shrubs, I. 180 (1905). Acer Prainii Léveillé in Fedde, Rep. Spec. Nov. x. 432 (1912).
Cutna. Kweichou: Pin-fa, J. Cavalerie, nos. 951, March 1903, and 1048, June 3, 1903 (two syntypes of A. Prainii; dupl. and photo. of no. 951 and photo. of no, 1048 in A. A.); Ma-jo, J. Cava- lerie, July 1903 (ex Léveillé; syntype of A. Prainiz).
After having seen a fragment of Hance’s type of A. Fabri I am con-
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 7
vinced that it does not belong to A. laevigatum Wall. where I had placed it formerly (in Sargent, Trees & Shrubs, 1. 180), but that it is conspecific with A. Fargesi which represents a red-fruited form of it. Acer Fargesii was based by the writer exclusively on red-fruited or red-flowered specimens; of the Farges specimens named by Franchet A. Fargesi in herb. the writer saw and cited only a flowering specimen with the flowers and young leaves deep red which obviously belongs to the red-fruited form; the specimen with whitish fruits also named A. Fargesi by Franchet was not cited and did not enter in the description, therefore A. Fargesi must be considered a red-fruited form of A. Fabri and becomes a synonym of A. Fabri var. rubrocarpum and need not be cited in part under the type. Acer Prainii was first identified with A. Fabri var. rubrocarpum by Metcalf.
Acer Davidi Franchet in Nouv. Arch. Mus. Paris, sér. 2, vir. 212 (Pl. David. mm. 30) (1884). — Léveillé, Cat. Pl. Yun-Nan, 252 (1917). — Fang in Contrib. Biol. Lab. Sci. Soc. China, vir. 174 (1932). — Rehder in Jour. Arnold Arb. xtv. 213 (1933).
Acer Cavaleriet Léveillé in Fedde, Rep. Spec. Nov. x. 432 (1911); Fl. Kouy- Tchéou, 383 (1915).
CHINA. Kwei ch ou: Ma-jo, J. Cavalerie, no. 3345, Sept. 1908 (holotype of A. Cavaleriei; merotype in A. A.)
Acer Cavaleriei was first identified with A. Davidi by Fang (I. c.).
Acer crataegifolium Siebold & Zuccarini in Abh. Akad. Muench. Iv. pt. ii, 155 (Fl. Jap. Fam. Nat. 1. 47) (1845). — Koidzumi in Jour. Coll. Sci. Tokyo, xxxm. art. 1, p. 13 (1911).
Acer cucullobracteatum Léveillé in Bull. Soc. Bot. France, L111. 590 1906
Japan. Hondo: in sylvis Hayachine, U. Faurie, no. 6732, June 5, 1905 (holotype of A. cucullobracteatum; photo. in A. A.).
Acer cucullobracteatum was first identified with A. crataegifolium by Koidzumi (1. c.). Acer crataegifolius [sic] as enumerated by Léveillé in his Flore de Kouy-tchéou (p. 383) and in his Catalogue des plantes de Yun-Nan (p. 252) refers probably to A. Davidi of which specimens were found in Herb. Léveillé in the cover of A. crataegifolium.
Acer Tschonoskii Maximowicz in Bull. Acad. Sci. St. Pétersb. xxxt. 24 (1886); in Mél. Biol. x11. 432 (1886). — Rehder in Sargent, Trees & Shrubs, 11. 26 (1907). — Koidzumi in Jour. Coll. Sci. Tokyo, xxxm. art. 1, p. 24 (1911).
ees het lac Léveillé & Vaniot in Bull. Soc. Bot. France, Li. 592 (1906). JapAN. Hokkaido: in alpibus Tokachi, 1500 m., in lacunis
8 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
secus nives, U. Faurie, no, 6729, July 1905 (holotype; isotype in A. A.).
Acer pellucidobracteatum: was first identified with A. Tschonoskii by the writer (1. c.).
Acer argutum Maximowicz in Bull. Acad. Sci. St. Pétersb. x11. 226 (1867); in Mél. Biol. vr. 368 (1867).
Acer palmatum var. a Léveillé in Bull. Soc. Bot. France, Litt. 592 (1906
Japan. Hondo: in ‘sipibus Nikko, U. Faurie, no. 2309, May 27, 1898 (holotype of A. palmatum var. plicatum; photo. in A. A.).
The specimen of A. palmatum var. plicatum in the Léveillé herbarium which consists of a small branch with pistillate flowers does not bear on the label the name given by Léveillé.
Acer Negundo Linnaeus, Spec. Pl. 1056 (1753). — Rehder in Sar- gent, Trees & Shrubs, 11. 26 (1907)
Acer Fauriet Léveillé in Bull. Soc. Bot. France, tii. 590 (1906).
Japan. Hokkaido: Sapparo, dans le jardin de M. Miyabe, U. Faurie, no. 6084 bis, July 1905 (holotype of A. Fauriei; photo. in A. As):
The same species is represented in this herbarium by a specimen of Faurie’s no. 6084, collected September 1904 in “Yezo in hortis culta.” Acer Fauriei was first identified with A. Negundo by the writer (1. c.).
SAPINDACEAE
Eurycorymbus Cavaleriei Rehder & Handel-Mazzetti.!
Rhus Cazvalerici Léveillé in Fedde, Rep. Spec. Nov. x. 474 (1912); Fl. Kouy-Tchéou, 411 (1915).
Eurycorymbus austrosinensis Handel-Mazzetti in Anz. Akad, Wiss. Wien, 1922, p. 104 (Pl. Nov. Sin. Forts. 16, p. 4) (1922).
Cuina. Kweichou: Pin-fa, J. Cavalerie, no. 1094, June 23, 1903, “petit arbre; il nourrit des insectes de couleur vermillon” (holo- type of Rhus Cavaleriei; photo. and fragments in A. A.); inter urbes Kutschou and Liping in fruticetis prope vicum Tsaimou, alt. 600 m., H,. Handel-Mazzetti, no. 10907, July 21, 1917 (syntype of Eurycorym- bus austrosinensis; isotype in A. A.).
Rhus Cavaleriei was first identified with Eurycorymbus austrosinensis by the writer. The species, though apparently rare, has a wide distri- bution having been found also in Kwangtung, on the Lungtoushan, 60 km. east of Siudsao by R. Mell (Pl. Mell. no. 17) and in Formosa, on the river Shin-chou, prov. Karenko by E. H. Wilson, (no. 11098) and 1 This combination is being published almost simultaneously by Handel-Mazzetti
in a sree of his Plantae Mellianae which probably will appear before the present paper 1
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 9
by R. Kanehira (no. 14225), the latter specimen with staminate flow- ers which apparently have not yet been described. They are small, about 4 mm. across, whitish; sepals 5, ovate to oblong-ovate, 1-1.5 mm. long, denticulate, outside like the pedicels densely appressed-pubescent, less so inside; petals 5, oblong-lanceolate about 2 mm. long, unguiculate, villous-pubescent outside and ciliate; disk shallowly cupuliform, with short rounded lobes; stamens 8, nearly twice as long as petals; fila- ments glabrous, folded twice in bud; rudimentary ovary minute, pubescent. SABIACEAE
Sabia gracilis Hemsley in Hooker’s Icon. xx1x. t. 2831 (1907).
Sabia Dunnii Léveillé in Fedde, Rep. aii cade 1x. 457 (1911) ; Fl. Kouy-Tchéou, 379 (1915). — Synon. n
CHINA. Kweichou: Pin-fa, J. en no. 21bis, April 4, 1902, “fleurs vertes-jaunes” (holotype of S. Dunnii; merotype in
A.).
Sabia yunnanensis Franchet in Bull. Soc. Bot. France, xxx1i1. 465 (1886). — Loesener in Ber. Deutsch. Bot. Ges. xxx. 543 (1914), — Léveillé, Cat. Pl. Yun-Nan, 250 (1917).
Celastrus Mairei Léveillé in Fedde, Rep. Spec. Nov. x11. 264 (1914).
Cutna. Yunnan: broussailles de Lou-ké-suin, alt. 3000 m., E. E. Maire, Apr. 1911, “arbuste grimpant tout-vert, fl. vertes, lévre inféri. noire” (syntype of Celastrus Mairei; photo. in A. A.); brousse des mont. derriére Tong-tchouan, alt. 2600 m., E. E. Maire, May 1911, “arbuste grimpant, feuilles caduques, fl. vertes’’ (syntype of C. Mairei; photo. in A. A.)
Celastrus Mairei was according to Loesener, |. c., first identified with Sabia yunnanensis by L. Diels.
Sabia Dielsii Léveillé in Fedde, Rep. Spec. Nov. rx. 456 (1911); Fl. Kouy-Tchéou, 379 (1915).
Sabia rigid Rehder & Wilson in Sargent, Pl. Wilson. 11. 197 (1914). — Chung in Mem. Sci. at China, 1. 153 (Cat. Trees Shrubs China) (1924). — Synon.
Cuina. Kweichou: Pin-fa, ue grotte, J. Cavalerie, no. 1008, May 14, 1903, “h. 1 m., fl. vertes-jaunes” (syntype of S. Dielsii; Bots in A. A.); without locality. J. Esquirol, no. 474, June 1905, “arbrisseau” (syntype of S. Dielsii; merotype in A. A.). Hupeh: Hsing-shan Hsien and Fang Hsien, E. H. Wilson, no. 2534a and b (syn- types of Sabia puberula in A. A.).
10 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
The species was collected in Kweichou also by Y. Tsiang (no. 4630) in 1930 near Cheng-feng.
Sabia parviflora Wallich in Roxburgh, Fl. Ind., ed. Carey, 11. 310 (1824). — Léveillé, Fl. Kouy-Tchéou, 379 (1915).
Cutna. Kweichou: Lo-fou, J. Cavalerie, no. 3415, Aug. 1909; route de Mou-you-se a Kouan-lin, J. Cavalerie, no. 3919, July 25, 1912, “‘arbuste, fl. vertes” (photo. in A. A.).
Both specimens cited are enumerated in the Flore de Kouy-Tchéou under S. parviflora, but only no. 3415 bears the name S. parviflora in Léveillé’s handwriting, while no. 3919 is labeled Celastrus with an apparently unpublished specific epithet.
Sabia parviflora Wall. var. nitidissima Léveillé, Fl. Kouy-Tchéou, 379 (1915)
Celastrus Esquirolit Léveillé in Fedde, Rep. Spec. Nov. x11. 262 914).
Curna. Kweichou:. without locality, J. Esquirol, Aug. 9, 1905 (holotype of C. Esquirolii and Sabia parviflora var. nitidissima; merotype in A. A.).
The variety differs from the type in the generally larger leaves, 10-12 (-15) cm. long and 2.5-4 cm. broad, ovate-oblong to lanceolate-oblong, lustrous and bright green above, lighter green and less lustrous beneath. It resembles Rock’s no. 2202 from Burma, but is more lustrous.
Meliosma Oldhami Miquel in Ann. Mus. Bot. Lugd.-Bat. m1. 94 (1867); Prol. Fl. Jap. 258 (1867).
Rhus Bofillii Léveillé in Mem. Acad. gage xir. 562 (Cat. PI. Kiang-Sou, 22) (1916). — Synon. n
CHINA. Kiangsu: Suo-se, Teha- Tchan, d’Argy (1846-66) (holotype of Rhus Bofilii; fragments in A. A.).
The specimen cited above represents typical M. Oldhamii not M. sinensis Nakai which is probably best considered a glabrous or gla- brescent variety of the former.
RHAMNACEAE
Zizyphus mauritiana Lamarck, Encycl. Méth. 1. 318 (1789). Zizyphus ‘ete (L.) Lamarck, Encycl. Méth. rr. 318 (1789), — éveillé, Cat. Pl. Yun-Nan, 228 (1917). — Non Miller (1768). Paliurus Maire: Léveillé in Fedde, Rep. Spec. hag xix, 535 (1913); Cat. Pl. Yun-Nan, 227 (1917). — Synon. Cuina. Yunnan: brousse, rives du nen Blea, alt. 400 m., E. E. Maire, July 1912 (holotype of Paliurus Mairei; photo. in A. A.).
Berchemia Giraldiana Schneider, Ill. Handb. Laubholzk. 1. 263,
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 11
fig. 182 m-n, 183 k, (1909); in Sargent, Pl. Wilson. 1m. 213 (1914). — Léveillé, Cat. Pl. Yun-Nan, 227 (1917). Microrhamnus Mairei Léveillé in Bull. Géog. Bot. xxv. ie (1915) ; Cat. Pl. Yun-Nan, 227 (1917); pro parte. — Syno ? Microrhamnus Franchetiana Léveillé, 1. c. (1918): ” C917 i Cat. Pl. Seu-Tchouan, 152, pl. 57 (1918), Mscr.; cae eae quoad ramum fructiferum
CuInA. Yunnan: brousse des coteaux a Tong-tchouan, alt. 2550 m., E. E. Maire, May 1912 “arbuste grimpant, feuilles caduques; rameaux long de 2 a 3 m.” (syntype of Microrkhamnus Mairei; mero- type in A. A.); brousse des mont. a Siao-ou-long, alt. 2600 m., E. E. Maire, June 1912 “arbuste grimpant a long rameaux” (fruiting branch of holotype of Microrhamnus Franchetiana, photo. in A. A.).
Both species of Microrhamnus cited above are mixtures; one of the specimens cited under M. Mairei apparently belongs here, the other to the following species. The specimen of M. Franchetiana consists of a flowering branch and of a leafless fruiting branch, the former represents Tripterygium hypoglaucum (Lévl.) Hutch. while the fruiting branch, judging from the shape of the inflorescence belongs here, or possibly to B. floribunda Wall.
Berchemia pycnantha C. Schneider in Sargent, Pl. Wilson. 1m. 215 (1914).
Microrhamnus Mairei Léveillé in Bull. Géog. Bot. xxv. a (1915); Cat. Pl. Yun-Nan, 227 (1917); pro parte. — Syno
CHINA. Yunnan: brousse des collines 4 Kiao-me- eth ak 3100 m., E. E. Maire, May 1913 “arbuste grimpant, long rameaux de 2 a 3 m., fl. blanches, baies noires” (syntype of Microrhamnus Mairei; merotype in A. A.).
Rhamnella Martini (Lévl.) Schneider in Sargent, Pl. Wilson. 11. 225 (1914) “Martinii.”” — Léveillé in Fl. Kouy-Tchéou, 341 (1915); Cat. Pl. Yun-Nan, 227 (1917).
Rhamnus Martini Léveillé in Fedde, Rep. Spec. Nov. x. 473 (1912).
Microrhamnus Cavaleriet Léveillé in Fedde, Rep. Spec. Nov. x11. 535 (1913); Fl. Kouy-Tchéou, 341 (1915). — Synon. nov.
Rhamnus yunnanensis Heppeler in Arch. Pharm. ccixvi. (Ber. Deutsch. Pharm. Ges. xxvitt.) 167 (March 1928), nomen; in Notizbl. Bot. Gart. Mus. Berlin, x. 343 (June 1928). —
. nov.
CuInA. Hupeh: Chanyang, E. H. Wilson, Veitch Exp. no. 1990, April 1901 (semiscandent shrub, 20 ft.) Kweichou: en- virons de Gan-pin, rochers au dessus de Ta-tong, L. Martin in hb. Bo- dinier, no. 2299, May 23, 1898 “arbuste, fl. verdatres” (holotype of
12 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
Rhamnus Martini; merotype in A. A.); Tsin-gai, bois, J. Cavalerie, no. 1152, July 15, 1903, “petit arbre’ (holotype of Microrhamnus Cava- leriei; merotype in A. A.); in fruticetis aridis prope oppidum Huang- tsauba, alt. 1400 m., Handel-Mazzetti, no. 10289, June 15, 1917. Yunnan: Mengtze mountains alt. 7000 ft., A. Henry, no. 10929 (shrub 4-6 ft., fl. green); in rupestribus calcareis in monte Tschang- tschuan, distr. Yunnan-fu, alt. 2200 m., O. Schoch, no. 136 “arbuscula” (holotype of Rkamnus yunnanensis in Berlin Mus.; photo. and isotype in A, A.).
Rhamnella rubrinervis (Lévl.), comb. nov.
Embelia rubrinervis Léveillé in Fedde, Rep. Spec. Nov. x. 374 (1912); Fl. Kouy-Tchéou, 285 (1914). Rhamnella hainanensis Merrill in Philipp. Jour. Sci. xx1. 349 2 ynon. nov.
CuInaA. Kweichou: without locality, J. Esquirol, no. 729 (holotype of Embelia rubrinervis ; photo. in A. A.); hauteurs de Tong- kai, J. Esquirol, no. 3028, July 25, 1911, “fl. vertes” (cited in Fl. Kouy- Tchéou; photo. in A. A.). Hainan: Five Finger Mt., Hop-lo- tsum, F. A. McClure, no. 8358, Dec. 6, 1921 (holotype of Rhamnella hainanensis ; isotype in A. A.).
The Kweichou specimens are in flower, while the Hainan specimen is in fruit, but in their vegetative characters the specimens agree exactly except that the dark reddish color of the veins of the under side of the leaves is lacking on the Hainan specimen; it is apparently more prom- inent on younger leaves. Also the transverse trabecular veinlets are not quite as conspicuous as in the Kweichou specimens. In all specimens the petioles are minutely puberulous, a character not mentioned by Merrill. The floral characters are not different from those of Rham- nella franguloides (Maxim.) Weberb. as figured by Maximowicz under Microrhamnus franguloides Maxim. (in Mém. Acad. Sci. St. Pétersb. sér. 7, x. no. 11, t. fig. 15-22. 1866) and by Schneider (Ill. Handb. Laubholzk. 11. 262, fig. 183 h-]. 1909). The flowers appear in fascicles of 3-7; occasionally two fascicles are borne on short axillary branchlets 5-10 mm. long or sometimes elongated axillary branchlets bear fascicles in the axils of small leaves, resembling in this respect Rhkamnus Es- quiroli Lévl. and Rh. napalensis Wall. The petals of RA. rubrinervis are emarginate and the sepals bear on the inside the lamella-like mid- rib indicated in Schneider’s figure, but not shown by Maximowicz.
Sageretia rugosa Hance in Jour. Bot. xv1. 9 (1878). — Schneider in Sargent, Pl. Wilson. 11. 227 (1914).
Quercus Dunniana Léveillé in Fedde, Rep. Spec. Nov. x11. 363 (1913). — Synon. nov.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 13
Cutna. Kweichou: without locality, J. Cavalerie (holotype of Quercus Dunniana; photo. in A. A.).
Quercus Dunniana was first identified with Sageretia rugosa by W. S. Evans according to a note on the label of the type specimen, but the identification was not published. Quercus Dunniana does not appear under Quercus in the Flore de Kouy-Tchéou and probably was re- ferred by Léveillé to some other plant.
Sageretia theezans (L.) Brongniart in Ann. Sci. Nat. x. 360 (Mém. Fam. Rhamn. 53) (1826).
Berchemia Chaneti aa in Fedde, Rep. Spec. Nov. x. 433 (1912). — Synon. Sageretia Chanetu Schneider in Sargent, Pl. Wilson. 11. 228 (1914).
Cuina, Hopei: montagnes de Ping-chan, L. Chanet, no. 232, June 15, 1908 “arbuste épineux” (holotype of Berchemia Chaneti; merotype in A. A.).
Berchemia Chaneti was referred to Sageretia by C. Schneider and doubtfully enumerated as a new species under S. theezans. It does not seem to me specifically distinct from that variable species.
Sageretia Henryi Drummond & Sprague in Kew Bull. Misc. Inform. 1908, p. 14; 1914, p. 175. — Schneider in Sargent, Pl. Wilson. 11. 623 (1916).
Berchemia Cavalerici Léveillé in Fedde, Rep. Spec. Nov. x. 433 (1912).
Sageretia Cavaleriei Schneider in Sargent, Pl. Wilson. 11. 228 (1914). — Léveillé, Fl. Kouy-Tchéou, 343 (1915 Cuina. Kweichou: Keou-tchang, Kouy-tin, J. Cavalerie, no. 725, Nov. 23, 1902 (holotype of Berchemia Cavaleriei; photo. in A. Peels Rhamnus crenatus Siebold & Zuccarini in Abh. Akad. Muench. Iv. pt. u. 146 (FI. Jap. Fam. Nat. 1. 38) (1845). — Schneider in Sargent, Pl. Wilson. 11. 232, 244 (1914). — Léveillé, Cat. Pl. Yun-Nan, 228 (1917). Rhamnus pseudo-frangula Léveillé in Be ah Rep. Spec. Nov. x. 473 (1912); Fl. Kouy-Tchéou, 343 (19 rte — Schneider in Sargent, Pl. Wilson. 11. 245 (1914). — Synon Celastrus Esquirolianus Léveillé in FI. patie Tchéou, 69 (1914). — nov.
Celastrus Kouytchensis tere in Fedde, Rep. Spec. Nov. x111. 263 (1914). — Synon. no Cutna. Kweichou:. environs de Kouy-yang, mont du Col- lége, le long de l’aqueduc (Ye-mou-kiang-tse), E. Bodinier, no. 1620, June 10, 1897 ‘‘fleurs jaunes-verdatres” (holotype of RA. pseudo- frangula; photo. in A. A.); chemin de Tong-tcheou, rochers, alt. 1000
14 JOURNAL OF THE ARNOLD ARBORETUM [VOL. xv
m., J. Esquirol, no, 3612 (cited in Fl. Kouy-Tchéou as 3618), June 1912 “frutescent, ne s’éléve guére qu’a 2 m., mais ses branches s’éten- dent” (holotype of Celastrus Esquirolianus,; photo. in A. A.); Pin-fa, Majo, J. Cavalerie, no. 3349 (holotype of Celastrus Kouytchensis ; photo. in A. A.).
Schneider speaks (1. c.) of Rh. pseudo-frangula as a species closely related to Rh. crenatus or perhaps a variety and gives a full descrip- tion of it based on Léveillé’s specimen. I cannot see any difference from the variable Rh. crenatus.
Rhamnus heterophyllus Oliver in Hooker’s Icon. xvi. t. 1759 (1888). — Schneider in Sargent, Pl. Wilson. 1. 232 (1914). — Lé- veillé, Fl. Kouy-Tchéou, 342 (1915).
Rhamnus Cavaleriei Léveillé in Fedde, Rep. Spec. Nov. 1x. 326 1910).
Cuina. Kweichou: prés Pin-yue, J. Cavalerie, no. 2477, Aug. 1, 1905, “‘arbuste de 1 m. 50” (holotype of RA. Cavaleriei; photo. in A. A.).
Schneider first identified Rh. Cavaleriei with Rh. heterophyllus and Léveillé cites in his Flore de Kouy-Tchéou the specimen under Rh. heterophyllus without citing a synonym.
Rhamnus Esquirolii Léveillé in Fedde, Rep. Spec. Nov. x. 473 (1912); FI. Kouy-Tchéou, 342 (1915). — Schneider, in Sargent, PI. Wilson. 11. 233 (1914).
Celastrus? Lyi peu in Fedde, Rep. Spec. Nov. xu. 264 (1914). — Synon. n Sageretia Bodiniert Léveillé. Fl. Kouy-Tchéou, 343 (1915).
Cuina. Kweichou: Pin-fa, sous bois, J. Esquirol, no. 392, June 15, 1905 (holotype of RA. Esquirolii; photo. in A. A.) ; environs de Kouy-yang, mont du Collége, E. Bodinier, June 3, 1898 (holotype of Celastrus Lyi and Sageretia Bodinieri; photo. in A. A.).
Schneider gives (I. c.) a complete description of the species. Lé- veillé cites in his Flore de Kouy-Tchéou Celastrus Lyi as a synonym of the new name Sageretia Bodinieri and enumerates besides Bodinier’s specimen Esquirol’s no. 3771 “route de Tong-tchéou, 1200 m., 1912” which I have not seen.
Rhamnus napalensis (Wall.) M. A. Lawson in Hooker f., Fl. Brit. India, 1. 640 (1875), “nipalensis.”
Coalitions — Wallich in Roxburgh, Fl. Ind. ed. Carey, 1. 375, 575 (1824).
Rhamnus paniculiflorus caps in Sargent, Pl. Wilson. 11. 233 (March 1914). — Synon
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 15
Celastrus tristts igen in Fedde, Rep. Spec. Nov. x11. 263 (May 1914). — Synon. n
Cuina. Kweichou: without locality, J. Esquirol, (holotype [2 sheets] of Celastrus tristis, photos. in A. A.).
I am unable to separate RA. paniculiflorus Schneid. from Rh. napa- lensis M. Laws. The distinguishing characters given by Schneider do not seem to hold; I find verrucose branches in a specimen from Assam (L. F. Ruse, no. 58a) and in a specimen from Sumatra (H. S. Yates, no. 725); and the branching of the panicle varies so greatly in the 28 specimens before me that no line can be drawn. The species has a very wide distribution; it occurs throughout China except the northern provinces and extends through India and Indochina to Malaysia.
The specific epithet was published originally by Wallich (1. c.) under Ceanothus as “napalansis,’ but this spelling was corrected in the same volume on p. 575 to “napalensis’”’ which must be accepted as the correct spelling of the name, though most later authors, following Lawson spelled it “nipalensis” or sometimes “nepalensis.”
Rhamnus Bodinieri Léveillé in Fedde, Rep. Spec. Nov. x. 473 (1912); Fl. Kouy-Tchéou, 342 (1915). — Schneider in Sargent, PI. Wilson. 11. 246 (1914)
Cuina. Kweichou: environs de Tsin-gay, rochers au bord de la riviére 4 Ché-tiou-tchay, E. Bodinier, no. 2657, June 27, 1899 (holotype; merotype in A. A.)
Schneider (I. c.) gives a complete description of the species partly based on additional material collected by Henry in Yunnan (nos. 10814 and 10814a).
Rhamnus Blinii (Lévl.), comb. nov.
Maesa Blinii Léveillé in Fedde, Rep. Spec. Nov. x. 376 (1912); FI. Kouy-Tcheéou, 286 (1914).
Cuina. Kweichou: Ma-jo, J. Cavalerie, no. 3103, July 24, 1907 (holotype of Maesa Blinii; merotype in A. A.). Szechuan: Kuan-hsien, alt. 3500-4000 ft., W. P. Fang, no. 2243, July 15, 1928 (shrub 4 m.)
Maesa Blinii is apparently conspecific with Rkamnus Sargentianus Schneid., but differs in several minor characters, so that the latter is probably best considered a variety of the former, the specific epithet “Blinii” having priority. Cavalerie’s no. 3103 differs from Rh. Sar- gentianus chiefly in the quite glabrous, chartaceous and sharply serru- late leaves, distinctly reticulate beneath, with fewer, 10-12, pairs of veins, while in typical Rh. Sargentianus Schneid. the leaves are pilose at least on the veins beneath, the serration is minute and the 10-18
16 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
pairs of veins are connected by conspicuous trabecular veinlets without visible reticulation between. Fang’s no. 2243 approaches the typical form, the veins are more numerous, the trabecular veinlets are more prominent and the serration is less conspicuous; it resembles W. Y. Chun’s no, 4023 from Hupeh which, however, is pubescent on the veins and partly on the veinlets beneath and the veins are closer.*
Rhamuus serpyllifolius Léveillé in Fedde, Rep. Spec. Nov. x1. 282 (1913), “serpyllifolia” ; Cat. Pl. Yun-Nan, 228 (1917) “serpillifolius.”
Cuina. Yunnan: rochers arides de Pan-pien-kai, 2550 m., E. E. Maire, May 1912 “arbrisseau épineux rampant, appliquées sur les rochers, fl. blanches” (holotype; merotype in A
Descriptioni adde: frutex glaberrimus, spinosus, ramulis brevibus suboppositis vel alternis; folia breviter petiolata, coriacea, obovata vel obovato-oblonga, 4-10 mm. longa et 2.5-4 mm. lata, apice rotundata vel leviter emarginata, basi cuneata, integra vel apicem versus sparsis- sime minute et indistincte denticulata, supra nitida, intense viridia, subtus pallidiora, costa et venis indistinctis; flores (deflorati) solitarii, breviter pedicellati pedicello ca. 1 mm. longo, 4-meri, sepalis triangulari- ovatis, acutiusculis, ca. 1 mm. longis.
This is a very distinct species on account of its very small coriaceous entire or nearly entire leaves. The collector notes the flowers as white, but this may be a mistake, for the few flowers on the specimen show neither petals nor stamens. The species seems best placed near RA. Rosthornii Pritz., Rh. leptacanthus Schneid. and Rh. Leveilleanus Fedde but these have membranous, distinctly denticulate and generally larger leaves. According to Heppeler’s disposition of the species (Beitr. Syst. Rhamnus in Arch. der Pharmazie, ccivr. 152-173. 1928) it should be placed into his series Coriaceae, though he refers Rh. Rosthornii and Rh, Leveilleanus to the series Parvifoliae.
Rhamnus Taqueti Léveillé in herb. apud Schneider in Sargent, Pl. Wilson. 11. 248 (1914).
1Rhamnus Blinii var. nig nes pay Se Seealar ), stat. oe wo ciate: eee in Sargent, Pl. Wilson, ‘Ir, 235 (1914).
CH We aoe oo and Sept. 1908 holotype) same locality, E. H. Wilson, Veitch Exp. no. 3339a June 1903 (paratype); Mupin, EF. H. Wilson, no. 862a, June 1908 (paratype)
without precise locality, E. H. Wilson, Veitch Exp. no. ” 3339, May ane (pa type). Western ek 6 li from Hsien-tien- tze, alt. 6500 Chun, no. 4023, Aug. 23, 1922 (tree, 6 m.). Yunnan: Chien-chuan- hebone petted lat. 26° 30-36’ N., long. 99° 40’ E., alt. 9-10000 tt, G. Forrest, nos. 21482
=i ie) _ oO ~— ae i &, te ~ =) Qa. > a _ tw = =e > ° ,& — so) ad Oo -o nm o 7S te] 7) co io) >| 3 a ~ Ss “3s ° +
e variety has apparently a much wider distribution than the typical form and sare must be considered the phylogenetic type.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 17
Prunus Taqueti Léveillé in Fedde, Rep. Spec. Nov. vir. 197 (1909).
KorEA. Quelpaert: Hallisan, E. Taquet, no. 104, Sept. 1907
(syntype of Prunus Taqueti; photo. in A. A.); same locality, U. Faurie, no. 1870, July 1907 (syntype of Prunus Taqueti,; isotype in A. A.).
Prunus Taqueti was first recognized by E. Koehne as a species of Rhamnus according to a note on the sheet of Taquet’s no. 104 and a reference in Sargent, Pl. Wilson. 11. 276 (1912).
Rhamnus Leveilleanus Fedde, Rep. Spec. Nov. x. 272 (1911). — Schneider in Sargent, Pl. Wilson. 1. 248 (1914). — Léveillé, Fl. Kouy- Tchéou, 243 (1915).
Rhamnus Cavaleriet Léveillé in Fedde, Rep. Spec. Nov. x. 148 (1911); nec Léveillé, op. cit. 1x. 226 (1910).
Cuina. Kweichou: Lo-fou, J. Cavalerie, no. 3348, April 1907 (holotype of Ré#. Cavaleriei; merotype in A. A.); environs de Kouy-yang, E. Bodinier, no. 2656, June 15, 1898 (cited in Fl. Kouy- Tchéou; dupl. in A. A
Rhamnus leptophyllus Schneider in Notizbl. Bot. Gard. Berlin, v. 77 (1908); in Sargent, Pl. Wilson. 11. 239 (1914). — Léveillé, Cat. PI. Yun-Nan, 228 (1917).
ene pruniformis Léveillé in Fedde, Rep. Spec. Nov. x11. 534 (1913); Cat. Pl. Yun-Nan, 228 (1917). — Synon. nov.
Cuina. Yunnan: Kin-tchong-chan, 2550 m., E. E. Maire, April 1912, ‘“arbuste épineux buissonant, haut 1.50 m., fl. jaunes, abondantes” (holotype of Rk. pruniformis; merotype in A. A.).
In the Herbarium Léveillé there is another specimen (merotype in A. A.) from the same locality with very young leaves which seems to be referable to Rh. Rosthornii Pritz.
Rhamnus Schneideri Léveillé & Vaniot in Fedde, Rep. Spec. Nov. vi. 265 (1908). — Schneider in Sargent, Pl. Wilson. 11. 250 (1914).
Korea: Nai-piang, U. Faurie, no. 234, July 1901 (holotype; photo. in A. A.)
Rhamnus hamatidens Léveillé in Fedde, Rep. Spec. Nov. x. 473 (1912); Fl. Kouy-Tchéou, 342 (1915). — Schneider in Sargent, PI. Wilson. 11. 252 (1914
CuHtInaA. Kweichou: Pin-fa, J. Cavalerie, no. 992, April 15, 1903, ‘‘arbuste, fl. vertes” (holotype; photo. in A. A.).
Hovenia dulcis Thunberg, Fl. Jap. 101 (1784). — Léveillé, FI. Kouy-Tchéou, 340 (1915); Cat. Pl. Yun-Nan, 227 (1917).
Ziz pls Esquiroli Léveillé in Fedde, Rep. ape fee x. 148 1S 8 a ae Kouy- Tchéou, 343 (1915). — Synon
nd Kweichou: without locality, J. Biase ok no. 861
13 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
(holotype of Zizyphus Esquirolii; merotype in A. A.).
Gouania javanica Miquel, Fl. Ind. Bat. 1. pt. 1, p. 649 (1855). — Léveillé, Cat. Pl. Yun-Nan, 227 (1917). — Exell in Sunyatsenia, 1. 93 (1933
Terminalia Kouytchensis Léveillé Cat. Pl. Yun-Nan, 35 (1915), sub
Cuina. Kweichou: Lo-fou, J. Cavalerie, no. 3457 (holotype of Terminalia Kouytchensis; merotype in A. A.).
Terminalia Kouytchensis was only incidentally mentioned by Léveillé in the description of his Terminalia Mairei (\. c.): “folia .. . basi brevissime attenuata nec cordata (ex quo a T. Kouytchensis Lévl. diag- noscitur)”. Terminalia Mairei is according to Exell (1. c. 87) identical with Combretum Wallichti DC. Terminalia Kouytchensis is not enu- merated by Léveillé in his Flore de Kouy-Tchéou; though Terminalia appears in the key of genera under Combretaceae (p. 74), no species is listed. Exell (1. c.) first published it as a synonym of Gouania javanica Miq., to which I had it already referred in this herbarium.
VITACEAE
Vitis flexuosa Thbg. var. parvifolia (Roxb.) Gagnepain in Sar- gent, Pl. Wilson. 1. 103 (1911). — Léveillé, Cat. Pl. Yun-Nan, 8, (1915)
Vitis Cavaleriet Léveillé & Vaniot in Bull. Soc. Agr. Sci. Arts Sarthe, xv.' 36 (1905) ; in Fedde, Rep. Spec. Nov. 11. 158 (1906).
Vitis flexuosa Léveillé, Fl. Kouy-Tchéou, 27 (1914), pro parte, vix Thunb.
Cuina. Kweichou: Kouy-yang, mont du College, J. Cava- lerie, April 25, 1898 (holotype of Vitis Cavaleriei; photo. in A. A.).
Vitis Cavaleriei was referred by Léveillé in his Flore de Kouy-Tchéou to Vitis flexuosa as a synonym together with Vitis Marchandii Lév). which, however, belongs to the following species.
Vitis Wilsonae Veitch apud Gard. Chron. ser. 3, xLvr. 236. fig. 101 (1909). — Rehder in Sargent, Pl. Wilson. m1. 428 (1917), pro syn. V. reticulatae Gagnep.; in Jour. Arnold Arb. x11. 339 (1932).
Vitis reticulata Pampanini in Nuov. Giorn. Bot. Ital. xvi. 429, fig. 13 (1910); non M. A. Lawson.
Vitis Marchandii Léveillé in Fedde, Rep. Spec. Nov. x11. 531 (1913).
Vitis flexuosa Léveillé, Fl. Kouy-Tchéou, 27 (1914), quoad syn. V. Marchandii; non Thunb. n the reprint or Léve illé’s Les vignes de la Chine, on the cover of the first ue
one aining the original ti icle and in Fedde’s Repertorium the volume number given as Lx. eae of x
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 19
Cuina. K weichou: Chouang-chan-po, J. Esquirol, no. 3143, May 1911 (syntype of V. Marchandii; photo. in A. A.); bois, Gan- chouen, J. Cavalerie, no. 3874 (syntype of V. Marchandii; photo. in A. A.).
Vitis Marchandii was reduced by Léveillé in his Flore de Kouy- Tchéou to a synonym of V. flexuosa, but it clearly is identical with V. Wilsonae Veitch.
There may be some doubt whether V. Marchandii, though later than V. Wilsonae, should not be considered the correct name for this species, as it was correctly published with a Latin description and based on type specimens, while the name V. Wilsonae was only incidentally mentioned, though with a figure of a leaf. As far as the two descriptions go, they are both unsatisfactory and incomplete, and particularly misleading is Léveillé’s statement “V. inconstanti Miq. affinis,’ while V. Wilsonae is much more correctly referred as belonging into the affinity of V. armata.
Vitis Davidi (Rom. du Caill.) Foéx, Cours Vitic. 44 (1886).
Vitis prunisapida Léveillé & Vaniot in sii Rep. ae Nov. III. 350 (1907); Fl. Kouy-Tchéou, 28 (1914). — Synon. n
Cutina. Kweichou: Touan-po, J. Cavalerie, no. p88: Aug. 13, 1905, “‘vigne a gros raisins, peu aqueux et a gout de pruneau” (holo- type of V. prunisapida; photo. in A. A.).
Vitis Thunbergii Sieb. & Zucc. var. adstricta (Hance) Gagnepain in Sargent, Pl. Wilson. 1. 105 (1911). — Léveillé, Cat. Pl. Yun-Nan, 8 (1915)
Vitis flexuosa Thunb. var. Mairei Léveillé in Fedde, Rep. Spec. Nov. vit. 340 (1909) ; Cat. Pl. Yun-Nan, 8 (1915). — Synon. nov.
Cuina. Yunnan: montagnes, E. E. Maire, no. 415, May 1906, (holotype of V. flexuosa var. Mairei in herb. Bonati, photo. in A. A.; merotype in herb. Leveillé).
Léveillé in his Catalogue des plantes du Yun-Nan gives for V. flex- uosa var. Mairei the author citation “Lévl. pr. sp. in Fedde, Rep. vu. 338, but this is apparently a mistake; Vitis Mairei on that page is partly Cayratia oligocarpa var. glabra Gagnep. and partly Tetrastigma serrulatum (Roxb.) Planch., while the variety appears on p. 340. The actual type of the variety seems to be the specimen in the Herb. Bonati, now in the herbarium of the University of California; the specimen in the Léveillé herbarium is only a fragment.
Tetrastigma Hemsleyanum Diels & Gilg in Bot. Jahrb. xxrx. 463 (1900). — Gagnepain in Lecomte, Not. Syst. 1. 317 (1911).
Vitis Labordei Léveillé in Bull. Soc. Agric. Sci. Arts. Sarthe, xv. 40
20 JOURNAL OF THE ARNOLD ARBORETUM [voL, Xv
(1905); in Fedde, Rep. Spec. Nov. 1. 158 (1906); Fl. Kouy- Tchéou, 27 (1914). — Synon. nov. Vitis Esquirolui Léveillé & Vaniot in Fedde, Rep. Spec. Nov. 111. 20 (1906); Fl. Kouy-Tchéou, 27 (1914); Cat. Pl. Yun-Nan, 8 . Nov
CHINA. Kweichou:. environs de Kouang-chouen, bois de Ma-yan-chan, J. Laborde, no. 2695, Aug. 27, 1899 (holotype of Vitis Labordei; photo. in A. A.); Pin-fa, J. Cavalerie, no. 1051, June 11, 1903 “fl. blanches” (holotype of Vitis Esquirolii; merotype in A. A.).
It is not without hesitation that I refer Vitis Labordeit as a synonym to T. Hemsleyanum, but I cannot see the slightest difference in the leaves of both species; also the inflorescence seems to be the same, though the type of 7. Hemsleyanum is in flower, while that of Vatis Labordei is in fruit and has the secondary axes longer. Gagnepain (1. c.) describes the fruit of 7. Hemsleyanum from a specimen col- lected by Farges in eastern Szechuan as being globose with two fur- rowed-ruminate seeds, while the fruit of V. Labordei is pyriform with one seed which has two deep furrows on the ventral side, but the shape of the fruit would change with the number of seeds which is variable in most species.
In the Flore de Kouy-Tchéou Léveillé enumerated under Vitis Es- quirolii another specimen, J. Cavalerie, no. 2471, July 26, 1905, which according to a duplicate before me belongs to Cayratia oligocarpa (Lévl. & Vant.) var. glabra Gagnep.
Tetrastigma serrulatum (Roxb.) Planchon in de Candolle, Monog. Phaner. v. 432 (1887). — Gagnepain in Lecomte, Not. Syst. 1. 322 (1911).
Vitis Mairei Léveillé in a! aan ove Nov. vir. 338 (1909), pro parte, quoad no. 412. — §
Vitis flexuosa Thbg. var. Maire on sp. Léveillé, Cat. Pl. Yun-Nan, 8 (1915), errore.
Cuina. Yunnan: forets des montagnes, EL. FE. Maire, no. 412, May 1904 “‘tiges perissantes tous les ans’? (syntype of Vitis Mairet in hb. Bonati; photo. in A. A.).
Léveillé cites under V. Mairei two numbers 412 and 414 which I have seen in the herbarium Bonati, now in the herbarium of the University of California, with the name Vitis Mairei sp. nov. in Léveillé’s hand- writing. Whether there are duplicates in the Léveillé herbarium I do not know. Maire’s no. 414 is Cayratia oligocarpa (Lévl.) var. glabra Gagnep
In bis Catalogue des plantes de Yun-Nan Léveillé cites under Vitis flexuosa var. Mairei Lévl. pr. sp. the reference to V. Mairei, apparently
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 21
confusing it with the plant described under the former name on page 340, which is identical with Vitis Thunbergii S. & Z. var. adstricta (Hance) Gagnep. In 1912 Léveillé used the name V. Mairei a second time for an en- tirely different plant belonging to the Cucurbitaceae (see p. 27). Tetrastigma obtectum (Wall.) Planch. var. B glabrum Gagnepain in Lecomte, Not. Syst. 1. 323 (1911). Vitis eee Hemsley in Jour. Linn. Soc. Bot. nee ad (1886). — Le n Bull. Agric. Sci. Arts Sarthe, xv. 45 05). V itis Polentille Lévl. & Vant. var. glabra Léveillé in nk Agr. Sci. Sarthe, xL. 43 (1905) ; in Fedde, Rep. Spec. Nov. 11. 160 (1906). Vitis obtecta Wall. var. Potentilla f. glabra (Lévl.) Léveillé, FI. Kouy-Tchéou, 28 (1914). Vitis obtecta Wall. var. glabra (Lévl.) Léveillé, Cat. Pl. Yun-Nan, 8 (1915).
Cuina. Kweichou: environs de Gan-pin, sur les rochers, aux Grandes Rocailles, L. Martin & E. Bodinier, no. 1879, p. p., Sept. 26, 1897 (syntype of Vitis Potentilla var. glabra, ex Léveillé); environs de Gan-pin, rochers a la depression, grotte; environs de Ou-la-gay, sur les rochers, L. Martin & E. Bodinier, nos. 2191, 2191bis, p. p., (syn- type of V. Potentilla var. glabra, p. p.; photo. in A. A.); Pin-fa, J. Cavalerie, nos. 1331, 1332, 3253, May & July 1902 and May 1908 (cited under V. obtecta var. Potentilla f. glabra in Fl. Kouy-Tchéou; photos. in A. A.).
In his original publication of Vitis Potentilla and var. glabra, Léveillé’ cites the types in one paragraph at the end without indicating which specimens represent the type of the species and which that of the variety, but the fact that both type numbers 1879 and 2191 consist each of both forms, may be his reason for not separating them in his citations. Of no. 1879 I have seen only var. pilosa, and no. 2191 he separates in Flore de Kouy-tchéou into no. 2191 and no. 2191bis.
Vitis umbellata Hemsl. = Tetrastigma umbellatum (Hemsl.) Nakai, of which I have seen the type in Kew, does not seem to be different from T. obtectum var. glabrum,; Nakai refers also Vitis arisanensis Hay. as a synonym to 7. umbellatum (in Tokyo Bot. Mag. xxxv1, 69. 1922). Léveillé (1. c.) cites in 1905 Cavalerie nos. 1331 and 1332 under Vitis umbellata Hemsl., but in his Flore de Kouy-Tchéou he enumerates them under V. obtecta var. Potentilla {. glabra. Gagnepain in his Essai de classification du genre Tetrastigma (in Lecomte, Not. Syst. 1. 306-326. 1911) does not mention Vitis umbellata Hemsl. at all.
Tetrastigma obtectum var. Potentilla (Lévl. & Vant.) Gagnepain in Lecomte, Not. Syst. 1. 323 (1911).
22 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
— Segoe Léveillé & Vaniot in Bull. Agric. Sci. Arts Sarthe, 3 (1905); in Fedde, Rep. Spec. Nov. 11. 159 (1906). Vitis ooh ta var. Potentilla f. pilosa (Pl.) Léveillé, Fl. Kouy-Tchéou, 28 (1914); vix T. obtectum var. pilosum Pl. apud Gagnep. Vitis obtecta Wall. var. Potentilla (Lévl. & Vant.) Leéveillé, Cat. PI. Yun-Nan, 9 (1915).
Cuina. Kweichou:. environs de Gan-pin, sur les rochers, aux Grandes Rocailles, L. Martin & E. Bodinier, no. 1879, p. p., Sept. 26, 1897 (syntype of Vitis Potentilla; photo. in A. A.); environs de Gan- pin, rochers a la depression, grotte; environs de Ou-la-gay, sur les rochers, L. Martin in herb. E. Bodinier, no. 2191bis (syntype of V. Potentilla; photo. in A. A.).
In his Flore de Kouy-Tchéou Léveillé places Vitis Potentilla under V. obtecta var. Potentilla f. pilosa, but in his Cat. Pl. Yun-Nan he separates var. pilosa and var. Potentilla following Gagnepain.
Parthenocissus himalayana (Royle) Planch. var. rubrifolia (LévI. & Vant.) Gagnepain in Sargent, Pl. Wilson. 1, 101 (1911). ises rubrifolia Léveillé & Vaniot in Bull. oe Agric. Sci. Sarthe 44 (1905); in Fedde, Rep. Spec. Nov. 11. 101 (1911). a Delavayana Léveillé, Fl. Kouy aoa, 27 (1914), tandem quoad synon. I’. rubrifolia; non V. Delavayana (Planch.) Franch.
Cuina. Kweichou: environs de Kouy-yang, mont du Col- lége, gorges de Yang-pa, E. Bodinier, no. 2457, p. p., July 20, 1898 “liane ligneuse, pedoncules rouges” (syntype of Vitis rubrifolia ; photo. in A. A.); mont de Kao-po, sur les rochers, E. Bodinier, no. 2457, p. Pp., Nov. 8, 1899, “pedoncules et feuilles rouges” (syntype of Vitis rubri- folia; photo. in A, A.).
Vitis rubrifolia was referred in 1911 by Gagnepain as a variety to Parthenocissus himalayana, but in 1914 Léveillé made it a synonym of V. Delavayana. Whether Cavalerie’s specimen from Pin-fa, also cited under V. Delavayana belongs to that species or to P. himalayana var. rubrifolia | do not know.
Parthenocissus heterophylla (Bl.) Merrill in Philipp. Jour. Sci. xi. Bot. 143 (1918). Vitis Feddei Léveillé in Fedde, Rep. Spec. Nov. vir. 231 (1909). — Synon. nov.
CuHInaA. Kweichou: route de Pin-fa a Tou-yun, J. Cavalerie, no. 3347, Aug. 2, 1908 (holotype of Vitis Feddei; merotype in A. A.).
Vitis Feddei is not mentioned by Léveillé in his Flore de Kouy- Tchéou nor do I find Cavalerie’s number cited under any other species of Vitis.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 23
Parthenocissus tricuspidata (Sieb. & Zucc.) Planchon in De Can- dolle, Monog. Phaner. v. 452 (1887). Vitis Taquetu Léveillé in Bull. Acad. Intern. Géog. Bot. xx. 11 (1910). — Synon. nov. Psedera Thunbergii (Sieb. & Zucc.) Nakai, Fl. Sylv. Kor. xi. 11, pl. 1 (1922). Korea. Quelpaert: scandens in muris agrorum Hongno, E. Taquet, no. 648, July 1908 (holotype of V. Taquetii; isotype in A. A.). Vitis Taqueti was first identified with P. tricuspidata by Nakai (1. c.); it represents the juvenile form of Parthenocissus tricuspidata; Taquet’s no. 648 is sterile, but his no. 5454 also under V. Taquetii in the Herb. Léveillé collected August 1911 at the same locality bears small inflorescences. Of the three other specimens before me collected by Faurie July 1908 at the same locality, no. 646 is similar to the type of V. Taquetii, no. 647 bears the large three-lobed leaves characteristic for the mature plant, and no. 649 is intermediate; all three have inflorescences with flower-buds.
Ampelopsis Bodinieri (Lévl. & Vant.), comb. nov. Vitis Bodiniert Léveillé & Vaniot in Bull. Soc. Agric. Sci. Sarthe, xL. 36 (1915); in Fedde, Rep. Spec. Nov. 11. 157 (1906). Vitis oo Léveillé, Fl. Kouy.Tehéou, 27 (1914), p. p.; vix
Ampelopsis micans Rehder in Mitteil. Deutsch. Dendr. Ges. xxr. 188 (1912). — Synon. nov.
Cuina. Kweichou: environs de Kouy-yang, mont du Col- lége, haies, rochers, E. Bodinier, no. 2291, May 29 and June 16, 1898 “liane ligneuse a long sarments” (syntypes of Vitis Bodinieri; photo. of specimen of May 29 in A. A.)
Vitis Bodinieri is referred by Léveillé in his Flore de Kouy-Tchéou as a synonym to V. heterophylla Thunb. (= Ampelopsis brevipedunculata | Maxim.] Koehne), but it apparently is identical with the closely re- lated A. micans Rehd., a name over which the specific epithet of V. Bodinieri has priority.
Ampelopsis Bodinieri var. cinerea (Gagnep.), comb.
Vitis Lyjoannis Léveillé in Bull. Bot. Soc. France, Liv. (1907).
Ampelopsis heterophylla var. cinerea Gagnepain in Sargent, Pl. Wil- son. 1. 101 (1911), p. p.
Ampelopsis micans var. cinerea Rehder in Mitt. Deutsch. Dendr. Ges. xxr. 189 (1912).
Vitis heterophylla Léveillé, Fl. Kouy-Tchéou, 27 (1914), p. p.; non Thunberg.
CuInA. Kweichou: Pin-uo, Ton-yun, terrains pierreux, J.
24 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
Cavalerie, no. 2478, Aug. 1, 1905 “plante de 2 a 4 metres, assez dressée”’ (holotype of Vitis Lyjoannis; photo. in A. A.).
Vitis Lyjoannis was referred as a synonym to V. heterophylla by Lé- veillé (1. c.), but apparently belongs to A. micans var. cinerea, though it differs from the type of that variety by the slightly or scarcely 3- lobed leaves, but the branch and the leaves beneath are rather densely pubescent.
Ampelopsis Delavayana Planchon in De Candolle, Monogr. Phaner. v. 458 (1887).
Vitis rigida Léveillé & Vaniot in Bull. Agric. Sci. Sarthe xL. 44 (1905); in i a Spec. Nov. 11. 160 (1906); Fl. Kouy- Tchéou, 28 (191 — Synon. nov
CHINA. Kweic % ou: environs de Kouy-yang, mont du Col- lege, E. Bodinier, no. 2402 p. p., May 28, 1898 (syntype of Vitis rigida; photo. and merotype in A. A.); haies prés de Kien-lin-chan, E. Bodinier, no. 2402, p. p., July 7, 1898 (syntype of V. rigida, ex Léveillé).
Ampelopsis Delavayana var. Gentiliana (Lévl. & Vant.), comb.
nov. Vitis Gentiliana Léveillé & Vaniot in Bull. Agric. Soc. Sarthe, XL. 1905); in Fedde, Rep. Spec. Nov. 11. 158 (1906); Fl. Kouy- Tchéou, 27 (1914). Ampelopsis heterophylla var. Gentiliana (Lévl. & Vant.) Gagnepain in Sargent, Pl. Wilson. 1. 100 (1911).
CutIna. Kweichou: environs de Kouy-yang, mont du Col- lege (N.-D. de Liesse), haies, E. Bodinier, no. 2301, June 9, 1898, “grande liane sous-ligneuse”’ (holotype of Vitis Gentiliana, merotype in A. A
Vitis Gentiliana was referred by Gagnepain (1. c.) as a variety to A. heterophylla S. & Z. = A. brevipedunculata (Maxim.) Koehne, as was also A. Delavayana, but they are hardly conspecific with A. brevi- pedunculata which always has undivided though more or less lobed leaves and fruits of a different color. Vitis Gentiliana and A. Dela- vayana though quite distinct in their extreme forms are apparently variations of a polymorphous species, since their typical forms are con- nected by intermediate specimens, as Wilson, no. 604, from Hupeh, Bock & Rosthorn, no. 1540! and Fang, no. 473, from Szechuan, Schoch, no. 168, and Rock, no. 6389, from Yunnan. Both forms vary from densely pilose on the veins to quite glabrous. In the shape of the foliage Wilson, no. 2734, is very similar to the type of Vitis Gentiliana,
1See Jour. Arnold Arb. xiv. 349 (1923).
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 25
but the leaves are quite glabrous, while in the latter the veins beneath are densely pilose.
Ampelopsis aconitifolia Bge. var. palmiloba (Carr.) Rehder in Mitt. Deutsch. Dendr. Ges. xx1. 190 (1912); in Sargent, Pl. Wilson, mr. 427 (1917).
Vitis Dunniana Léveillé in Fedde, Rep. Spec. Nov. x1. 297 (1912). — Synon. nov.
KorEA: in dumosis Chinnampo, U. Faurie, no. 535, Aug. 1906 (hol- otype of Vitis Dunniana; merotype in A. A.).
There are no specimens either of A. aconitifolia or of this variety from Korea in this herbarium, nor can I find any records. Faurie’s no. 535 is not represented in this herbarium, though we have a good set of his collections.
Ampelopsis Chaffanjoni (Lévl.), comb. nov.
Vitis Chaffanjoni Léveillé in Bull. Soc. Agric. Sci. Sarthe xi. 37 (1905).
Leea Dielsii Léveillé in Fedde, Rep. Spec. Nov. virr. 58 (1910) ; FI. Kouy-Tchéou, 25 (1914). — Synon. nov. Meliosma? Cavaleriei Léveillé in Fedde, Rep. ee cia Ix. 457 (1911); Fl. Kouy-Tchéou, 378 (1915). — Synon Vitis dings eae Fl. Kouy-Tchéou, 28 “(1914) ; nom. nudui
Ampelopsis desea es ‘Wilson in Jour. Roy. Hort. Soc. xui1. 37 16), nom. nud. — Rehder & Wilson in “Sargent, Pl. Wilson. III. or (1917). — aie nov.
Cuina. Kweichou: environs de Kouy-yang, mont du Col- lége, gorges de Yang-pa, J. Chaffanjon in herb. Bodinier, no. 2386, June 20, 1898 “grande liane sous-ligneuse” (holotype of Vitis Chaffan- joni; photo. in A, A.); Tsin-gay, J. Cavalerie, no. 1243, Aug. 1903 (holotype of Leea Dielsii; photo. in A. A.); Pin-fa, J. Cavalerie, no. 1027, June 3, 1903 “fl. vert-jaunes” (holotype of Meliosma Cavaleriet; photo. in A. A.); Pin-fa, Kiang-ouang, J. Cavalerie, no. 2362, May 25, 1905 (holotype of Vitis megaphylia; photo. in A. A.).
The type specimen of Vitis Chaffanjoni is enumerated by Léveillé in his Flore de Kouy-Tchéou under Leea Dielsii without V. Chaffanjoni being cited as a synonym. Vitis megaphylla Lévl. is published by Lé- veillé (1. c.) without description and I was unable to find any previous reference to it. From a pencil note apparently not in Léveillé’s hand- writing on the type specimen reading “Vitis aff. V. megaphylla” I assume that this name is probably a slip of the pen for V. megalophylla, and Léveillé took it up by mistake as V. megaphylla Lévl. when compil- ing his Flore de Kouy-Tchéou. The only complete leaf of the type
26 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
specimen is from a vigorous shoot and has coarsely serrate unusually large leaflets (up to 11 cm. long).
Ampelopsis cantoniensis (Hook. & Arn.) Planchon in De Candolle, Monog. Phaner. v. 460 (1887).
Vitis multijugata Léveillé & Vaniot in Bull. Soc. Agric. Sci. Sarthe 1 (1905); in Fedde, Rep. Spec. Nov. 11. 159 (1906); FI Kouy-Tchéou, 28 (1914). — Synon. nov. Leea theifera Léveillé in Fedde, Rep. Spec. Nov. vir. 58 (1910) ; “l, Kouy- Tcheou, 25 (1914). —Synon. nov.
Cuina. Kweichou: environs de Tou-chan, J. Cavalerie, June- July 1899 (holotype of Vitis multijugata; photo. in A. A.); Pin-fa, cat- aracte au sud-ouest, fourrés, J. Cavalerie, no. 243, Aug. 21, 1902, “les Miao-tsé font leur thé avec le fruit mtr violet de cette plante” (holo- type of Leea theifera; photo. in A. A.).
The two specimens cited above differ considerably in the shape of their leaflets; in the specimen from Tou-chan they are generally ovate and shallowly toothed, while in that from Pin-fa they are oblong-ovate to oblong-lanceolate and coarsely toothed. The latter is apparently referable to var. grosse-dentata Hand.-Mazz. (in Anz. Akad. Wiss. Wien, 1922, p. 105 [Pl. Nov. Sin. Forts. 16, p. 43]), which is closely connected with the type by intermediate forms.
Cayratia oligocarpa (Lévl. & Vant.) Gagnepain in Lecomte, Not. Syst. 1. 348, 359 (1911).
~~ oligocarpa Léveillé & rs in Bull. Soc. Agric. Sci. Sarthe, 41 (1905); in Fedde, Rep. Spec. Nov. 11. 159 (1906); FI. a Tchéou, 28 (1914); Cat. Pl. Yun-Nan, 8 (1915). Columella oligocarpa (Lévl.) Rehder in Jour. Arnold Arb. 11. 177 (1921).
CHINA. Kweichou: environs de Kouy-Yang, C. dans les haies, E. Bodinier, no. 2387, June 29, 1908 (holotype of Vitis oligo- carpa (2 sheets, flowers and fruits); 2 photos. in A. A.).
I have taken up again the generic name Cayratia Juss., since Colu- mellia Ruiz & Pav. will be proposed and doubtless accepted as a nomen conservandum to prevent the renaming, on account of the priority of Columella Lour., of Columellia which represents a distinct family. Both names must be regarded as orthographic variants, having been given in honor of the same man.
Cayratia oligocarpa var. glabra Gagnepain in Lecomte, Not. Syst. 1. 360 (1911).
Vitis Maire: Léveillé in Fedde sep Se Nov. vir. 338 (1909), quoad Maire, no. 414. — Synon.
CuHInA. Yunnan: foréts des ones E. E. Maire, no. 414,
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 27
May 1904 (syntype of V. Mairei [in herb. Bonati]; photo. in A. A.).
As in the case of Maire, no. 412 (see under Tetrastigma serru- latum), I have seen a specimen of Maire, no. 414, only in the herbarium Bonati. This is apparently the actual type specimen and has the name, Vitis Mairei sp. nov., in Léveillé’s handwriting.
Cayratia dichromocarpa (Lévl.), comb. nov.
Vitis dichromocarpa Léveillé in Fedde, Rep. Spec. Nov. ut. 350 (1907) ; Fl. Kouy-Tchéou, 27 (1914).
CuIna. Kweichou: Pin-fa, J. Cavalerie, no. 2416, July 19, 1905, “fruit jaune avant compléte maturité, devient ensuite noir lui- sant” (holotype of Vitis dichromocarpa; merotype in A. A.).
I have not been able to identify this plant with any described spe- cies; it looks quite distinct on account of its large leaflets rather densely covered beneath with gray villous pubescence, the veins being densely pilose; the middle leaflet is oblong, acuminate, about 11 cm. long and 4.5 cm. wide, the lateral ones oval or ovate, about 7 cm. long and 4 cm. wide, acute or obtuse and mucronate, coarsely crenate-dentate. It is perhaps nearest to C. oligocarpa, but the peduncle is articulate and bracteoled in the middle; the seeds are similar to those of that species and prominently T-shaped in cross-section.
Three of the new species of Vitis proposed by Léveillé do not belong to the Vitaceae but to the Cucurbitaceae. These are: Vitis Martini Léveillé & Vaniot (in Bull. Soc. Agric. Sci. Sarthe, xt. 41. 1905), Vitis quelpaertensis Léveillé (in Fedde, Rep. Spec. Nov. x. 351. 1912)" and Vitis Mairei Léveillé (in op. cit. x1. 299. 1912; not Léveillé 1909) which apparently all belong to the widely distributed and variable Gyno- stemma pentaphyllum (Thunb.) Makino (G. pedatum B1.).
(To be continued)
HERBARIUM, ARNOLD ARBORETUM, HarvaArRD UNIVERSITY.
ae holotype of Vitis ste ocieca is Taquet no. 2745, as are by Léveillé, but n his herbarium this s en bears only the name V. Taque i in Léveillé’s hand- cara a name given OY pre the mae before to tal seal tricuspidata
28 JOURNAL OF THE ARNOLD ARBORETUM [VOL. xv
SPECIATION IN UVULARIA
Epcar ANDERSON AND THoMAS W. WHITAKER
With plates 82 and 83 and five text figures
‘I. A CYTOLOGICAL SURVEY OF UVULARIA
THE PURPOSE of this section of the investigation was to determine by cytological examination the evolutionary importance of certain phenomena already known to occur in the genus Uvularia. Belling (1925) had shown that in U. grandiflora non-disjunction, lack of pair- ing, fragmentation of chromosomes, and the duplication of single chro- mosomes or the entire chromosome complement may take place under experimental conditions. Our problem was to determine whether in the genus Uvularia these phenomena are of importance in the differen- tiation of species. If chromosome duplication has been one of the causes of specific differentiation, these two species of Uvularia should differ in their chromosome number. If the entire chromosome set has been duplicated, one would expect to find polyploid strains within the species, such as have been reported for Tradescantia (Anderson and Diehl 1932), Tripsacum (Mangelsdorf and Reeves, 1931) and other genera. If chromosome fragmentation has been involved in species differentiation, one would expect to find single individuals or whole geographic races characterized by the possession of fragmented chro- mosomes.
The material is excellent for such a study. The species under con- sideration have well-marked specific differences and the chromosome complement of one species has been worked out in detail (Belling, 1925). It has seven (n) chromosomes, each of which can be identified by its morphological peculiarities. Therefore if gross chromosomal differences are responsible for specific differentiation in this genus, the point should be rather easily detected.
An effort was made to obtain plants for cytological examination from as many widely separated points as possible. Uvularia grandiflora was collected at five points and U. perfoliata at two. The data are summarized in Table I.
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 29
TABLE IT. No. OF PLANTS CHROMOSOME .
SPECIES LocaLity EXAMINED NUMBER (n) U. grandiflora Herculaneum, Mo. 2 7 Cliff Cave, Mo 3 7 Farmington, Ark. a 7 Schoolcraft, Mich. 1 7.
Hamilton, N. Y. 1 Faia U. perfoliata Great Smoky Mts., N.C. zZ v New Canton, Virginia 3 7
As shown in Figure 1 there are no evident differences in chromosome ‘size or configuration between the species, and there is only one case of an intra-specific difference, the plant examined at Hamilton, N. For this observation we are indebted to Dr. C. L. Stebbins, HO kindly made a cytological examination of local material. He found the char- acteristic chromosome complement for the genus, and in part of the microspores an additional fragment chromosome, much smaller than the others and similar to those reported by Belling (1925). Frag- mentation, therefore, can and does take place in nature as well as in the laboratory, though there is no evidence that it has been of any importance in the differentiating of species within the genius Uvularia.
ttl cccttl cecetll
Figure 1. Chromosome complements of Uvularia grandiflora (‘G’), U. perfoliata (‘P’), and Oakesia sessilifolia (‘S’), redrawn diagrammati- cally from the original camera lucida sketches.
The investigation was extended by examining the chromosome com- plement of the closely related Oakesia sessilifolia (L.) S. Wats. (Uvu- laria sessilifolia L.)
Plants were collected in the vicinity of Boston and the somatic chro- mosomes were studied in root tips. Fourteen chromosomes are present and it is possible to identify the seven types. Figure 1 (S) is a dia- grammatic drawing of the seven, made to scale from camera lucida sketches.
CoNCLUSION: There is no evidence that chromosome duplication or chromosome interchange are of phylogenetic importance in Uvularia.
30 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
Although these phenomena are known to have occurred under experi- mental conditions and may well occur in nature, they apparently have no direct evolutionary significance in this genus. So far as can be determined by cytological examination, specific differentiation in Uvu- laria has not involved gross differences in the number or relative size of the chromosome complement. This conclusion is strengthened by the fact that Oakesia sessilifolia, although usually classified in another genus, has a chromosome complement essentially identical with that of the two species of Uvularia. These results are of particular interest in view of the data presented in Part II, which show that U. grandiflora and U. perfoliata are well differentiated species. We do not mean to suggest that chromosome duplication and interchange are never of phylogenetic importance but merely call attention to the fact that there are groups of plants, like the genus Uvularia in which they are not species-forming forces. II. A MORPHOLOGICAL SURVEY OF UVULARIA
The following morphological survey of Uvularia grandiflora Sm. and U. perfoliata L. is an attempt to present objectively, in a codified form the essential facts as to resemblances and differences within and _ be- tween these two similar but distinct species. It is an attempt to repro- duce in a concise manner, for non-taxonomists, the kind of data which are consciously and unconsciously used by taxonomists in the delimita- tion of species. Fundamental biological unit though it may be, the Linnean species is still definable only by example (definable that is in terms which carry meaning to workers in other scientific fields). Those biologists who in the herbarium and in the field have had the oppor- tunity to study a number of such examples, have come to an apprecia- tion of the importance of these fundamental units and to a general notion of their relative magnitude, as compared with individual differ- ences on the one hand and generic differences on the other. There are, however, many non-taxonomists who though they do not have the time nor the training to participate in taxonomic work would like to gain a rough working knowledge of a Linnean species. If a species cannot as yet be defined in terms which are meaningful to workers in other fields of biology, one can at least present the range of variation within and between two closely related species in such a summarized form that the results may be digested in a comparatively short time.
The following pages present such a summary. It is not a taxonomic treatise on the two species. They were chosen for study not because their taxonomic position was in doubt, but for precisely the opposite
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 31
reason. ‘They were taken as an example because there seemed to be a very general agreement, on the one hand that they were specifically distinct from one another, and on the other hand that they were closely related members of the same genus. As far as practical taxonomic work is concerned, the data reported below merely confirm the cus- tomary disposition of Uvularia grandiflora and U. perfoliata as two distinct species of one genus.
After a preliminary study of herbarium material 15 sheets of each species were selected from the collections at the Gray Herbarium and were subjected to intensive study. Several times as many would have been desirable but we were obliged to limit the sample in order to obtain well-developed material in which the characters chosen for study could be measured accurately. The following collections were studied:
Uvularia grandiflora
Canada: Quebec, Lake Memphremagog, dry ground, J. R. Churchill, Aug. 17, 1914. Vermont: Middleburg, E. Brainerd, May 18- June 5, 1879; Shell House Mt., Ferrisburg, E. and C. E. Faxon, June 19, 1891; Hubbardston, W. W. E., no. 2, July 17, 1898; Hyde Manor, Sudbury, Geo. G. Kennedy, May 23, 1908; Manchester, M.A. Day, no. 179, July 7, 1898; near Hyde Manor, Sudbury, E. F. Williams, May 23, 1908. New York: North Fort Ann, Wash- ington Co., S. H. Burnham, May 30, 1920. Illinois: Starved Rock, La Salle Co., in rich woods, J. M. Greenman et al., no. 26, June 1-7, 1909. Wisconsin: St. Croix Falls, C. F. Baker, July 8, i900; J. M. Greenman, no. 2165, June 5, 1907. Missouri: Taney Co., rich woods, along rocky bluffs, EF. J. Palmer, no. 19237, Sept. 29, 1920; Creve Coeur Lake, M. W. Lyon, June 12, 1904; Clark Co., B. F. Bush, no. 2, Aug. 26, 1892; Cape Girardeau Co., rich woods along Miss. River, E. J. Palmer, no. 17998, June 21, 1920.
Uvularia perfoliata
Massachusetts: Sudbury, EZ. F. Williams, no. 1, May 30, 1900; Sudbury, EZ. F. Williams, no. 1, May 30, 1900; Granville, Hamp- den Co., deep rich, deciduous wood, F. C. Seymour, no. 191, June 24, 1914; South Georgetown, open woods, E. F. Williams, Aug. 1, 1899. Connecticut: Southington, shade, L. Andrews, no. 358, June 3, 1898; Cromwell, E. Wright, June 7,1878. New York: Tomp- kins Co., in woods, A. Gershoy, no. 7845, June 27, 1917; South Nyack, L. M. S., June 2, 1888. Pennsylvania: Berks Co., woods, eastern slope, H. B. Meridith, June 14, 1922; Delaware Co., woods,
32 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
A. MacElwee, no. 97567, June 16, 1899. New Jersey: Somer- dale, Camden Co., rich wooded slope, H. B. Meridith, May 27, 1921. Virginia: Bedford Co., A. H. Curtiss, Aug. 26, 1871. Dis- trict of Columbia: Washington, EZ. S. Steele, May 4, 1896. North Carolina: Waynesville, E. E. Magee, June 9, 1897; Asheville, B. L. Robinson, no. 75, Aug. 2, 1893.
A number of interesting characters could not be studied, either be- cause they were poorly preserved or because they were present on only a few of the sheets. For these reasons no detailed observations are re- ported on roots, seeds, fruits, and scale leaves (cataphylls).
Figure 2. Diagrams, to scale, showing node number, etc. in 5 speci- mens of U. perfoliata (above) and 5 of U. grandiflora (below). Further explanation in the text. '
STEM.
Five flowering specimens of each species are shown diagrammati- cally in Figure 2. A few poorly preserved details (such as the number of scale leaves at the base of the stem) have been ignored. The draw- ings are made to scale so far as internode lengths are concerned, and the angles of the branching have been approximately reproduced. The representations of the leaves and flowers are purely conventional. A
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 33
study of Figure 2 reveals the following differences between the two species.
U. grandiflora is somewhat larger, on the whole.
U. grandiflora has fewer nodes below the lowest branch.
U. grandiflora has more nodes on the lowest (sterile) branch.
U. grandiflora has a higher average number of flowers per plant. In U. grandiflora the upper end of the main axis is less perpen- dicular than in U. perfoliata.
BOO >
TABLE II. Number of nodes eee eae eee eae: | | Number of nodes with leaves | below branch | | | | U. perfoliata | 51h: tt | | | U. grandiflora 7|8 | Number of nodes on sterile branch | U. perfoliata 1;9|4]1 U. grandiflora 1 feel Fee
Of these five differences, B and C are relatively unaffected by the age of the plant and can be readily ascertained from herbarium speci- mens. They were chosen accordingly for more extensive investigation and are recorded in Table II, for all of the specimens under considera- tion. It will be seen that when a number of specimens are considered that neither character is discontinuous. This is true of all five of the characters of the stem enumerated above. After a little study any single specimen could be correctly placed, by a combination of these five characters, no one of which would suffice, if taken separately.
LEAVES.
Tracings of leaf outlines from the same five specimens are presented in Figure 3. To make the comparison more exact a particular leaf was
34
JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
chosen, namely, the leaf located at the axis of the main branch and the fertile shoot.
The following differences, or tendencies to differ, can be found between leaves of the two species.
Ficure 3. Leaf outlines from five specimens of U. grandiflora
(above) and five of U. perfoliata (below). The four features measured are shown on the lower left hand specimen.
A; B.
The leaf of U. grandiflora is a little larger on the average.
The leaf of U. grandiflora has a strong tendency towards some- thing like the outline of a wooden shoe, with a narrow diameter through the heel and a broad one through the sole. This was made the subject of a more extended investigation, the results of which are reported in a later section.
The leaf of U. grandiflora is usually pubescent on the under side. That of U. perfoliata is usually glabrous. Microphotographs of the leaf surfaces of 4 specimens of each species are shown in Plate 82. While this difference is a discontinuous one for the material considered in the present study, it would probably lose its discontinuity if a larger series of each species were examined. The specimen from Creve Coeur Lake, Missouri, was practically glabrous though it was in other respects a typical plant of U. grandiflora. Nor is there any evidence that it might have re- sulted from hybridization. The region in which it was collected
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 35
is far outside the natural range of U. perfoliata and the particu- lar locality is a semi-public reservation in which the original vegetation has been practically undisturbed.
D. Uvularia grandiflora has usually three veins in the leaf which are much larger than the rest. In U. perfoliata there are large veins and small ones, but there are seldom three which are clearly of another order of size than the remainder.
FLOWERS.
Photographs of representative flowers are shown in Plate 83. They exhibit a number of differences which are characteristic of the two species. If the flowers could be more easily examined without injury to the specimens, several of these characters would have been studied statistically. A survey of the 30 sheets shows the following differences.
A. Uvularia perfoliata has large glandular outgrowths on the side of the perianth segments. These outgrowths are not present in U. grandiflora.
B. In U. grandiflora the style is much longer in proportion to the stamens than in U. perfoliata.
C. The perianth segments of U. grandiflora are longer on the aver- age. The actual extremes were 2.4 — 4.7 cm. in U. grandiflora and 2.4 — 3.6 cm. in U. perfoliata. The respective medians for 7 specimens of each species are 3.2 and 2.7 cm.
D. The stamens of both species have a pointed sterile tip. This tip is on the average much longer in U. perfoliata.
STATISTICAL ANALYSIS OF DIFFERENCES IN LEAF SHAPE.
Four measurements were taken on each leaf, the points to be meas- ured having been chosen in such a way as to include the most out- standing differences in leaf shape between the two species. The four distances measured are indicated for the lower left-hand leaf of Figure 3. In interpreting this diagram it should be remembered that the leaf of Uvularia is perfoliate. The four measurements may be defined as follows:
a. The distance from the stem to the rear margin of the leaf, along
the median line.
b. The width of the leaf at the stem.
. The distance from the stem to the point of maximum width.
d. The maximum width at right angles with the mid vein.
The results are summarized in Table III. It will be seen that as far as absolute values are concerned, that there is very little difference
QO
36 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
between the two species. Quite another picture is presented by the ratios between these four measurements. There are six possible ratios and they are summarized in Table III. It will be seen that though in every case the two species overlap, their averages are quite different.
The three ratios in which the two species differ most markedly are a/d, b/c, and b/d. It was desired to utilize all three and in some way combine them into a simple ratio or index which would measure simul- taneously all three ratios. This was done quite simply by use of the following formula: leaf index —
a\2 b\2 b\2 VG) t+ @)
Figure 4 shows graphically the significance of the ieaf index in terms of analytical geometry. Analytically the index is a diagonal through a box, each dimension of which is determined by one of the three ratios. The ratio a/d determines the width of the box, the ratio b/c its height, and the ratio b/d its depth. Since U. grandiflora on the whole has
+ a/a——> Ficure 4. Explanation in the text.
lower values for all three ratios, the ‘“‘boxes” are generally smaller, and the diagonals through these “boxes” generally shorter than is the case in U. perfoliata. The calculated values for the leaf index are col- lected in Table III. While they still give an overlapping distribution for the two species, they come much nearer to separating them than does any simple ratio taken by itself.
By the addition of two values calculated from characters of the stem, we can separate the specimens statistically. To accomplish this in a
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 37
simple way, we need characters for which the values of U. perfoliata are usually higher than those of U. grandiflora and which have a range of values roughly similar to those of the three leaf ratios (0.1 to 1.3). Characters fitting these requirements were obtained from, (1) the num- ber of leaves below the lowest branch, and (2) the number of leaves on the sterile branch. In the case of the first, the actual number of leaves was divided by 10 and multiplied by 2, yielding values from 0.2 to 0.4 for U. grandiflora and from 0.4 to 0.8 for U. perfoliata. The new scale for the second character was obtained by subtracting the leaf number from 10 and dividing the result by 10. This yielded values of 0.2 to 0.6 for U. grandiflora and 0.6 to 0.9 for U. perfoliata. These two
A gee MEASUREMENTS B Aah peer on ESO C : os : D a x > t+ rn ~ * RATIOS 4 a X& — % penne LEAF INDEX INDICES meen GENERAL INDEX 10 20 30 40 50
TascLe III. Variation in leaf measurements, ratios of measurements, and index numbers for 15 specimens each of U. grandiflora (dotted lines ) and U. perfoliata (solid lines). The arrows show the limits of variation and the dots the position of the average values.
values may be added to the three ratios in the same simple way that was used in building up the leaf index, to form what we may call for lack of a better term the “General Index.” General Index = VY (a/d)? + (b/c)? + (b/d)? + (1)? + (2)? Analytically we are still dealing with diagonals and although they are diagonals in multi-dimensional space, they are still capable of linear measurement, and the lengths of these diagonals may be grouped on a single scale as we have done in Table III
It will be seen that the values of the “General Index” are completely discontinuous. In other words we have demonstrated statistically the
38 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
same point that was brought out in the graphical analysis of differences in the branching of the stem. The species in question may be sepa- rated by a combination of tendencies even when none of these tend- encies, taken by itself, will suffice.
Quite as important as the morphological differences in a presentation
Ficure 5. Distribution of U. perfoliata (open circles) and U. grandi- flora (black dots).
of specific characteristics are physiological ones. For the most part they cannot be studied effectively with herbarium material. That they exist has been shown by Wiegand and Eames (1925) in their report on the two species in their “Flora of Cayuga Lake Basin.” They state that U. perfoliata flowers two weeks or more later than U. grandiflora
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 39
and that it is found in “sandy acid or subcalcareous soil” while U. grandiflora grows in “gravelly or alluvial calcareous rich soils.”
The ranges of species are in part, at least, determined by their physiological constitution. This point can, of course, be investigated with herbarium material and the distributions of the species of Uvularia as determined by that method are shown in Fig. 5. It will be seen that throughout the general calcareous areas of the Middle West, Uvularia grandiflora is a common species while in New England it penetrates only to the limestone areas of western Vermont. U. perfoliata con- versely is a fairly common plant in New England and is found south- ward mainly along the mountains. There is an appreciable area where the two ranges overlap but for the most part they occupy different situations within that area.
DISCUSSION,
The above comparison of Uvularia perfoliata and U. grandiflora has shown four differences in the leaves, five in the nodes, and four in the flowers. Fundamental physiological differences were indicated by habi- tats and ranges, other morphological differences are known to exist in characters not available for study in our material. Of all the above differences, only three were discontinuous and there were indications that at least one of these would not have remained so if a larger series of specimens could have been examined. In the case of the intergrading differences, the two species could be separated by compounding several, no one of which would be sufficient if considered by itself.
If one attempts to think in general terms of the total hiatus between U. grandiflora and its nearest relative, he reaches a conception of spe- cific differences as a combination of many minor tendencies. The acknowledged discontinuity between the two species, taken in their entirety, is a discontinuity of combinations, reinforced by a few discon- tinuous differences in single characters.
It does not seem to be generally realized that species may be, and customarily are, thought of in two quite different ways by different groups of biologists. Those biologists engaged in purely taxonomic work will unavoidably think of species in terms of the precise differ- ences which permit their ready classification; which make it possible to arrange species in an herbarium or to construct a morphological key to a genus. To them the really essential differences between U. grandi- flora and U. perfoliata will be those few discontinuous ones which are ordinarily used in identifying the species.
40 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
With this attitude of mind the present authors have no quarrel recog- nizing taxonomy as a difficult and necessary business, and that those who have it in hand are to be thanked for doing the work and must be allowed to develop their own methods of going about it.
On the other hand there are those who are interested not so much in the technical nature of the classifying process as in the biological make- up of the units which are being classified. This group will include some taxonomists, for the separation of the two lines of thought is not absolute.
To us, members of the latter group, the difference between two spe- cies is the difference between one kind of germ-plasm and another. As geneticists currently think of germ-plasms made up of an enormous number of separate units, so the difference between species, even be- tween very closely related species, will be the resultant of a large num- ber of minor differences. The seat of these differences is certainly the individual cells and will be expressed not in any one or two characters. It will extend throughout the plant, being more sharply manifest in some parts than in others.
From this point of view such tenuous characters as aspect and tex- ture may have a very real biological basis. The characteristic leaf texture of a particular species, for instance, is ultimately dependent upon the size, form and arrangement, of the cells which make up the leaf. If the cells are small and regularly arranged, the leaf will have a close even surface; when the cells are irregular and large, the surface will be coarse and rough. Differences in aspect may in a similar way be traced back to the reactions of different germ-plasms.
Such specific differences, though slight and subtle, will on the whole manifest themselves similarly throughout the plant. Those internal forces which tend to produce small regular cells in the leaf will gen- erally have a similar effect in the stem, and in the flower. The late Edward L. Greene has been credited with having separated species on the basis of ‘‘a certain indescribable grace.” He may have been a good biologist, though a poor taxonomist, in relying on that distinction as an evidence of specific difference.
To us the many slight overlapping differences between species are more characteristic than the few sharply discontinuous ones. If we were forced to summarize the many differences between these two species of Uvularia in the fewest possible words, we could do no better than to say that U. perfoliata is neater and more delicate than U. grandiflora. In this way we might pack into a single phrase something
1934] ANDERSON AND WHITAKER, SPECIATION IN UVULARIA 41
of the impression we have gained of two different life-stuffs, each of which reacts variously with the environment. Individual plants pro- duced cooperatively by the germ-plasm and the environment, will show only one facet of the possibilities of that germ-plasm. For the full expression of the capabilities of a particular species there will be required a whole series of individuals produced under various environments. SUMMARY
Part I. 1. Differentiation of species in the genus Uvularia has not involved
gross differences in the number or relative sizes of the chromosomes.
Part II.
Uvularia perfoliata and U. grandiflora were chosen for a study of the nature of specific differences. A summary is presented of such differences as can be investigated in herbarium material.
bo
WwW
A statistical method is described which compounds two or more separate measures into a single index. By the use of this method it is possible to separate the two species by a combination of char- acteristics, no one of which would suffice if considered separately.
>
The nature of specific differences is discussed in the light of these results,
LITERATURE CITED
ANDERSON, E. and DIk D. G. Contributions to the Tradescantia prob- lem. te erie ren 13:213-231. 1932.)
BeLiinc, J. The origin of chromosomal mutations in Uvularia. (Jour. Genetics, 15:245-266. 192 e
BLa URN, K. B. On the ‘cage pay geographic races and poly- ploidy j in Silene ciliata Pourr. (Genetics, 18:49-66. 1933.)
MANGELSpoRF, P. C. and se R. me "Hybridization of Maize, Trip- sacum and Puchlaena Vee Hered. 22:329-343. 1931.)
WIEGAND, K. and Eames, A. J. The flora of Cayuga Lake basin, New York. prea Univ. ‘Agr. Exp. Sta. Mem. 92. 1925.)
42 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
EXPLANATION OF THE PLATES Plate 82
ower leaf surfaces (X +6) of Uvularia grandiflora (‘G’) and U. perfoliata (‘P’). The leaves in order from top to bottom are on the following 4 collections:
yon s. " erg aa
5 The leaves of U. Ae ade are from the following 4 collections : Wright s.n. 6/7/18 Williams s.n. £11809 Williams s.n. 5/30/1900 Meridith s.n. 5 5/27; 1921
e reader’s attention is called to the fact that this Plate is not so much a Epeaeron ition of points made in the paper as an attempt to present an objective summary of one kind of specific difference (namely leaf-tex- ture). So far as possible the same lighting, eines iea, and photo- graphic development have been used throughout. The lack of sharpness in the cuts of U. perfoliata is not inferior photography; it is due to the fact that that species has a smoot th lea with a somewhat waxy surface.
st
ter (presence or absence of hairs) which is “customarily used in keys descriptions.
Plate 83 age dens er (<6) of Uvularia waa an herbarium speci- me The stigmas have been retouched with ia
= U. perfoliata yan Burnham s.n. 6/2/1907 Lake George, N. Y.) = U. grandiflora (from Williams s.n. 5/22/1908 Brandon, Vt.)
ARNOLD ARBORETUM, HARVARD UNIVERSITY.
< ‘ Ss fo < — 5 - > 2
IN
OL.
Jour. ARNOLD ARB.
Jour. Arno_tp Ars. VoL. XV. PLATE 83
SPECIATION IN UVULARIA
1934] ANDERSON AND ABBE, DIFFERENCES IN THE BETULACEAE 43
A QUANTITATIVE COMPARISON OF SPECIFIC AND GENERIC DIFFERENCES IN THE BETULACEAE
EpGAR ANDERSON AND ErRnNsT C. ABBE
With one text figure
THE CRITERIA on which the sciences of Morphology and Taxonomy are largely based have two disadvantages. They are not objective and they are not commensurable. The first of these difficulties may be resolved through improvements in morphological technique. Such simple qualitative categories as ‘ovate’ or “obovate” are already quite as objective as the quantitative measures of chemistry or physics. Through closer analysis and more exact definition it may eventually be possible to bring the larger categories of taxonomy to a comparable degree of objectivity. Even with this improvement, the second diffi- culty would still remain; the categories are not reducible to a common basis. Is, for instance, the change from an ovate leaf to a lanceolate one, greater or smaller than the change from a lanceolate leaf to a linear? How do the differences between subgenera compare to generic differences on the one hand and specific differences on the other? To such questions as these the qualitative units of morphology permit no precise answers.
The following paper is a crude first attempt to work out a commen- surate method for dealing with such morphological and taxonomical problems. It is an attempt to carry on consciously and mathematically the same sort of process which with a good naturalist is subconscious and unmathematical. When a naturalist distinguishes between two species which are well known to him, he subconsciously takes the com- ponent of a very large number of variables. He knows Acer rubrum, for instance, not by any one technical distinguishing feature, but by the total impression of a very large number of features, any one of which may vary considerably in different individuals of Acer rubrum. In mathematical phraseology, his generalized idea of an Acer rubrum is the largest common denominator of all the individuals he has pre- viously recognized as belonging to that species. Any biologist who has had practical field experience with a species or a genus will probably agree with this definition. A slightly more objective demonstration
44 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
will be found in a recent survey of specific differences in Uvularia (Anderson & Whitaker, 1934).
If such a process is to be made conscious and turned into commen- surate units, we must attempt as far as possible to carry on the same sort of procedure mathematically. We must select a large number of features for measurement, features which by previous analysis are known to be individually significant, and then determine their aggre- gate effect mathematically. This paper is an attempt to measure the relative magnitudes of various categories in the Betulaceae by such a method. Six features are chosen and their aggregate effect is calcu- lated by a comparatively simple method whose essential features have been described elsewhere (Anderson & Whitaker, 1934). The sig- nificance of such an index calculated from diverse components may be made clear by a comparison with well-known indices in other fields. As a measure of business activity it is common practice to combine diverse categories such as car-loadings per week, electric power output, and the price of food-stuffs. With an intrinsically meaningless scale of this type, we measure the rise and fall of business from week to week or contrast the condition in one part of the country with that in another. In just such a way there are combined in the following analysis, six features of known importance in the Betulaceae:
—"
: the basic chromosome number
the number of perigon segments in the female flower
the average number of stamens per flower
the total number of bracts in a bract-complex
the number of flowers per pistillate bract-complex
the maximum number of cell rows in the rays of the secondary xylem.
Am W DH
Table I shows the actual data used for the 40 species of the Betu- laceae for which figures were readily available. It should be noted, however, that in the case of the perigon segments of the ovary, in the number of bracts in a bract-complex, certain factors must be taken into account. Thus there is anatomical evidence for the presence of vestigial perigon segments in the pistillate florets of some species of Alnus (Abbe, 1933). In the absence of a satisfactory means of indicating this mathematically they have been uniformly treated as indicated in Table I. The number of stamens per floret introduces perhaps the most variable character possessed by the members of the family. Here the data are based on representative individuals rather than on a math- ematical average. These individuals have been chosen to illustrate the
1934] ANDERSON AND ABBE, DIFFERENCES IN THE BETULACEAE 45
TABLE I* — No. of | No. <P ; Bracts per Florets Maximum ; Basic (Female) No. of ae resi Pistillate No. of Cell Genus & Species Chromo- Perigon Stamens Brac Rows in ‘ egments per Floret Copies Rone the Ray Alnus SALNOBETULA A. crispa 0) 5 4 2 1 §GYMNOTHYRSUS A. incana 7 ) 4 5 2 1 A. cordata 7 @) 4 5 2 1 A. maritima i ) 3 5 2 1 A. rhombifolia 7 0 3 5 Z 3 A. hirsuta 7 ) 4 5 2 1 A. rubra 7 0) 4 5 2 1 Betula SEUBETULA §§ CoSTATAE B. nigra 7 ) 2 3 3 4 B. utilis 7 ) 2 3 3 Z B. Schmidtii 7 0 2 x 3 3 B. grossa 7 0 3 3 3 3 B. costata 7 @) 3 3 3 3 B. lenta 7 0 3 3 3 4 B. lutea if 0) 3 3 3 4 §§ Nanae B. Michauxii 7 0 1 3 1 1 B. pumila 7 0 2 3 3 3 §§Albae B. pendula 7 0 2 3 3 3 B. japonica 7 0 2 5 3 3 B. populifolia 7 0 2 3 3 3 B. pubesc 7 0 2 3 3 + B. papyrifera 7 0 2 3 3 3 B. davurica 7 0 2 3 3 3 §BETULASTER §§ ACUMINATAE B. Maximowicziana 7 0) 3 3 3 3 B. alnoides 7 ) 2 1 3 3 orylus Cc. Roni 7 4 2 3 2 1 C. americana 7 4 2 3 2 il C. heterophylla 7 4 2 3 2 1 C. Colurna 7 4 2 3 2 1 C. cornuta 7 4 2 3 2 3 Ostryopsis O. mph creey 8 4 2 1 2 1 SiisTEROCARPUS C. japonica 8 o 6 2 4 G. data 8 6 1 5 3 SEUCARPINUS C. betulus 8 4 4 1 2 4 C. caroliniana 8 4 4 1 2 3 C. laxiflora 8 4 3 1 2 5 C. Tschonoskii 8 4 3 1 2 3 C. orientalis 8 4 4 i 2 2 O. ice ag 8 4 4 1 2 3 O. japo 8 4 4 1 2 2 O. jarainifolis 8 4 3 1 2 2
specific names are based on Rehder (1927). The subgenera, etc., are based on ane (1904).
46 JOURNAL OF THE ARNOLD ARBORETUM
[VOL. Xv
salient trends of differentiation within the group. While the number of bracts in a given staminate bract-complex is generally a relatively constant character for a genus or a species, it sometimes, as in Alnus crispa and Carpinus japonica, is variable.
sary to choose a significant figure, in the absence of averages. basic chromosome numbers are taken from Woodworth (1931).
Here again it has been neces-
The The
differences between species were calculated from the data in Table I
by the formula:
TABLE II
SUMMARY OF INTRAGENERIC, INTRASECTIONAL DIFFERENCES
Names of pau and Sectio
ALNUS oo BETULA 8§Cos 8§N
af 8A Ibae ae
ae §§Acuminatae
CoryLus
Carpinus §Distegocarpus i Eucarpinus
OstTRYA
0-1.0
Range of Differences
1.1-2.0 2.1-3.0 3.1-4.0 4.1-5.0 5.1-6.0 6.1-7.0 Total Av’g
4 7 2
1 1
1
pnenaiel OF INTRAGENERIC, INTERSECTIONAL DIFFERENCES
§AIngbetla—$Gymnothyrss Bet §§ Costatae—§ § rae —§§A
e win eater g§Nanae—$8 §Albae cuminatae §5Albae—§ ee minatae Carp Sie ae 8 Eucarpinus
SUMMARY OF INTERGENERIC DI
ALNUS—BETUL A Cory
—OsTRYOPSIS He ente US —OsTR Berura—Corvte S —— —CAaRPINUS (a9 —QOSstTrR YA CoryLus—Ostrvopsis “
Ostrvopsis—C — YA
Cue.
bo me AMO U1
4 2
4 3
13
3 6 5
1
6 1
1 6
6
ANAA WU
2 1 2 3
FFERENCES
15 4
18 7
2
13
20
20 5
3 14
10 12
9 16
14 6 2 9
1 1 1 3
1934] ANDERSON AND ABBE, DIFFERENCES IN THE BETULACEAE 47
i = 2 2 2 2 2 2
Aggregate difference V AP+taAs+ alt aZ+Ait+ A,
Thus the aggregate difference between Alnus cordata and Alnus mari- tima is calculated from the data in Table I as follows:
V (7-1)? (0-0)? (43)? (55)? (2—2) 7+ 1) =V0+0+1+0+0+40
—V1 == 1.0
The difference between Alnus cordata and Carpinus japonica becomes:
V (7--8)?-+ (04)? (46)? (53) (22)? (4-1) =V1-—-16+4+4+4+049
= /ea
—20o Analytically the magnitude of the difference is the length of a line between two points each of which is defined in six dimensions. Mathe- matically it would not be necessary to give the index a graphical inter- pretation, but many biologists find it easier to think geometrically than arithmetically.
In Table II, the magnitudes of the differences between the various species are tabulated and summarized. They are considered within sections of the genera, between sections within genera, and between genera. For the inter-specific comparisons all the possible combina- tions were calculated when the numbers of species in the genera under comparison were small. In the larger genera, 25 comparisons were chosen by chance, and the average calculated from these choices.
Table II] shows a number of interesting facts:
1. The differences between genera are of greater magnitude than the differences within genera.
2. The differences between species within the same section are of relatively the same magnitude throughout the Betulaceae. The average difference in Ostrya, on the scale of the index, is 1.1; in Betula, 0.8; in Carpinus, 1.2; in Corylus, 0.8; in Alnus it is 1.0. This is an indication that the qualitative work of the taxonomists and morphologists who have dealt with the group has been surprisingly constant from genus to genus. This is particularly interesting in view of the fact that the six features used in computing the index are far from identical with those used in taxonomic work. Two of them indeed (Nos. 1 and 6) bear no direct relationship thereto.
48 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
Using the average distances between genera as obtained in the last column of Table II, a three-dimensional model was constructed. Just as the distance between any two points can be represented on a single straight line, but the distances between any three may require a plane for graphical representation, so in this problem we are entitled to five dimensions to show the relationships between our six genera. For- tunately the relationships are simple enough so that three dimensions will give a working approximation. A study of the model reveals sev-
Figure 1. Models showing the comparative distances between the genera of the Betulaceae and the sub-sections of the genus Betula. For the larger model, AL, Alnus; CA, Carpinus; CO, Corylus; OP, Ostry- opsis; OS, Ostrya. For the smaller model, N, §88$Nanae; A, §§Albae; B, 8§ Acuminatae ; C, §8$Costatae.
eral interesting points (Fig. 1). It should be borne in mind that the lines connecting genera in the model are not meant to diagram phylo- genetic lines of development. They indicate nothing more than com- parative distances between genera as determined by the method de- veloped above.
(1) There are two groups of genera, Carpinus — Ostrya — Ostry- opsis at one end of the family and Alnus — Betula at the other. Corylus occupies a position off at one side and not clearly to be iden- tified with either of the other groups.
(2) Carpinus, Ostrya, and Ostryopsis form a straight-line sequence.
1934] ANDERSON AND ABBE, DIFFERENCES IN THE BETULACEAE 49
That is, the change from Carpinus to Ostrya, carried on still further in the same direction, will reach Ostryopsis. The model, of course, gives no hint of the direction of this apparent phylogenetic sequence; whether it is from Carpinus to Ostryopsis, or vice versa.
Betula was the only genus for which the intersectional figures were complete enough to warrant detailed comparison between sections and genera. A model of the sub-sectional relationships in Betula on the same scale as the other model is shown in Figure 1. The magnitude of sub-sectional differences in Betula is clearly less than that of the differ- ences between Betula and its closest relatives. The Nanae sub-section, however, as measured by this index shows a divergence from the other three sub-sections greater than that between Ostrya and Carpinus.
SUMMARY
(1) The qualitative nature of taxonomic units is briefly discussed. 2) A quantitative method is advanced for dealing with these units, and generic and specific differences in the Betulaceae are compared. It is shown that the latter are much smaller, and that they are approxi- mately equal in average magnitude from genus to genus.
(3) Models are constructed showing on the same scale the compara- tive divergence between the genera of the Betulaceae and of certain sub-sections in the genus Betula,
—_—
LITERATURE CITED
ABBE, Ernst C. (1933). The inter- ee of the genera of the Betulaceae, based on anatomical stu of the inflorescence, as flowers, and the secondary xylem. (Thesis, pene University [M
ANDERSON, EDGAR WHI eae ares W. (1934). one in eas Pia Arnold rey re
Woopwor He( 1931), Bolyploidy in the Betulaceae. (Jour. Ar-
nold Arb, 12: 506217)
ARNOLD ARBORETUM HarVARD UNIVERSITY.
50 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
THE BIOLOGY OF MILESIAN RUSTS J. H. FAu iyi
With plates 84 to 86
As PART of a contribution towards a monograph on Milesia I have recently presented an account of the taxonomy and geographical dis- tribution of the species of this genus (2). The paper now offered, the third of a series on the Pucciniastreae, is devoted, except for a short section on economic considerations, to topics on the biology of Milesian rusts. These topics comprise (1) hosts, (2) life history studies, (3) developmental periods, (4) habits of spore production, (5) host restrictions.
HOSTS
Ferns and ferns only serve as hosts for the diploid generation of all known species of rusts belonging to the genus Milesia. These are repre- sentatives of about sixteen genera scattered among various subfamilies of the one family, the Polypodiaceae. This limitation is an interesting phenomenon the significance of which can only be conjectured because the Polypodiaceae are regarded as the most recent family of homo- sporous Filicales. The genera and the subfamilies to which they belong are as follows: AcRoSTICHEAE - Elaphoglossum; ASPIDIEAE - Dryop- teris, Polystichum; ASPLENIEAE - Asplenium, Blechnum, Diplazium, Scolopendrium,; DAVALLIEAE- Nephrolepis, Odontosoria; PoLyPo- DIEAE - Drynaria, Polypodium; PTERIDEAE - Coniogramme, Crypto- gramma, Histiopteris, Pellaea; Woovst®Ark - Dennstaedtia. Lacking from the list are but the two small subfamilies ViTTARIEAE and OLEANDREAE,
In addition to the fact that the fern hosts compiled above are all polypodiaceous they are characterized in common by the general habit of ‘overwintering’ some of their fronds in a more or less green condi- tion. Recognition of this feature is important both to the collector and the student of Milesian rusts. This fact is emphasized because asso- ciated with it is the postponement beyond the current season, typical of so many species, of the formation of teliospores; they develop with exceptions in the spring on overwintered, affected fronds. Associated also is the production of crops of uredospores (often the most prolific)
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 51
in the spring on the overwintered, affected fronds. The latter phe- nomenon explains how it is that a species of Milesia may be perpetu- ated indefinitely without the necessity of teliospores and a haploid stage; it likewise provides a basis for a reasonable understanding of the world-wide distribution and the course of evolution beyond the range of the one known generic alternate host of this remarkable group of rusts.
Alternate hosts have been more or less satisfactorily determined for ten (nine if .V. fructuosa is the same as M. intermedia) species of Mile- sia, and in every instance they have proved to be species of Abies. Heretofore no consistent effort has been made to discover how many species of Abies might be susceptible to any one particular species of Milesia. But as Klebahn (7) found that A. alba and A. cephalonica, the only ones he tested, served as hosts for M@. Blechni, and Kamei (4) found that A. Mayriana, A. firma and A. sachalinensis, the only ones he tested, served as hosts for M. exigua, there appeared to be ground for anticipating that host restrictions are not prevalent in firs to any- thing like the same extent as in ferns. The results of experiments re- corded in this paper in which M. fructuosa was tested out on a wide range of fir species add confirming evidence to this supposition. In addition to the nine or ten species whose alternate hosts are known there are approximately fifteen species whose teliospores have been de- scribed but for which as yet the haploid stage remains to be discovered. All but one of these occur within the range of Abies; they include such well-known species as M. murariae, M. Feurichii, M. Polystichi. It is almost certain that they, too, pass to Abies. Regarding the, remainder of the species, those for which uredospores only are known, any from north temperate regions, such as M. Magnusiana, doubtless produce teliospores and likely have Abies as alternate host. Naturally there is uncertainty regarding the haploid hosts of those species that are re- stricted to areas beyond the range of Abies. As a matter of fact our acquaintance with such species is still so limited that we know of the existence of teliospores in but the one species, M. australis. Elsewhere (2) I have advanced the hypothesis that at least some of the Milesian species of the tropics, subtropics and Southern Hemisphere may be imperfect species, evolutions from the diploid phase of perfect species and perpetuated solely by means of uredospores. In any case it is clear that a great deal of fascinating observation and experimentation re- mains to be done on host relationships of a large proportion of the species of Milesia.
52 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
LIFE HISTORY STUDIES
The first successful study of the complete life history of any species of Milesia was made by Klebahn (7); the species studied was M. Blechni (Syd.) Arthur. Professor Klebahn employed as primary inocu- lum telial material from overwintered fronds of Blechnum Spicant (L.) With. collected in the spring of the year. At various dates between May 16 and June 5, 1915, portions of the inoculum were suspended over small potted plants of Picea Abies (L.) Karst., Abies alba Mill. and A. cephalonica Loud.; these were then covered by bell jars and kept covered for as long as six days. Picea remained free from infec- tion; but on Abies evidences of aecial formation began to appear June 21 on the earlier inoculated plants, and the first mature aecia were observed July 10, or 55 days after the experiment was started. This interval may be taken as an approximately correct measure of the de- velopmental period of the rust on Abies, for while it would actually be shorter at least by the time that elapsed between the setting of the inoculum in place and the shedding of the basidiospores, yet this could not have been more than a few days at most because the inoculating material in any instance was not in use for more than six days. All of the experimental plants of both species of Abies produced a copious crop of aecia, but on needles of the current season only. Some of the aecial-bearing needles were distributed in Jaap’s Fungi selecti exsiccati under number 774a; the remainder were used for the inoculating of ferns of the species Blechnum Spicant (L.) With., Dryopteris spinulosa (O. F. Miller) Kuntze and Scolopendrium vulgare Smith.
The ferns were inoculated at various dates beginning July 20. A marked degree of infection resulted on Blechnum, the plants first inocu- lated showing clear indications of rust along in August; the develop- mental period in all instances was protracted. No mention was made of any signs of infection on Dryopteris. Injuries were observed on Scolopendrium, but their cause was not determined; somewhat similar injuries on Blechnum were ascribed to the effects of manipulation of the fronds during the course of the experiment. These cultures were supple- mented by others in which uredospores from pustules on overwintered fronds of B. Spicant were transferred under suitable conditions to new, healthy fronds of the same species; the results were positive. Thus, inoculations made June 21 were followed by the production of uredinia which began to sporulate July 12, indicating a developmental period of three weeks. Summarizing, Klebahn established the outstanding facts that Abies, equally true for both species tested, serves as the alternate
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 53
host of Milesia Blechmi, and that this rust is capable of perpetuation from year to year on its fern host alone without the alternation of the aecial stage.
Life history studies of certain other species of Milesia have been reported more recently by Faull (2) and by Kamei (4, 5, 6); in addi- tion, Dr. Eugene Mayor informs me that he has obtained uredinia of M. Kriegeriana (Magnus) Arthur by sowing aeciospores of a white- spored Peridermium on fronds of Dryopteris spinulosa (QO. F. Miller) Kuntze, but his publication has not yet appeared.
To Kamei we owe accounts of determinations of the host interrela- tionships of Milesia exigua Faull (Milesina vogesiaca Kamei, not Sy- dow), Milesia jezoensis (Kamei) Faull, M. Miyabei (Kamei) Faull
The telial materials of M. exigua with which Kamei initiated his studies (4) of that species were collected in the fall from fronds of the current season of Polystichum Braunii (Spenner) Fée. They were placed in cotton bags and hung on the shady side of a wall to over- winter. The following spring, at the time the new leaves of Abies were unfolding, portions of the fronds carrying teliospores were placed in Petri dishes lined with moist paper and kept there until basidia began to emerge-—an interval of two or three days as it proved to be. The pieces so treated were next placed on the foliage of potted firs freshly sprayed with water, and all covered with bell jars for two or three days. After the bell jars were removed the inoculated firs were trans- ferred to a cool place and watered daily. In the course of his experi- mentation three species of Abies were tested, namely, A. Mayriana Miyabe & Kudo,’ A. sachalinensis Mast. and A. firma Sieb. & Zucc. Infection resulted in all three species and in the new needles only. Spermogonia appeared in from 13 to 20 days following inoculation, and aecia in from 21 to 28 days. Six out of nine tests on A. Mayriana were successful, one (spermogonia only) out of three on A. sachalinensis and one out of two on A. firma. From aecia that formed on A. Mayriana and A. firma one sowing of aeciospores each was made on fronds of P. Braunii, the first on pinnae of a potted plant, the second on detached pinnae in a Petri dish. Uredinia developed within 17 and 13 days respectively. While we might wish that the data were more volumi- nous we can unreservedly accept Kamei’s conclusion that M. exigua alternates between Abies (three species tested) and P. Braunit.
Milesia jezoensis was the next species studied by Kamei (5). Telial
14bies Mayriana is often considered a variety of A. sachalinensis - A. sachalinensis var. Mayriana Miyabe & Kudo.
54 JOURNAL OF THE ARNOLD ARBORETUM [VOL. Xv
material was collected in the spring on overwintered fronds of Poly- podium vulgare L. The portions to be employed as inoculum were placed in a moist chamber and left there until basidia could be de- tected—a period of two or three weeks. Just what the unexpected length of this interval means is not clear, that is, whether the teliospores were immature at the time of collection or that in this species they require a resting period before germination will take place or that the early growth of the basidia is exceptionally slow or that conditions suitable for immediate rapid growth were not provided. As soon as the basidia emerged the inoculum was suspended over the foliage of small, potted trees of A. Mayriana on which the new needles had recently expanded, the same technique being employed as for M. exigua. Four out of nine tests were successful, and in these spermogonia appeared on needles of the current season only and within about 20 days. For some reason, probably because of poor condition of the firs, no aecia developed and most of the infected needles dropped before the end of the summer. Naturally because of the lack of aeciospores inoculations in the opposite direction, that is, from Abies to Polypodium, were im- possible. While the incompleteness of this experimentation and the paucity of data obtained call for a repetition of the work, yet Kamei’s conclusion as to the host connections of M. jezoensis may be reasonably accepted as correct.
Kamei has likewise reported (6) partial life history studies of M. Miyabei and M. Dryopteridis, but his published data on these are par- ticularly meager. For both species the telial materials employed as inocula were collected on overwintered fronds—on Dryopteris crassirht- zoma Nakai for M. Miyabei and on M. viridescens (Bak.) Kuntze for M. Dryopteridis. Inoculations were made on Abies Mayriana and in due course lesions appeared, but on the new needles only. Eventually these bore spermogonia and aecia in the case of M. Miyabez, but sper- mogonia only in the case of M. Dryopteridis. For the latter the devel- opmental period up to the production of spermogonia was about 38 days; no time records were given for M. Miyabei. Here again, while the results warrant the reasonable conclusion that A. Mayriana may carry the haploid stage of M. Miyabei and M. Dryopteridis, fuller and more complete data are desired.
My own researches have pertained to the species Milesia intermedia Faull (probably a condition of M. fructuosa Faull in which the ure- dinia are omitted or if present at all are much reduced), M. fructuosa Faull, M. marginalis Faull and Watson, and M. polypodophila (Bell)
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 55
Faull. These were conducted during the summers of 1924-7 at the University of Toronto Field Laboratory of Forest Pathology located on Bear Island, Lake Timagami, Timagami Forest Reserve, Ontario, and during the summer of 1933 at the Arnold Arboretum; Harvard University, Jamaica Plain, Mass. In the former locality all these rusts are readily obtainable in ample quantities on their fern hosts; and their coniferous host, Abies balsamea (L.) Mill., abounds. In the latter locality neither the fern hosts nor spontaneous plants of Abies occur; the material used as inoculum was collected on ferns in the Berkshire Mountains of western Massachusetts, and the firs inoculated were trees that at one time or another had been planted in the Arboretum.
Studies on each of the species began with the telial stage. At the time the new needles of Abies were expanding overwintered fronds bear- ing teliospores were brought into the laboratory, and, so as to force spore germination, they were placed between folds of wet paper or laid on damp sphagnum moss in large loosely covered cans. Small test pieces were also kept under moist conditions in Petri dishes to afford greater ease of observation. Close watch was kept for the emergence of basidia and the first formation of basidiospores. As soon as basidio- spores were freely appearing the material was ready for making inocu- lations. In setting up an experiment a portion of a frond carrying inoculum was either laid, abaxial surface down, directly on the foliage of a small branch of the host to be inoculated, or it was first prepared for easier handling and greater economy in use by laying it between two strips of galvanized wire netting, separated from the upper strip by a piece of wet absorbent paper. The chosen branch by way of its preparation was sprayed with water from an atomizer and so flexed that its lower face would come in juxtaposition with the basidial-bearing face of the piece of frond. A celluloid cylinder about six inches long by two inches in diameter (E. E. Hubert—Celluloid cylinders for inocu- lation chambers. Phytopath. 6:447—-450. 1916) plugged at one end with wet sphagnum was then slipped over the branch and its adjust- ment completed by packing its proximal end around the branch in its axis with wet sphagnum. Whenever necessary the branch was propped to keep it as nearly as possible in a horizontal position until the tube was finally removed some days later. The experiment was visited daily as long as the tube was in use, and the sphagnum plugs were kept wet by the addition of water as needed. Similar technique was employed in subsequent inoculations of ferns except that the inoculum consisted of a suspension of spores in water which was sprayed on the fronds.
56 JOURNAL OF THE ARNOLD ARBORETUM
TABLE 1. Milesia intermedia—
from Dryopteris spinulosa intermedia to Abies balsamea
[VOL. Xv
First ap- First 0. of Date of First ap- pearance pearance No. of needles inocula- pearance of spermo- of peri- Date needles with peri- tion of lesions gonia dermia harvested infected dermia 27. vi.24 14 days — 36 days 40 days 50 33 27. vi.24 14“ — 360 40 “ 100 80 27. vi.24 14“ —_— 34 45: * 150 140 27. vi.24 14“ — 34“ 40 “ 90 80 28. vi.24 13 * a 357: °F 39 32 26 28. vi.24 sk — 33°-* 39 24 21 28. vi.24 12.” — 33°." 41 “ 100 90 28. vi.24 12 * —_ 33. 41 ‘ 90 80 28. vi.24 12 “ _— oa. 39 100 80 4.vii.27 14 * 16 days 34 43 “ — 2 4.vil.27 13. 15 “ 38 43“ — aa 4.vi1.27 13“ 15 * 34“ 43% — 8 4.vil.27 19 19 “ 34“ 43 —_— 5 4.vii.27 19 “ 21 “ Ae 43 * — 5 4.vii.27 13 “ 19‘ 32° 43 * ~- 13 4.vil.27 13° -* 19 “ 34. 43 “ —_ 15 4.vii.27 — — 34 43“ — 5 4.vil.27 16. * 19 * 35 43 “ — 7 5.vil.27 15‘ 20 * 34“ 45 * _— 10 5.vil.27 13.‘ 18 “ a4: = 45. — 17 5.vii.27 13: * 20“ 34 45 “ — 17 5.vil.27 18 “ 20: “** 34 45 “ — 8 5.vil.27 10 ‘ 20 “ 36°“ 45“ — 5 5.vii.27 18 “ 18 “ 37. * 45“ —_— 7 5.vil.27 18“ 20 “ —_— — = 0 6.Vil.27 17,“ 17 * 33:. 45 * — 13 6.vi1.27 : 14“ aa 45 “ — 11 6.Vil.27 11 “ 14“ 33." 45 “ — 23 6.vil.27 11 17 “ 36 45 “ — 2 6.vil.27 14“ 17 $3 45“ — 5 6.vii.27 12. “ 17“ 33.-** 45 “ — 16. Averages 14 * 18 “ 34
Average no. of peridermia per needle bearing peridermia—7.
Peridermia ruptured in from 1 to 6 days or an average of about 3 days.
Percentage of needles infected varied from 1 to 33.
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 57
All of the experiments were conducted out-of-doors. Two sets of experimental plants were used in the Timagami experiments, one set consisting of well-established potted plants located in a protected “gar- den” where natural infection was altogether unlikely (in our experience none ever occurred) and under as nearly normal conditions as possible; and a second set consisting of undisturbed plants in a favorable natural habitat, in localities well removed from occurrences of Milesian rusts. The latter set likewise remained free from natural infections. In the Arnold Arboretum experiments the undisturbed trees scattered about the Arboretum were used; an auxiliary set was not necessary because these trees are entirely free from Milesian rusts and far removed from any possible source of spontaneous infection. Uninoculated branches and uninoculated plants in the immediate neighborhood served as con- trols. Data on the experiments are assembled in Tables 1-7.
TABLE 2. Muilesia intermedia—
from Abies balsamea to various ferns
Date of Date Fern inoculated Inoculation Harvested Results Dryopteris spinulosa intermedia 12.viii.24 5.vi.25 III ms : a 16.vili.27 27.vi.28 III “ S vs 12.viii.24 5.vi.25 III a ss es 24.viii.24 5.vi.25 Ill Dryopteris spinulosa 16.viii.27 30.vi.28 Ill “ oe 12.viii.24 5.vi.25 Ill Dryopteris marginalis 16.viii.27 30.vi.28 No infection = Re 19.viii.27 28.vi.28 A rs sh 19.viil.27 28.vi.28 as . ie 19.vili.27 28.vi.28 - : ve ss 20.viii.27 27.v1.28 - rs Dryopteris cristata 16.vili.27 28.vi.28 No infection Drvopteris fragrans 20.viii.27 29.vi.28 No infection - 20.viii.27 29.vi.28 .: oe a “ 20.viii.27 29.vi.28 s - o 20.viii.27 29.vi.28 § t Ss 20.viii.27 29.vi.28 ? Polystichum acrostichoides 16.viii.27 27.vi.28 No infection ae 19.vili.27 27.vi.28 . x
The aeciospores used as inoculum were produced in the experiments recorded in Table 1.
58 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
TABLE 3. Milesia marginalis—
from Dryopteris marginalis to Abies balsamea
First ap- First ap- No. of Date of First ap- pearance pearance No. of needles inocula- pearance of spermo- of peri- Date needles _ with peri- tion of lesions gonia dermia harvested infected dermia _ 27. vi.24 13 days -—— 47 days 53 days 180 160 25. vi.24 zi — 51 55 “* 14 14 25. vi.24 19. * ~= 51 “ 55 * 25 25 27. v1.24 13. a 42 “ 53: = 150 130 27. v1.24 13. “ -—— 42 “ 53: 21 20 27. vi.24 3. * — 42 “ 55. 400 310 28. vi.24 13: * — 46 “ 54“ 170 130 28. v1.24 13. * — 46 “ 52. “ 380 270 28, vi.24 13.“ -—— 46 “ 52 * 250 190 28. vi.24 13 = — 41: -* 52 “ 200 170 28. vi.24 13. CO — 41 “ S2- 290 250 28. v1.24 13. * -—— 41 * 52% 190 140 28. vi.24 is — 43 54“ 330 260 28. v1.24 13. * —_ 41 “ 52. % 55 50 28. v1.24 13° “* 41 “* 52. + 260 210 29. vi.27 12 * 18 days 40 “ 54“ = 105 29. vi.27 18 “ 18 “ 5.°* 54“ — 8 29. v1.27 14 * 19: * 43“ 54 “ — 19 29. v1.27 12 “ 18“ 42 “ 54“ — 58 29. v1.27 12 “ 18 “ 43“ 54 * co 21 29. vi.27* ya 18“ 39 54“ — 106* 29. v1.27 12. = 18 “ 30. * 54.“ — 87 29. vi.27 14 “ 19 “ 40 “ s7 a 52 29. v1.27 14 “ 19 “ 42 “ 57. * — 77 29, v1.27 12 * 19 42 * ny — 26 29. vi.27 14 * 19“ 42 “ by — 57 29. vi.27 12 “ 18 “ 42 “ 57 % —_ 83 29. v1.27 14 “ 18 “ 45 “ 57 ” — 26 29. vi.27 12 -* 19“ 45 *“ 7 — 41 30. vi.27 18 “ 23:.% 42 “ 57 — 40 30. v1.27 18 “ 21 .* 44“ 57“ — 43 30. vi.27 15.“ 23° S* 42 “ Ss; < — 51 5.vli.27 12 “ 20 ‘ i a 52. — 21 5.vii.27 4 =“ 20 * 45 “ sz °° a 3 5.vii.27 Fy 2G, -& 45 “ 52. * — 4 5.vii.27 12: * 20° * 40 “ 52 —— 29 5.vii.27 12. 20: # 41‘ 52. >. m- 25 5.vii.27 2. * 20.“ A1.: 3 52. — 34 Averages 14—“ 19+“ 43 Material was ready for inoculation one day after ses oma Average no. of peridermia per needle bearing peridermia—7. Peridermia ruptured in from 1 to 8 days or an nse “of about 4 days
Four expe ; Percentage of needles infected varied from 2 to 65, an average of 33%. Percentage of needles infected when inoculation was at best, an average of 40%.
*Plates 84 and 85 are photographic records of this experiment.
1934]
TABLE 4.
Milesia marginalis—
FAULL, BIOLOGY OF MILESIAN RUSTS 59
from Abies balsamea to various ferns
Date of Date Fern inoculated Inoculation Harvested Results Dryopteris marginalis 20.vili.24 22.vi.25 «XII, III abundant a a 20.viii.24 5.vi.25 P es s fs 24.viii.24 Savizs; “Lt oe . 22.vili.27 27.vi.28 II, III very ae oe 22.viii.27 27.vi.28 II, III abun cS os 25.viil.27 28.v1.28 IL. II mea « es 25.vili.27 28.vi.28 II. was injured by animal s “ 26.viii.27 28.vi.28 ‘II, III i “ 26.viii.27 29.vi.28 ‘IT, III abundant Dryopteris cristata 23.viii.24 22.vi.25 No infection 23.viil.24 22.vi.25 a a - ok 23.vill.24 22.vi.25 a = oe es 23.viil.24 22.vi.25 Me i: ‘ o 22.viii.27 28.vi.28 Si a eS a 25.viii.27 28.vi.28 a 3 Dryopteris fragrans 25.viil.27 29.vi.28 No infection “ 25.viii.27 29.vi.28 i se sf se 27.viii.27 29.vi.28 an ie f oh 27.viii.27 29.vi.28 a . Dryopteris spinulosa 23.vili.24 22.vi.25 No infection : 22.Vii1.27 29.vi.28 < it = ss 22.viii.27 30.vi.28 st “
a - 25.viii.27 27.vi.28 2 se Dryopteris spinulosa intermedia — 20.viii.24 22.vi.25 No infection i ss Y: 20.viil.24 22.vi.25 : “
SS ae + 22.vili.27 27.vi.28 ~ sp Polypodium virginianum 25.vili.24 22.vi.28 No infection - 25.vili.24 22.vi.28 : Polystichum acrostichoides 22.viii.27 27.vi.28 No infection
= : 25.viii.27 27.vi.28 de “¢ se a 26.viii.27 27.vi.28 “ .
The aeciospores used as inoculum were produced in the experiments recorded in Table 3.
60 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
TABLE 5. Milesia polypodophila—
from Polypodium virginianum to Abies balsamea
Date of First appear- First appear- First appear- inocula- ance of ance a Sa ance of peri- i lesions mo dermia Dates of harvest 21.vi.24 Summer of June, 1927 25.vii.27 (on 7.viii. if a 1926 needles of oO 1924) (Herb. as 21.vi.24 Summer of June, 1927 25.vii.27 (on 25.viii.28 (needles 1926 needles of of 1923-5 1924) (Herb. #8596) 21.vi.24 Summer of June, 1927 19.vii.27 (on 28.vii.27
needles of (Herb. #8380)
1 12.vii.28 es of 1924-5)
(Herb. #8508)
The telial material of M. polypodophila, on overwintered fronds of Polypodium virginianum, was collected June 19, 1924. The fronds were placed at once in a moist chamber. There was an abundance of basidiospores in evidence two days later, and that same day, June 21, fourteen experiments were set up. The inoculated plants were closely examined from time to time during the remainder of the summer, but no signs of infection could be detected. Likewise, during 1925 there were no obvious signs of infection; but some needles were perhaps very slightly paler than others; an examination of a few of these revealed the presence of mycelium. By midsummer of 1926 needles on some of the experiments were more certainly paler and samples of these were found to contain mycelium. During the winter or spring of 1926-7 some of the experimental plants disappeared or were broken. In four of those that remained infection could not be demonstrated in 1927 or subsequently. But in three of them infection was not only evident in 1927, but their inoculated branches bore the first crops of spermogonia and aecia. In every instance the sori were restricted to the needles of 1924, that is, to the needles of the primary infection. Spermogonia appeared first in order and they were followed about a month later by peridermia. From four to eight twigs of the season of 1924 (along with needles of earlier years) had been enclosed in each inoculation tube, and in the case of each of the three successful experiments several of these showed primary infection, but only on needles of 1924.
Final examinations were made in 1928 and in all of the infected twigs referred to above the rust had spread to the needles of 1925, and in two of them to the needles of 1923. Likewise a stimulated growth
1934]
(The aeciospores used were
FAULL, BIOLOGY OF MILESIAN RUSTS
TABLE 6.
from Abies
Milesia polypodophila—
balsamea to various ferns
from field collections on Abies balsamea)
; Results irst Date of observa Total Fern inocu- tion of Date tee Total vin inoculated lation uredinia harvested Stages Location tions pinnae Li ‘Herb. No. ssc lati 2.vili.24 13.1x.24 13.ix.24 II Garden 24 — — _ 8264 virgin m ; as 2.viil.24 — 22.vi.25 II Mi - —- — — a “ 2.vili.24 13.1x.24 — II ue —- —- — — =: . 2.vill.24 13.ix.24 13.ix.24 II fe 9° —_—- — —_ (lesions only ) = ie 2.vill.24 13.ix.24 13.1x.24 II 400 — — — be as 2.vill.24 13.ix.24 13.ix.24 II 3 27 _- —_— _— ss 2.vili.24 13.ix.24 13.ix.24 II : 68 —- — os oe 5.viii.24 June 1925 22.vi.25 II “ —- — — 8266 . os 5.viii.24 June 1925 22.vi.25 II, Ill — — — 8262(a) - Ss S.vili.24 June 1925 22.vi.25 II, Ul s — — 8262(b) ai es 9.viii.24 June 1925 5.vi.25 II, III Field — 13 5 7374(a) ss ce O.viii.24 June 1925 — 5.vi.2 II 17 3. 7374(b) s os 9.viii.24 —_— 5.vi.25 II Ue —- LS 6 7374(c) : 9.vili.24 ~= .Vi.25 II bi — 21 9 7374(d) a. as O.viii.24 June 1925 5.vi.25 II, Ill : — 26 13 7374(e) ue . O.vili.24 15.ix.24 21.vi.25 II, III : 41 25 17 7375(a) (lesions only ) i a 9.vili.24 —_— ZACVIE2 Si OEE 3 26 23 13 7375(b) ie Ee O.vili.24 15.ix.24 15.ix.2 II ss 235 —- — Dryopteris 4.vili.24 = 5.vi.25 No infection ma 5 4.viii.24 5.vi.25 a ae oe a 8.vili.24 S.viie5 a - hice thal S.vilil.24 = 5.vi.25 No infection 8260 a spinulos raipeeige es S.vili.24 = 5.vi.25 $¢ ~ 8270 a
The “garden” (pot) cultures showed about 200 primary infections; these bore hundreds of edi
uredinia.
The age cultures ee about an equal number of primary infections; these bore hun-
of ure
nia
The pie under oot pa 7375a showed lesions on 17 out of 25 pinnae; these lesions were rich in ri ers and they carried about 155 uredinia, the number per pinna ranging
from 1t
Out of 21 cine 3 only showed no infection.
remained free from infection.
Many controls were carried and all
62 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
was obvious, so that one could say with certainty that the loose broom, a characteristic phenomenon in trees affected by M. polypodophila, was well started. One of these brooms (Herb. no. 8596), representing the entire result of an experiment, was harvested as a unit. It possesses four primary infection areas. Spermogonia and peridermia occur on the needles of 1923 and 1925 as well as on those of 1924 that still remained (several hundreds in all), and infection has plainly spread to the proximal needles of 1926 in many places.
The percentage of infected needles in a given infected branch area is high—commonly from 90 to 95 per cent. The number of peridermia on individual infected needles of the experiments ranged from one to fourteen, and averaged about seven. Some needles on a successfully inoculated twig remain free from primary infection, but most of them, though not necessarily all, eventually are invaded by mycelium that is perennial in the stem. It should be stated that the spread of the fungus from season to season is by means of mycelium that persists in the stem—mycelium that doubtless originated in primary needle lesions. Needles of all ages are susceptible to infection, but needles of the cur- rent season only are susceptible to primary infection. The subsequent course of the mycelium is a growing downward into the contiguous stem, and thence through the cortical tissues of the stem upwards and down- wards, the mycelium as it advances sending out hyphal branches into uninfected needles of any age through their bases. Usually affected needles fall off during the winter following the summer in which they bear peridermia. It happens at times, however, that some persist, espe- cially those that produce spermogonia only in that year, and in the following year they produce what is potentially or actually a second crop of peridermia.
The field inoculations were made with aeciospores from a collection made on Timagami Island in Lake Timagami, a part of which collec- tion is incorporated in my herbarium under number 7267. The loca- tion chosen was Island 1024, a small island distant from known infections and itself free from infected firs. The island is rocky and in the part chosen for the experiments was largely covered by black spruce (Picea mariana), white birch (Betula papyrifera), a few scattered white pines (Pinus Strobus) and a single sapling of balsam fir (Abies bal- samea). There was a limited number of Polypodium virginianum on the floor of the forest—not more than about 100 fronds all told. Each of these was first carefully examined for signs of rust; no traces of rust were found, and those that were not inoculated remained perfectly free from rust at the end of the experimentation. Experiments were set up
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 63
on several each of the fronds of eight plants; one of these plants could later not be located; the selected fronds of the remaining seven became infected as a result of the inoculations made,
TaBLE 7. Milesia fructuosa—
from Dryopteris intermedia americana to various species of Abies
First ap- Date of pearance of Date Fir inoculated inoculation peridermia harvested Remarks A. amabilis 15.vi.33 None 47 days Very scanty. Spermogonia only. A. balsamea 5.vi.33 42 days 41 Scanty. oe a 12.vi.33 Non-erumpent 50 ‘“ canty. ae 12.vi.33 None 50. = Scanty. Spermogonia only. A. cephalonica 24.v.33 35 days (Non- 55 “ Scanty. erumpent ) A. concolor 24.v.33 35 days 48 “ Occasional, : a 2AeN253) 390 |: 5 a Occasional. . = 1.vi.33 37 “ S73 ) : “ s Divison 7428 os qe Occasional. i a 125y1-335 41035 44 “ Frequent. A. Fraseri 1.v1.33° 40. “‘ Aq ss Frequent. Distorting. > os LEVIS oS Os SO Frequent. Insect eaten. a ss 1.vi.33 36 “ 40 ‘ Frequent. Distorting. Es a UZ Soon 40 “ Frequent. Distorting. us i W230) 30 44 “« Abundant. Distorting. A. Fraseri prostrata O°V1,53" 245-5" 46 “ Frequent. Distorting.
A a 6.vi.33 42 “ 42 “ Frequent. Distorting. Piguse sf 13:vi33 35.5 35 9: Abundant. Distorting. eine = 131.33. 35. * 39°" Frequent. Distorting. 14.vi.33 41 “ 42. 2 Frequent. Distorting. ah ee 14.vi.33 41 “ yy equent. A. magnifica 14ivio3. 230 SSoa, Very abdt. Very distorting. a, ce 14 vss) -GOue 38 Very abdt. Very distorting. Hu: i 14.vi.33 36 “ 38“ Very abdt. Very distorting. oe 14.vi.33 34 “ 34“ Abdt. Very distorting. A. nephrolepis 13.vi.33 42 “ (Non- 49 “ Scarce.
erumpent)
se sy 13.vi.33 38 days 43“ Occasional.
The scope of the investigation, the positive results of which are recorded, was much more extensive than would appear from Table 7. Two hundred and eighty-four inoculations in all were made on twenty- eight species and eight varieties of Abies. The species and varieties on which no infection occurred are not enumerated because the experi-
64 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
mental conditions were such that negative results were not interpreted as proof of immunity. Temperatures and moisture conditions were unfavorable for the main part throughout the period of the experiment. From May 24, the date of the first inoculations, up through the end of June the weather was abnormally dry and warm, drying winds were above normal, precipitation meager and relative humidity and cloudi- ness below normal. The rainfall for the last week in May totalled 0.1 inches, for the entire month of June 1.2 inches and for the first half of July 1.4 inches. Moreover, most of the trees inoculated were unavoid- ably in exposed situations with the result that more or less drying of the needles took place and in many instances there was some sun- scorching of the foliage. The most striking infection was obtained on Abies magnifica, and in this instance the trees inoculated were the most protected of all. They are small and are situated in a close copse adja- cent to and at the base on the north side of a steep hill covered by a thick stand of hemlocks. The contrast afforded by the scanty, though positive, infection on Abies balsamea is particularly worthy of note because this species is the usual haploid host of WM. fructuosa, and in its natural habitat is often very heavily infected. The trees inoculated in this experiment are small and unprotected on all sides. Measured by these facts it is reasonable to suppose that some at least of the species that remained free from infection would prove susceptible under con- ditions more favorable to infection.
On the other hand it cannot be assumed that some species of Abies may not be immune to M. fructuosa, nor that susceptible species differ from one another in their resistance to this rust. As to the latter my experimentation affords grounds for concluding that differences in re- sistance do exist among the host species. Such is clearly indicated when comparisons are made as between the effects produced by M. fructuosa on Abies balsamea on the one hand and on A. Frasert, A. Frasert prostrata and A. magnifica on the other. A. balsamea in nature is often very heavily rusted by M. fructuosa, in some instances to the extent that nearly all of the needles of the current season may be rusted, but the infected needles show little evidence of dwarfed growth or dis- tortion. As for the others named and particularly A. magnifica checked growth and distortion are pronounced. It is apparent that for A. mag- nifica attacks of M. fructuosa might easily be so disastrous as to prove of considerable economic importance in connection with the natural or nursery production of that species.
Since provision had not been made in the experiment for a supply of
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 65
plants of the suitable fern host reciprocal inoculations with aeciospores were not made. Admittedly desirable yet under the circumstances such were not necessary to the establishment of the identity of the haploid hosts of M. fructuosa because spontaneous infections from this or sim- ilar rusts not only did not occur but were practically impossible. It may, therefore, be safely recorded that the following species of Abies can serve as hosts for the haploid phase of M. fructuosa : A. amabilis, A. balsamea, A. cephalonica, A. concolor, A. Fraseri, A. Fraseri prostrata, A, magnifica and A. nephrolepis. With respect to geographical dis- tribution of the hosts in their native habitats this list is of added inter- est because it comprises species of Abies from eastern North America, western North America, Europe and Asia.
The reactions of the rust and the various species of Abies infected with M. fructuosa exhibited some differences. The spermogonia varied somewhat in size and shape; nevertheless they approximated quite closely the spermogonia of the M. intermedia form of this species as described (under the erroneous name Milesina Kriegeriana) by Hunter (3), and in all instances were clearly distinct from the larger spermo- gonia of M. marginalis. They varied in their distribution on the needles. On all but Abies magnifica they were mostly hypophyllous, but in some instances there were occasional epiphyllous spermogonia as well. On A. magnifica they were abundantly amphigenous, almost equally distributed on all surfaces. The affected leaves of A. Frasert and A. Fraseri prostrata were often more or less deformed; those of A. magnifica were much deformed and smaller than normal; those of the other species showed no evident abnormality in size or form. The various morphological phenomena in connection with these infections are being studied in detail by Miss L. M. Hunter.
DEVELOPMENTAL PERIODS
The period of development of the haploid phase on Abies, that is, the length of time that elapses between inoculation with basidiospores and the first production of lesions, spermatia and aeciospores respec- tively, has been partially recorded for eight (seven if Milesia fructuosa and M. intermedia are the same species) and fully recorded for three of the species of Milesia whose complete life histories are known. The data are collated in Table 8.
In the case of each of the species mentioned above the new needles only of Abies are liable to infection following inoculation with basidio- spores. This cannot be taken to indicate that the older needles are less susceptible; it is more probable that their constant resistance to invasion
66 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. Xv TABLE 8
Developmental period of haploid phase of species of Muilesia
Period up to Period up to Period up to first appear- first appear- first appear- Species, of Species of ance of le- ance of sper- ance of peri- Milesia Abies sions mogonia dermia M. Blechni {. alba | — 36 days. (?) 55 days A. cephalonica § M. exigua 1. cue | A, fir \ — 13-20 days, 21-28 days, A. inte nsis | av. (2?) av. (?) M. jezoensis A. Mayriana — 20 days — M. Dryopteridis A. Mavyriana -—- 38 days M. intermedia 4. balsamea 10-19, av. 14 14-21, av. 32-38, av. days 18 days 34 days M. marginalis A. balsamea 12-21, av. 14 18-23, av. 39-51, av. ays 19 days 43 days M. polypodophila A. balsamea 1-2 years 3 years 3 yrs. 1 mo, M. fructuosa 1. — ) A. bal ed | A. se etn | 1. concolo | 1. Fra — — 35-45, av. 1. Frasert | 38 days Aint pe | A, magn | A. ne Abi eis |
is due to the nature of the barrier afforded by their maturer cuticles and outer epidermal walls. Evidence in support of these conclusions is furnished from plants occupied by the mycelia of MM. polvpodophila. Although primary infection with this species occurs in young needles of the current season only and never in older needles, yet once infection is established hyphae regularly invade healthy needles of the older ages by passing internally from the stem cortex through the tissues of the petioles out into the laminae. Indeed, it would seem that the proto- plasmic resistance of such needles is even less than that of needles of the current season, for while the struggle between the latter and the parasite is continued for two years beyond the year of infection before reproductive organs are formed, it is shorter by one year and possibly in some instances by two years in the case of needles secondarily in- fected. Protoplasmic resistance apparently wanes with age.
The period of development of the diploid phase on ferns, that is, the length of time that elapses between inoculation with aeciospores or uredospores and the first production of lesions, uredospores and telio- spores respectively, cannot as yet be expressed with the same exactness
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 67
as for the haploid phase. But observations have been made on a con- siderable number of species and experimentation conducted with a few. As for the latter some statistical information is available for five species, namely, M. Blechni, M. exigua, M. marginalis, M. intermedia and M. poly podo phila.
Klebahn (7) inoculated Blechnum Spicant with aeciospores of M. Blechni experimentally produced and found that the developmental period following infection was protracted; inoculations made July 20 showed lesions along in August; no statement was made as to when uredinia appeared. Inoculations made with uredospores in June were followed in about three weeks by uredinia. Kamei (4) made two sow- ings of aeciospores of M. exigua experimentally produced on Poly- stichum Braunii and obtained uredinia in 13 and 17 days respectively. Data on the remaining three species are derived from my own experiments.
Dryopteris marginalis inoculated with experimentally produced aecio- spores of M. marginalis, in pot cultures, on various dates from August 20 to August 22, 1924 and August 22 to August 27, 1927 showed no external evidences of infection up to September 15, the last date of observation for both years. It is regrettable that observations could not have been continued for a month or six weeks longer, because in spite of many frosty nights in the locality of experimentation from the middle of September onward some fairly temperate weather is experi- enced until towards the end of October, perhaps warm enough at times to permit further growth of the fungus. It should likewise be noted that lesions on D. marginalis caused by M. marginalis are very slow in discoloring. It is a common experience to find tendrils of white spores standing out from the inconspicuous uredinia on portions of the frond not perceptibly less green than the unaffected parts. Hence it is not improbable that uredinia were formed before the onset of winter—a fact determinable only by very close examination. However this may have been, the first uredinia we observed in our experimental plants were in May of the following springs. Teliospores developed about one month thereafter. No inoculations were made with uredospores, but I have concluded that the developmental period is much prolonged following natural inoculation on new fronds with uredospores in the spring from uredinia on overwintered fronds. Throughout the summer there is no sign of the rust and not until fall are lesions and uredinia evident. I have never found uredinia on fronds of the current season earlier than September, and even then rather sparsely. The abundant crop is on
68 JOURNAL OF THE ARNOLD ARBORETUM TvoL. xv
overwintered fronds the following spring. The uredospores then begin to emerge in snowy tendrils soon after the fronds are exposed, even before the last traces of snow disappear. The teliospores do not de- velop until about a month later, that is, until about the time the buds of Abies are swelling and opening.
Inoculation experiments with experimentally produced aeciospores of M. intermedia from Abies balsamea were conducted on D. spinulosa and D. spinulosa intermedia—all in pot cultures. Inoculations were made August 12 to 14, 1924 and August 16, 1927. Neither lesions nor uredinia were observed up to September 15, the last date of observa- tion for both years. Here again, just as with M. marginalis, it is pos- sible that there was further development before winter set in. If so lesions probably became evident, for on both hosts mentioned lesions in nature are in evidence during the fall. It was likewise not known at that time that the teliospores of M. intermedia may sometimes develop in the fall on fronds of the current season. All that can be said further is that uredinia did not form at any time on our experimental plants and that teliospores were not located until June of the following springs. All these phenomena were typical of natural occurrences for the same periods.
Coming to M. polypodophila a difference in the story is presented. Inoculations were made with aeciospores from field collections on both potted plants of Polypodium virginianum in our garden and on undis- turbed plants in the forest. These were made at various dates between August 2 and August 9, 1924. In almost every instance lesions were evident by the middle of September, and many of these bore uredinia by September 13. The elapsed period culminating in production of uredospores ranged from 37 to 42 days. In no instance, however, did teliospores develop on the infected fronds until June of the following spring. These phenomena coincide with observations on natural infec- tions whether in infections following aeciospores or uredospores. Ure- dinia in such cases may occur as early as August, but I have never found teliospores on fronds of the current season. They appear first on overwintered fronds at about the time the buds of Abies are swell- ing and breaking. The data recorded above are summarized and em- bodied in Table 9.
Any information on the developmental period of the diploid phase of species of Milesia on ferns derived from labels that accompany herbarium specimens can at best be little more than suggestive. Such records do indicate, however, that there are considerable differences
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 69 TABLE 9
Developmental period of diploid phase of species of Milesia
Species of Milesia Host Inoculum used and results M. Blechni Blechnum Spicant Aeciospores—protracted apie areas Ure- ores — ure in 3 eeks. M. exigua Polystichum Braunii Aeciospores—uredinia in 13 and 17 days. M. marginalis Dryopteris marginalis Aeciospores—no sees fan _ peg at end of da ode ure- dinia aay hav for mea in see fiom but seen ‘st in of follow —teliospores developed in pide f following year. M. intermedia Dryopteris spinulosa re eterna signs of lesions at D. spinulosa intermedia § a of 3 paced no uredinia med a Ae developed in spring following yea M. polypodophila Polypodium virginianum Aeciospores—uredinia in 37 a 42 days.
—teliospores ica a in spring of following ye according to the species. The most striking instance of prompt devel- opment is afforded by M. murariae on Asplenium Ruta-muraria in col- lections from Switzerland; specimens coliected as early as June 12 carry a prolific crop of uredinia on the new fronds. This is in striking con- trast to M. fructuosa on Dryopteris spinulosa intermedia in which the earliest production of uredinia occurs in the fall. As intermediates between these two there is a range from early June to late August as illustrated by the series M. Feurichii, M. Blechni, M. dilatata, M. Sco- lopendrii, M. carpatica, M. vogesiaca, M. fructuosa (on D. spinulosa americana), M. Polypodii, M. Polystichi and M. Kriegeriana. HABITS OF SPORE PRODUCTION
(a) SPERMATIA
Some observations were made on the production of spermatia and their period of discharge in connection with my field studies of M. inter- media, M. marginalis and M. polypodophila. As might be inferred from the relatively large size of their spermogonia and the ample protection afforded them by the host tissues spermatia are produced profusely and their discharge is continued over many days. Examinations of these phenomena were made daily with respect to M. polypodophila. In this species as in the others there is an outpouring of the spermatia in liquid suspension, but in M. polypodophila it is especially copious. When
70 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. Xv
the weather is humid the liquid spreads over the entire lower surfaces of the needles on which they are borne, and at times it fairly drips from the needles. It is somewhat sticky and sweetish, but no odor was detected. Discharge was found to continue for about thirty days.
The following excerpts from my field notes (4843) in the Timagami Forest Reserve, Ontario, under date June 9, 1924, embodying observa- tions on the spermogonia of M. polypodophila and their habit of discharge, present a graphic picture of the spermogonia of M. poly- podophila and their behavior. ‘The affected leaves are now moist on the lower surface, films and droplets—due to the active exudation of spermatia and a colorless fluid. A needle (leaf) pressed against a clean dry glass slide leaves an almost continuous smear of spores and liquid. The spores are hyaline, one-celled, cylindrical to ellipsoidal, and occa- sionally almost rod-shaped. They measure 2.5-3.2 x 3.5-7 u. The majority are about 3 x 5 yu. The needles along stretches of the axis of the branch (1) continuous with those that bore peridermia last year, (2) or that carried peridermia-bearing needles last year (now fallen off) show many inactive spermogonia. These are greenish-yellow in color with a blackish dot in the center. On the same needles are new spermogonia interspersed with the old ones, on restricted portions of the needles (usually towards their tips) on which areas there are few or no old ones. The new spermogonia are actively discharging: they are slightly discolored (bluish with yellowish tinge) and the central point is moist and not discolored. The affected needles along these stretches of the stem axis are distinctly paler (slightly yellowed) than normal leaves. Contiguous with these stretches are lengths of the axis on which the needles carry new spermogonia only. These needles are scarcely paler than normal needles on their lower surfaces and not paler on their upper. The spermogonia are commonly scattered along the full length of the needles, but are sometimes restricted to limited portions from the bases outward and not extending all the way to the tips. This is especially true of the needles farthest removed from the affected areas of last year. Nearly all of the needles along a given affected stem axis are infected. The length of an axis affected in a given season bears no relation to a season’s growth.
‘The number of spermogonia on a needle varies from ten to twenty and they are spaced in lineal rows, one on each side of the midrib. Frequently there may be a stoma more or less eccentrically located on the epidermis that covers a spermogonium. This opens wide, the opening with a jagged line of broken cuticle, and through the opening the spermatia are being discharged.”
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 71
(b) AECIOSPORES
The peridermia of these species with their first apparently mature aeciospores appear in from two weeks to a month after the spermogonia begin their spermatial discharge—an average of 16 days for M. inter- media, 24 days for M. marginalis and about 30 days for M. polypodo- phila. he peridermia rupture at their apices and discharge of the aeciospores begins. Rupture is delayed for a few days, however, the length of the interval apparently depending to some extent on weather conditions, being shorter when the weather is rainy or when there are heavy dews. The intervals were found to vary from less than a full day to six days for M. intermedia, with an average of three days, and from one to eight days for M. marginalis, with an average of four days. The discharge of aeciospores is protracted over a comparatively long period of time. For example it has been frequently noted that peri- dermia of M. intermedia formed soon after the middle of August were shedding spores until towards the end of September.
(c) UREDOSPORES
The habits of uredospore production characteristic of Milesian rusts exhibit considerable diversity, as was discovered from comparative observations on various species. An interpretation of the factors under- lying the differences manifested is complex, but much of interest could probably be elucidated by experiment. The factors involved are to be sought for in the innate characters of the individual species, the degree of compatibility as between host and parasite, the influence of the host and the influences of moisture and temperature.
Uredospore production from the same frond is usually active over a long period of time, interrupted only during periods of low temperature or drought. To what extent this applies to individual uredinia is un- certain. In some species, as in M. Blechni and several tropical forms, it seems to be relatively short as indicated by the circumstance that often many uredinia at the time of gathering are completely empty— sometimes to such an extent that few or no spores can be found in a collection. In others spores can be found in practically all uredinia regardless of age, as in most of the temperate zone species. The length of the time during which uredospores are produced on a frond naturally depends on the promptness with which uredinia are formed on new fronds and the longevity of these fronds after being overwintered. Striking examples of long sporulating periods are afforded by such spe- cies as M. murariae and M. polypodophila. In the former the new
72 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
fronds become infected in spring and sporulation is at once abundant; the same fronds are still producing spores copiously a year later side by side with heavily infested new fronds. Sporulation does not begin as early in M. polypodophila, but it is active from midsummer onward and continues throughout all of the subsequent year even late into the fall. A remarkable example of suspension of spore production due to drought is found in M. laeviuscula on Polypodium californicum. Infection must take place in the spring; but during the drought period, which in parts of California where the rust occurs may extend over several months, the fronds and the rust they carry are inactive. With the coming of the winter rains fronds and rust are revived and sporulation begins, to be continued well on into the following spring. An example of a Milesia in which the sporulating period is comparatively short is afforded by M. marginalis. Uredinia of this species according to my observations do not form until fall on fronds of the current season, and spore pro- duction that season is quite limited. The heavy spore discharge takes place the spring following, but it is rarely continued beyond June, and in dry springs the old fronds may die even before there has been normal time for teliospore formation.
Of all species of Milesia I have found none quite so perplexing with respect to uredospore production as M. intermedia. Widely distributed throughout the range of Abies balsamea, a common species the fre- quency of which is made possible by the almost constant presence every- where of its fern hosts in coniferous or mixed forests, it seems to have wholly escaped the notice of collectors because of its common habit of omitting uredinia. My finding of M. intermedia resulted from a delib- erate search for teliospores on Dryopteris spinulosa intermedia after having discovered those of M. marginalis on D. marginalis. Once hav- ing found the teliospores the abundance of this rust soon became evi- dent. Yet several years elapsed, though constantly watching for them, before its uredospores were encountered. The uredinia as I first found them were scattered, very small, immersed, very inconspicuous and their spore output was scanty. Collections made previously were again painstakingly examined for uredinia but with negative results. Thus far my collecting had been done in the eastern Provinces of Canada and the two most northeasterly States of the United States. In 1932, however, in-the course of collecting trips in the more southerly range of Abies balsamea in western Massachusetts and Vermont, I found a rust on varieties of D. spinulosa, easily manifest because of the abun- dance of its closely-studded, conspicuous, though small, pustular ure-
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 73
dinia and abundant discharge of white uredospores. The latter so closely resembled those which I had hitherto found so scantily on D. spinulosa intermedia as to be practically indistinguishable. Because of the differences in uredinia and in the habits of uredospore production, the rust with abundant uredinia was described under the name M. fructuosa and the other under the name M. intermedia. Regardless of whether or not both names will stand, a useful purpose has been served in calling attention to a rust that otherwise might be passed over be- cause of a condition so often characterized by the entire omission of uredinia.
For the purpose of obtaining fuller knowledge of M. fructuosa and M., intermedia I made a special collecting trip to Timagami in northern Ontaria, the Green Mountains of Vermont, the Berkshires of Massa- chusetts and the Adirondacks of New York in the neighborhood of Lake Placid during the middle and latter part of October, 1932. The visits were made at a time when presumably there would be no further growth of the rusts that year. Timagami and the higher elevations of White- face Mountain in the Adirondacks were already partly under snow, and all of the localities visited had already experienced killing frosts. The results warranted the efforts made by the extension of our knowledge of distribution, host records, the existence of intergradations of the two “species” and a fuller acquaintance with the habits of teliospore pro- duction in M. intermedia. To give as complete a picture as possible as a basis for conclusions as to the identity of these species and as a help towards needed further life history experimentation a list of all of my materials of M. fructuosa and M. intermedia with herbarium numbers and pertinent data is recorded in Table 10.
With these data before us the question arises as to why Milesian uredinia on Dryopteris spinulosa and D. spinulosa intermedia were not found in Timagami during the years 1924-7, years during which Mile- sian rusts were being assiduously studied there. I think the correct reply is that if present at all they were rare. The collecting grounds tn Timagami were well-defined and were visited often each year. One of these, a small area covered by low-growing Abies and interspersed clumps of Dryopteris spinulosa and D. spinulosa intermedia through the midst of which a trail led was traversed by myself and assistants many times each season. This particular area was one of the best localities for M. intermedia both on Abies and the ferns, and I repeat- edly examined the fronds in search of uredinia; not once were any found. On revisiting the same spot in October, 1932 I found them
74 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV TABLE 10
I. Milesia fructuosa (continued)
A. On Dryopteris spinulosa (O. F. ori Kuntze
No. Place Stages Age of fronds 10,853 Timagami, Ont. Oct. Be 1932 Il, 111 Current season 10,852a ™ “ Oct. 11, 1932 II, Ill . 10,847 ‘ é‘ ct. 14,1932 ‘II, III «“ a O,85 ne 7 Oct. 14, 1932 II, Il “ se 10,859 7 = Oct. 14, 1932 TE, Ti “ a B. On Dryopteris spinulosa americana age ) Fernald
No. Place Stages Age 10,668 Mt. Greylock, Mass. Heed a 1932 I, Il ieee (dead) 10, oe 7, 1932 II, 111 Current seaso n 10,76 “ = 7 te 7, 1932 II, I “
10, ah “ _ . Aug. 7, 1932 II, Il - a
10,883 “ ‘ “ Oct. 22, 1932 II, I “ 10,887* “ a Oct. 22,1932 UL, Il o « 10,771 Sherburne erage ae Aug. 10, 1932 I, Wl - af 10,768* Whiteface Mt., Aug. 9, 1932 II (rare) Overwintered
III (empty) at this date) 10,876* Whiteface Mt., N. Y. Oct. 24, 1932 II, 11] Current season sii margin of nge of a 10,877* Whiteface Mt., N. Y. Oct. 24, 1932 II (rare) 7 (a rotected, aes III (abdt.)
exposure, higher up)
ae On — spinulosa fructuosa oe Trudell
Stages Age of fronds 10, ey Mt. Greylock, Mass. June oe 1932 II, HI Overwintered 10,627 June 17, 1932 , Hl 7 10,628 a ‘ June 17, 1932 lI, HI " 10,764 7 Aug. 6, 1932 I Current season 10.885 7 = Oct. 22, 1932 II, Ill . im 10, Ss 6 a9 a9 ae Oct. 22, 1932 i is9 a 10,629 Searsburg, Vermont June 17, 1932 a Overwintered 10,665 Mt. Mansfield, Vt. July 15, 1932 10,669 “ oe se a 15, 1932 te iT as
D. On Drvopteris spinulosa ee —— — Age of fronds
No. Place 10, 847a Timagami, Ont. Oct. ey 1932 WW Current season 10,848 SL Oct. 14, 1932 II, III e 7 10,851 . = Oct. 12, 1932 Il, HI ‘ “ 10,852 - - Oct. 11, 1932 II, Il * “s 10 oo ¢ " Oct. 14, 1932 II, Ill = bd 10,85 ‘ e Oct. 14, 1932 II, Ill a “ 10, 892 Becket, Mass. Oct. 21, 1932 II, III a i 10,888 Sherburne Pass, Vt. Oct. 22, 1932 Il, II “ “ 10,889 Oct. 22, 1932 II, Il as mf 10,893 a ee Oct. 22, 1932 II, UI ae
gsi 10,876 is from the lower margin of range of host. oan _ 877 is from an open sunny osure higher up and within the main ra of the host and aeahue oat by “M. intermedia.” In Numbers 10, 768 a “10, 877 the II stage is rare.
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 15
Il. Milesia intermedia AL ha eget spinulosa (O. F. paca Kuntze Stages Age of fronds 8.272 Timagami, Ont. ae . 1925 Ill Overwintered (Cul-
media were used.) 9,850c a : June 30, 1928 III Overwintered (Cul-
tures, ditto.) 10,915 Gaspé Co., Que. Sept. 8, 1928 III Current season
BL On ee spinulosa americana (Fisch.) Fernald Date Stage Age of fronds
10, ari Whiteface Mt., N. ne Oct. 24, 1932 III Current season 10,875* Oct. 24, 1932 III a 3 10,878* ee i. ie Oct. 24, 1932 III ne si * on weed lt spinulosa scsaoea aia ) Underw. Stages Age of fronds 487 Tiatdan Ont. June oi 1924 III Overwintered ihe: June 16, 1924 Ill “ nel June 18, 1924 Ill A 8,267 ie be June 5, 1925 Ill ie ore dit under "3272) 8,260 ss is June 5, 1925 Ill 8,270 ‘ i. June 5, 1925 Ill oy s 9,850d - : June 27, 1928 Ill se = 9,924 Mt. eas eae N.H. July 5, 1931 II*, III Overwintered 10,670 Beck: July 12, 1932 DE gra dee a 10,6 Mt aed Vite July 15, 1932 II*, III : 10,800 Sherburne Pass, Vt. Aug. 11, 1932 Ill “(fronds ead) 10,880 Lake Placid, N. Y. Oct. 23, 1932 *e Current season 10.854 Wilmington Notch, N. Y. Oct. 23, 1932 se i: sh
*Prevalent throughout the main range of host and very abundan **Mostly just sporeless lesions; a few spores of stages II and ttt in a few of the oldest lesions found after careful searching, in such cases making decision difficult as a whether the rust should be called M. fructuosa or M. intermedia. *** Affected fronds of current season found in Maine, Quebec and Nova ‘Scotia bore spore ie lesions only.
abundantly and on practically every affected frond. A few hundred yards farther on, in a narrow swampy valley just on the other side of an intervening hill is the place where the type material of M. inter- media was collected. Many hours were spent there springs and falls of 1924-7 in intensive, undivided search of uredinia; none were ever found. Yet in October 1932, though snow had to be kicked away to locate the ferns, several fronds bearing uredinia were uncovered. A
76 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
subsequent re-examination of my earlier collections (all on overwin- tered fronds) made from that particular area confirmed my recorded observations that though teliospores were abundant there were no ure- dinia. Half a mile eastward is still another locality where, except for a single frond bearing uredinia of the M. fructuosa type found in 1926 or 1927 (subsequently lost), nothing but non-uredinial M. intermedia had been seen. There, in October 1932 almost every affected frond both of D. spinulosa and D. spinulosa intermedia bore quantities of the uredinia of M. fructuosa.
With the contrasting situation in Timagami, in 1924-7 as compared with the situation in 1932, it is illuminating to refer to the contrasting situations that existed on Mt. Greylock, Mass. and Whiteface Mt., N. Y. in the season of 1932. In the former locality, that is, on Mt. Greylock, spring came earlier, rust developed on Abies earlier and fall was later. A rust of the M. fructuosa type abounded there from early August onward on fronds of varieties of D. spinulosa of the current season, while the M. intermedia type was absent. On Whiteface Mt. rust abounded on D. spinulosa americana and except in the lower reaches and in sunny exposures farther up the mountain uredinia did not occur. Even where uredinia did occur they formed very late in the season and quite scantily. The trail up Whiteface Mt. from Wilming- ton (northern side of mountain) traverses a mile or more of terrain covered with D. spinulosa americana; the fronds of an enormous num- ber of them in the fall of 1932 were laden with teliospores and without uredinia except as just indicated. In other words the condition of the rust was typical of M. intermedia.
A similar parallel existed in 1932 as between Wilmington Notch, N. Y. and Sherburne Pass, Vt. with reference to the rust on D. spinulosa intermedia. In Wilmington Notch, for the main part sporeless lesions were present; the fronds entered the winter in the same condition I have so often found to prevail in northern Ontario, Quebec, Nova Scotia and the White Mountains of New Hampshire. At Sherburne Pass, on the other hand, there was an abundance of both pustular uredinia and telia, though they developed later in the season than on D. spinulosa americana from the same locality.
One of two explanations of these phenomena is probably correct. Kither both M. fructuosa and M. intermedia occur on D. spinulosa, D. spinulosa americana and D. spinulosa intermedia, varying in relative frequency periodically and locally, or M. fructuosa and M. intermedia are one and the same species, the production of uredinia potentially
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 17
being always possible but actually existent under certain conditions only. From the assembled data even though direct demonstration is lacking, I incline to the latter view.
It might be urged that a compelling argument in favor of regarding M. intermedia as a distinct species is the record of no uredinia in the six successful cultures recorded above (Table 2). But it is to be borne in mind that since the experiments were conducted out-of-doors the results would have been expected to coincide (as they did) with the occurrences on naturally infected plants in the same territory. Frankly, I suspect that M. intermedia is a condition of M. fructuosa, a condition characterized by an omission or reduction of uredinia according to environmental influences. This problem and that of biological strains and host preferences afford interesting topics yet to be conclusively investigated.
(d) TELIOSPORES
The first discovery of teliospores of a Milesia was made by Magnus (8) in specimens of Dryopteris spinulosa affected with M. Kriegeriana. They were found on fronds of the current season and it has been fully established that this is the typical habit for that species. Extended acquaintance with other species, however, has revealed the fact that for most of them the teliospores, if known, develop on affected over- wintered fronds only. My discovery of this phenomenon resulted from persistent search for teliospores in an ample collection made in early spring of fronds of Dryopteris marginalis rich in uredinia-bearing le- sions. A prompt examination for teliospores proved fruitless, but that outcome was not regarded as final because under the belief that these spores are rare I thought they might possibly have been overlooked. Hence as the collection was a large one and only a part of it had been overhauled the fronds were placed in a moist chamber to be kept fresh until the search could be completed. About three weeks elapsed before opportunity was afforded to continue the examination; at once, to my astonishment, I found the lesions throughout crowded with teliospores. Immediate reference to plants in the forest revealed that in them also teliosporic production had taken place. The missing clue to the time of occurrence of the III-stage which subsequent observations have shown to be so prevalent in Milesia was uncovered. As the periods of telio- spore occurrence for the various species have already “been published (2) there is no necessity for listing them here.
A point of interest relevant to those species in which teliospores are
78 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. Xv
formed on fronds of the current season is the question as to when the teliospores are capable of germination. According to Magnus’ account of M. Kriegeriana the inference is that the teliospores of that species are immediately germinable. Magnus did not state, however, whether or not germination occurs in the fall under natural conditions, nor did he comment on its relative frequency, Obviously if germina- tion does take place abundantly in the fall such a habit would be waste- ful because the needles of Abies are not susceptible to infection at that season. Dietel (1) in a later communication recounted that he found the teliospores of M. Kriegeriana still ungerminated in the early spring on overwintered fronds bearing them, and I have observed the same phenomenon in suitable specimens of that species which have come to my hand. The nearest parallel to Magnus’ observation was my find- ing in early spring before the new needles of Abies unfolded quantities of teliospores with empty cells and open germ pores through which basidia had evidently grown in the case of M. fructuosa on D. spinulosa americana and D. spinulosa fructuosa. Under the circumstances it was impossible to know whether germination had taken place the preceding fall or on exceptionally warm days after the snow cover had melted off in the first blush of spring. That the latter is the more likely interpre- tation is supported by my recent attempts to germinate the teliospores of this species. Materials were collected in October 1932 from the same hosts and in the same locality, teliospores that must have formed sometime between the middle of August and early October. They showed no signs of germination at the time of collection and all attempts to induce germination in the laboratory have ended in failure up to the time of writing (end of December 1932). It is, therefore, reasonable to conclude that as for M. fructuosa its teliospores overwinter in a dormant condition and that germination normally first takes place the following spring at about the time the new needles of Abies have just expanded.
In conclusion some comments should be offered with respect to the habit of teliospore formation in the species “M. intermedia.’ From the time of my first finding the species I have observed again and again that sporeless lesions in the fall on fronds of the current season was the prevailing phenomenon, while teliospore-bearing lesions was the pre- vailing condition on the overwintered fronds the following spring in the same localities. These experiences led me to suppose that the telio- spores of this species were formed on overwintered fronds only, but recent observations have revealed the fact that they can also be formed
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 79
in the fall on fronds of the current season. On examining a collection of supposedly rusted fronds of D. spinulosa made in September 1928 in Gaspé, Que., a collection that had been set aside because the lesions bore no uredinia and presumably no teliospores at that season (Faull, no. 10,915), I found that while some of the lesions were sporeless others already contained teliospores of M. intermedia. Similar findings came to light in specimens of rusted D. spinulosa intermedia made in October 1932 at Lake Placid, N. Y. and Wilmington Notch, N. Y. My conclu- sion is that the teliospores of 7. intermedia do form on the fronds of the current season as well as on affected overwintered fronds depending on the earliness of infection and the nature of the environmental conditions. HOST RESTRICTIONS
All that is known of host restrictions of various species of Milesta has already been referred to or indicated elsewhere in this paper so that a brief recapitulation will suffice. The sole data prior to those pre- sented in this paper on the restrictions of the O-I phase have been con- tributed by Klebahn and Kamei. Klebahn (7) found that M. Blechni would not infect Picea Abies, but that the two species of Adzes he tested served equally as hosts. Kamei (4) found that three species of Abies tested were susceptible to M. exigua. This paper records a wide range of species of Abies susceptible to MW. fructuosa (Table 7
With reference to fern hosts Klebahn (7) tested /. Blechni on Blechnum Spicant, Dryopteris spinulosa and Scolopendrium vulgare and found that the first-named alone gave positive results. In my own experimentation the three species M. intermedia, M. marginalis and M. polypodophila were tried out on several fern species the inoculations being made with aeciospores in each instance. Aeciospores of M. inter- media were sown on Dryopteris spinulosa, D. spinulosa intermedia, D. marginalis, D. cristata, D. fragrans and Polystichum acrostichotdes. Infection occurred on the first two, but there was no infection on any of the others (Table 2). Aeciospores of /. marginalis were sown on Dryopteris marginalis, D. cristata, D. fragrans, D. spinulosa, D. spinu- losa intermedia, Polypodium virginianum and Polystichum acrosti- choides. (Infection resulted on D. marginalis only (Table 4). Aecio- spores of M. polypodophila were sown on Polypodium virginianum, Dryopteris marginalis and D. spinulosa intermedia but infection occurred on P. virginianum only (Table 6). It may, therefore, be con- cluded that the diploid phase of these species is closely host restricted. The data assembled from collections of other species indicate like close
80 JOURNAL OF THE ARNOLD ARBORETUM [VOL, XV
host restrictions. It is true there are a few species like M. vogesiaca and M. exigua in which several species of ferns are found to be suscep- tible, but always they are of the same genus. Even here it remains to be determined to what extent biological strains more closely restricted still may exist. The one case in which a Milesia may infect ferns of more than one genus is that of M. nervisequa, a species the exact generic status of which will be open to question until we know for cer- tain whether or not its spores are white in the fresh condition; but to whatever genus it properly belongs its biological entities will remain unknown until comparative inoculations have been made. ECONOMIC CONSIDERATIONS
So far as I am aware there are no statistical records extant of the amount of losses to firs or ferns caused by species of Milesia. As for ferns the economic damage, in America at least, is probably not of much consequence. Whatever injuries there might be would most con- cern the owners of properties in summer resort regions where native ferns are so universally grown as outdoor ornamentals. According to my observations the damages from Milesian rusts do not compare with the often distressing injuries caused by species of Uredinopsis.
As for Abies it is conceivable, judging from what I have seen in the forest, and from what ! have seen of the severe effects on the foliage of A, magnifica in culture experiments (see Table 7 above), that appre- ciable injury might result in young plantations where infected ferns are in the immediate neighborhood. Certainly it is true that Milesian rusts do affect natural reproduction to a varying extent. The most important species in this connection covering the range of Abies bal- samea in eastern America, as deduced from my own more or less casual observations on the subject, are M. polypodophila and M. fructuosa (including M. intermedia).
The former induces a broomy growth of firs, involving practically the entire plant if the attack is made when the tree is young, as is usually the case. The resulting growth is unsightly and worthless. Moreover, in such areas not only is the prospective timber value of the firs spoiled but also none of those badly affected ever produce crops of seeds. Of course, such effects are possible only where Polypodium is abundant as part of the under-cover.
Undoubtedly M. fructuosa (including M. intermedia) is of much wider significance in relation to the natural reproduction of the balsam fir, partly because of its wholesale destruction of new needles and be- cause of the general prevalence of its fern hosts. Seedlings on the forest
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 81
floor are either killed outright where infection is heavy or greatly re- tarded in getting a start. Older plants, up to four or five feet in height, sometimes suffer from the loss of an appreciable part of their foliage of the current season. Indeed I have seen areas in eastern America where Dryopteris spinulosa or its varieties abounded, in which rust injury almost certainly played a part in reducing the balsam fir as a constituent of the forest stands or in preventing it from becoming as fully established as it otherwise would. This feature in the competitive struggle for existence applies all the more to firs in the southerly limits of their range, that is in regions where ferns most abound and where the firs are subjected to a greater struggle against unfavoring climatic conditions. Reduced in numbers and checked in growth the firs be- come outgrown and overgrown by other species. It is probably true that such checks to fir reproduction in many of these instances would be considered desirable from the standpoint of forest values. But, however that may be, the fact remains that whenever intensive studies are being conducted for the purpose of quantitatively forecasting forest successions fern rusts as influencing factors should not be disregarded. Plate 86 illustrates the disastrous effect of a_fern rust (Milesia inter- media) on seedlings of balsam fir growing under natural conditions. The forester should bear in mind the principle that firs and ferns of certain kinds are incompatible neighbors where fern rusts are strongly entrenched. ACKNOWLEDGMENTS
For assistance in making the inoculation experiments recorded in this paper I am indebted to Messrs. W. R. Watson, G. D. Darker, E. H. Bensley, I. H. Crowell and Miss L. M. Hunter. I wish particularly to acknowledge the devoted efforts of Mr. Watson throughout two sum- mers of fruitful field work.
SUMMARY
1. Milesian rusts are now known in nature (including a few yet unpublished) to occur on species of about sixteen genera of ferns—all of the farnily Polypodiaceae—distributed among the polypodiaceous subfamilies Acrosticheae, Aspidieae, Asplenieae, Davallieae, Poly- podieae, Pterideae and Woodsieae. They are also known either in nature or cultures on eleven species and one variety of Abies, namely, A, alba, A. amabilis, A. balsamea, A. cephalonica, A. concolor, A. firma, A, Fraseri, A. Fraseri prostrata, A. magnifica, A. Mayriana, A. nephro- lepis and A. sachalinensis.
2. Hosts of both the haploid and diploid stages in the life histories
82 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
are now known for ten species of Milesia (nine if M. fructuosa and M. intermedia are identical), namely, M. Blechni, M. Dryopteridis, M. fructuosa, M. exigua, M. intermedia, M. jezoensis, M. Kriegeriana (Mayor, unpublished), M. marginalis, M. Miyabei and M. polypodo- phila, This paper records the data of life history studies on M. fruc- tuosa, M. intermedia, M. marginalis and M. polypodophila.
3. The developmental period of the haploid phase (the phase on Abies) of Milesian species varies widely and is approximately constant according to the species. The period up to the first appearance of lesions ranges according to the species from 10 days to 2 years, up to the first appearance of spermogonia from 13 days to 3 years and up to the first appearance of peridermia from 21 days to 3 years 1 month (Table 8). The developmental period of the diploid phase (the phase on ferns) does not show such great differences, but it too varies widely according to species (Table 9).
4. Spermatial discharge is relatively abundant and long continued. The most striking instance is that of M. polypodophila for which spe- cies it may continue for a month, the spermatial fluid at times fairly dripping from the needles.
Aeciospores make their appearance in from two weeks to a month following the first discharge of spermatia. The period under normal conditions averages about 16 days for M. intermedia, 24 days for M. marginalis and 30 days for M. polypodophila. The discharge of aecio- spores may continue over several weeks.
Uredospore production from the same frond is usually active over a long period of time, but there are striking exceptions. Perhaps the most remarkable instance of longevity is afforded by M. polypodophila— discharge is known to begin in midsummer and to continue until late in the fall of the subsequent year.
5. Significant variations have been observed in M. intermedia as originally defined with respect to the time of formation of teliospores (Table 10, 11). Ordinarily these form on Dryopteris spinulosa and D. spinulosa intermedia throughout most of the localities with which I am familiar in the spring on overwintered fronds, but I have recently dis- covered that some at times develop in the fall on fronds of the current season.
Equally significant variations have been observed in what I take to be M. fructuosa as originally defined with respect to the relative abun- dance of uredinia (Table 10,1). Ordinarily these occur abundantly on D. spinulosa americana and D. spinulosa fructuosa; but in the upper
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 83
reaches of Whiteface Mountain of the Adirondacks I found on D. spinulosa americana a complete lack of them on the majority of affected fronds.
Owing to the absence of sharply cut morphological differences be- tween M. intermedia and M. fructuosa, the modifications noted above serve to break down clear distinctions between them. My tentative conclusion is that M. intermedia and M. fructuosa as originally defined refer to different manifestations of the same rust species, manifesta- tions that appear to be determined by host or climatic factors. The name M. fructuosa has priority. Its definition then should be amended by stating that uredinia vary in occurrence from none to many, in morphology from few-spored, vesicular, depauperate to many-spored, pustular, fairly conspicuous, and that teliospores form in the fall on fronds of the current season, or in spring on overwintered fronds, or part in the fall and part in the spring.
Pondering on the factors that determine the formation of teliospores in Milesian rusts one thinks of such interesting work as was accom- plished by Waters (9) on the subject of this phenomenon in certain other rusts. Waters recorded that for nine species of rusts studied by him teliospore and uredospore formation could be influenced by experi- mental methods. He stated as part of his summary that teliospores were formed when “the host plants were placed under environmental conditions unfavorable for their development.” It will be interesting to learn to what extent the metabolic state of the host determines telio- spore formation in Milesian rusts, a state that in some hosts presum- ably would be attained in the fall, in others in the spring only after the fronds had overwintered. In the absence of direct observation on this point conclusions would be speculative, particularly when we know of such rusts as Melampsorella Caryophyllacearum and Hyalopsora As pidi- otus in which teliospores form solely in unfolding young leaves in the spring, leaves in which metabolic conditions would appear to be at their best.
6. Teliospores of M. fructuosa formed in the fall could not be made to germinate before the onset of winter. Evidently they require a resting period and their natural time for germination is in the follow- ing spring at the time the needles of Abies are unfolding.
7. Evidence is accumulating that species of Milesia are not closely restricted with respect to the species of Abies they may infect. This paper records (Table 7) successful inoculations of M. fructuosa on A. amabilis, A. balsamea, A. cephalonica, A. concolor, A. Fraseri, A. Fra-
84 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
sert prostrata, A. magnifica and A. nephrolepis—a list of hosts com- prised of representatives from eastern North America, western North America, Europe and Asia.
On the other hand evidence is accumulating that species of Milesia are much more closely restricted with respect to their fern hosts. This paper records the results of experiments pertaining to this question in which three species were tested each on several species of ferns. (a) Milesia intermedia from Abies balsamea was inoculated on to Dryop- teris spinulosa, D. spinulosa intermedia, D. marginalis, D. cristata, D. fragrans and Polystichum acrostichoides ; infection resulted on D. spinu- losa and D. spinulosa intermedia only (Table 2). (b) M. marginalis from Abies balsamea was inoculated on to D. marginalis, D. cristata, D. spinulosa, D. spinulosa intermedia, Polystichum acrostichoides and Polypodium virginianum,; infection resulted on D. marginalis only (Table 4). (c) Milesia polypodophila from Abies balsamea was inocu- lated on to Polypodium virginianum, Dryopteris marginalis and D. spinulosa intermedia; infection resulted on P. virginianum only (Table 6).
8. The reactions of the rust and the various species of Abies in- fected with M. fructuosa exhibited some differences. There were slight variations in the size and form of the spermogonia; but the type re- mained approximately constant. A greater difference was found in the distribution of the spermogonia, particularly on the needles of Abies magnifica. In this fir they were strikingly amphigenous, almost equally distributed over all surfaces of the affected needles and not as in the other firs characteristically hypophyllous. The rusted needles of three of the firs showed some distortion, namely, in A. Fraseri, A. Fra- sert prostrata and A. magnifica. The greatest effect resulted on A. magnifica—marked distortion and dwarfing.
9. Economically certain Milesian rusts are of some importance with respect to the unfavorable effect they exercise on the natural repro- duction of Abies. Milesia polypodophila and M. fructuosa are note- worthy examples. The former causes unsightly and worthless plants of A. balsamea and the latter may kill or hold in check seedlings and sap- lings where the balsam fir is associated with rusted ferns. It is appar- ent from culture experiments that M. fructuosa is potentially a menace to young A. magnifica. Whenever intensive studies are being conducted in regions in which Abies constitutes a part of the forest stand, for the purpose of quantitatively forecasting forest successions, fern rusts as influencing factors should not be disregarded.
PLATE 84
ARNOLD ARB. VOL, XV.
JOUR.
Brotogy oF MILEstAn Rusts
THE HELIOTYPE CORP. BOSTON
JouR. ARNOLD ARB. VOL. XV. PLATE 85
BroLtocy or MILESIAN Rusts
THE HELIOTYPE CORP, BOSTON
Jour. ARNOLD ARB. VOL. xv.
PLATE
Brotoay or MILESIAN Rusts
THE HELIOTYPE CORP. BOSTON
1934] FAULL, BIOLOGY OF MILESIAN RUSTS 85
LITERATURE CITED
1. DieteL, P. (1903). Ueber die iepasdehiseeaiicet ig von Uredo a uscula D. et H. und tuber Melampsora Fagi D. et Neg. (Anna Mycol. 1: 415-417.)
2. Fautt, J. H. (1932). Taxonomy and oa rers distribution of the genus Milesia. (Contrib. Arnold Arb. 2:1-138.)
3. Hunter, L. M. (1927). Comparative study of spermogonia of rusts of Abies. (Bot. Gaz. 83:1-23.)
4. Kame, S. ee . Notes on Milesina vogesiaca Sydow on Poly- stichum Braunii Fée and its peridermial stage on the needles of
ies Magee ae et Kudo, A. firma Sieb. et Zucc. an sachalinensis Mast. (Trans. Sapporo Nat. His. Soc. 11:141- 147.)
5, ————— a): w new species of Milesina parasitic on Polypodium vulgare a (Trans. Sapporo Nat. His. Soc. 12:27-33.)
6. ———— (1932). On new species of Hetereecous fern rusts. (Trans. Sapporo Nat. ee Soc. 12:161-174.)
7. KirepAun, H. (1916). Kulturversuche mit Rostpilzen (Zeitsch. Pflanzenkr. 26:257-277 .)
8. Macnus, P. (1901). Weitere alipme uber die auf Farn- krautern auftretenden Uredineen. (Ber. Deutsch. Bot. Ges. 19 :578-584. )
9. Waters, C. W. (1928). The control of teliospore and urediniospore formation by experimental methods. (Phytopath. 18:157-213.)
EXPLANATION OF PLATES (Photographs made by Mr. A. B. Hatch) Plate 84 Milesia marginalis Faull and Watson. Inoculation experiment #180 (E. H. B.) of 1927. Abies balsamea neculated with telial material from Dryopteris ira datasans collected Jun . . ate of inoculation June
29, 1927; first spermogonia July 17. fully mature eee August 13; harvested August 22, 1927. x 1. (Herbarium J. H. Faull no. 8497.)
Plate 85 Milesia marginalis Faull and Watson. Enlarged portion of Plate 1. Plate 86 ilesia intermedia Faull on seedlings of Abies balsamea growing on the floor of the forest; natural infection. Collected September 12, 1925.
left and five only of pla t on the right will survive; all the others will succumb from the rust attack. & 1. (Herbarium J. H. Faull no. 7393.)
LABORATORY OF PLANT PATHOLOGY, ARNOLD ARBORETUM, HARVARD UNIVERSITY.
86 JOURNAL OF THE ARNOLD ARBORETUM [ VOL. XV
A REMARKABLE SPRUCE RUST, PERIDERMIUM PARKSIANUM, N. SP.
J. H. FAuLy
A spruce rust of exceptional interest from western North America has been brought to light in collections made by Mr. Harold E. Parks on Picea sitchensis Carr. The rust is a Peridermium and its spermo- gonia indicate that it is the haploid phase of a species of Melampsorop- sis. But whatever its diploid phase may prove to be its spermogonial and aecial characters are sufficiently distinctive to separate it easily from other known spruce rusts. Hence, even though the name pro- posed, Peridermium Parksianum, will be superseded by the name of the diploid phase when the latter is determined, in accordance with the accepted nomenclatorial rules, there is ample justification for formal designation of the haploid phase. The most striking characters are those manifested by the spermogonia and the aeciospores. The former are triangular in vertical section and broad-based, resembling tiny Erlen- meyer flasks. Most of the aeciospores are elongated, commonly so much so as to be properly described as linear. Often they are some- what curved and their tips are tapered, the shape suggesting that of a slightly curved Closterium. Occasional spores are even slightly sigmoid.
This rust is restricted to needles of the current season and is often very abundant. The lesions caused by it may involve entire needles or limited portions only; they are somewhat lighter green than normal by the time the spermogonia first appear, but eventually they are straw- colored. The spermogonia form in late spring or early summer. Speci- mens collected June 29, 1933 carry an abundance of spermogonia from which there is a copious outpouring of the globular to broadly ellip- soidal spermatia, still being actively produced (apparently catenu- lately) from the elongated, unbranched spermatiophores that crowd the interior of the spermogonia. The peridermia develop slowly. First signs of them in 1933 were evident August 25 but they were not mature until the middle of September or later. They occur over a compara- tively long period. The earliest collection in my herbarium bears the date September 21 and the latest November 23, the latter carrying many of the yellow-orange aecia still unruptured.
I have attempted to culture the rust on Rhododendron californicum
1934] FAULL, PERIDERMIUM PARKSIANUM 87
Hook., which appears to be a likely host, but so far without success. The negative results cannot, however, be accepted as final because diffi- culty has been encountered in securing germination of the aeciospores.
Peridermium Parksianum is reported from Trinidad, California and Florence, Oregon. The description of the spermogonia given below is based on a collection (Herb. J. H. Faull, no. 11,076) from Trinidad, California made June 29, 1933; the description of the peridermia is based on material (no. 11,002; Herb. H. E. Parks, no. 4052) collected October 15, 1932 from the same locality and checked against materials collected likewise at Trinidad November 23, 1933 (no. 11,592) and September 24, 1933 (no. 11,540). All collections were made by Mr. Harold E. Parks, and the species is named in his honor.
Peridermium Parksianum Faull, n. sp.
Pycnidia hypophylla, numerosa in foliis hornotinis, biseriata, con- spicua, immersa, parum prominula, mellea, dein rufo-brunnea, in facie verticali triangula, subepidermalia, 228-315 \: lata et 125-155 y alta, plus minusve 260 140 1; spermatia globosa vel ellipsoidea, 2.7- 3.2 & 3.0-5.0 up. Aecidia hypophylla, in foliis hornotinis, irregulariter biseriata, maculis flavis insidentia, cylindracea vel lateraliter com- pressa, 0.4-1.2 mm. longa et 1.0-2.2 mm. alta, ad apicem dehiscentia; cellulae peridii albae, subimbricatim positae, 18-26 48-88 1, pariete exteriore levi, 0.7-1.0 ,) crasso, interiore dense et subtiliter verrucoso, circa 1.5 uy crasso; aecidiosporae fusiformes vel anguste ellipsoideae, rarius subglobosae, rectae vel curvulae, flavae, 13-24 & 25-111 1, prae- cipue fusiformes et plus minusve 16 X 70 1, dense et subtiliter verru- cosae, episporio circa 1 u crasso.
Hab. in foliis Piceae sitchensis in California et Oregon Americae bor.
LABORATORY OF PLANT PATHOLOGY, ARNOLD ARBORETUM,
HARVARD UNIVERSITY.
88 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
NOTES
The Beach Plum in Michigan. — In September 1932 we received for identification from Mr. James E. Armstrong of Chicago, fragmen- tary material of a plum that he had found growing on gravelly ridges in Lake County, west-central Michigan. From the somewhat withered leaves and the fruit this appeared to be a form of the variable beach plum (Prunus maritima Wang.), which has been previously known only from a limited area along the Atlantic coast from Maine to Virginia, where it grows, usually as a straggling shrub, along beaches and on sand dunes not far from the sea.
In subsequent correspondence with Mr. Armstrong he stated that the Michigan plum formed thick clumps of shrubs about six feet tall, and that it was apparently local and not very abundant in the vicinity of Walhalla, in the Lake region.
On my trip to the west this year, in company with Mr. E. D. King, I visited the region, and following the careful directions of Mr. Arm- strong had no difficulty in locating the place and finding the plum, although not in the exact locality where he had found it.
Walhalla is a village in Lake County, about twenty or twenty-five miles east of Luddington, which is on Lake Michigan. The region is glaciated and includes a number of small lakes. Most of the sandy and gravelly uplands, where not cleared, are covered with a mixed growth of coniferous and deciduous forest.
Where we found the plums, a mile or two south of the village, they seemed to be locally abundant in openings of the sandy upland woods. The shrubs are usually scattered and are seldom more than four or five feet in height. They were in young fruit at the time of our visit, June 19th, but due apparently to a frost a short time before, most of the fruit was blighted, except in a few protected places where it was still abundant. I was told by a resident that the sand plums, as they seem to be called in that section, were common and that they were gathered for making jams and jellies.
A series of herbarium specimens was made and living plants were secured for propagation at the Arboretum. Study of this material and observation of the plants in the field leave no doubt as to their identity and confirms this interesting and remarkable extension of the known range of the beach plum. — Ernest J. PALMER.
1934] NOTES 89
Quercus ellipsoidalis in Missouri. — While collecting last summer in northern Missouri, in company with Dr. Julian A. Steyermark, we came upon an oak in Harrison County of a species that apparently had not been previously found or reported for the state. The distribution of the jack oak (Quercus ellipsoidalis E. J. Hill) and its relation to that of the scarlet oak (Q. coccinea Muenchh.) has recently been worked out and discussed in an interesting manner by Mr. S. C. Wadmond.' On the distribution map accompanying his paper the range of the jack oak is shown as extending from northern Indiana, across Michigan, Wisconsin, northern Illinois, Minnesota and Iowa, with several records in the last named state immediately north of the Missouri locality reported here.
Harrison County, one of the border counties of northwestern Mis- souri, is in a glaciated region underlaid by Pennsylvanian formations of sandstones, shales, and limestones, and the topography and _ general character of the country, and that of the immediate locality, on the bor- der of prairie and upland forest where Quercus ellipsoidalis was found, are quite characteristic for the species. Only a single tree was noted on our hurried trip, but no doubt others could be found in the vicinity, and the species should be looked for in other northern counties of Missouri, where a number of other northern plants reach their southern limits west of the Mississippi. — ERNEST J. PALMER.
1Trans. Wisconsin Acad. Sci. Arts and Letters, 28:197—203 (1933).
JOURNAL
OF THE
ARNOLD ARBORETUM
VOLUME XV. APRIL, 1934 NUMBER 2
NOTES ON THE LIGNEOUS PLANTS DESCRIBED BY LEVEILLE FROM EASTERN ASIA?
ALFRED REHDER
ELAEOCARPACEAE
Sloanea sinensis (Hance) Hu in Jour. Arnold Arb. v. 230 (1924). Castanopsis Cavaleriei Léveillé in Fedde, Rep. Spec. Nov. x11. 506 (1913); Fl. Kouy-Tchéou, 125 (1914).— Camus, Chataigniers,
; nov. Sloanea Hanceana Léveillé, Fl. Kouy-Tchéou, 420 (1915), vix enisley. Sloanea chengfengensis Hu in Sinensia, ur. 85. (1932).— ynon. nov.
Cuina. Kweichou: sud de Pin-fa, J. Cavalerie, no. 2514, Sept. 24, 1905 “arbre moyen, rare” (holotype of Castanopsis Cavaleriei; merotype in A. A.); Yang-kia-wan, Cheng-feng, in mixed woods, IY. Tsiang, no. 4641, Oct. 17, 1930 “‘tree 27 m., bark grey; fruit red” (holotype of S. chengfengensis; isotype in A. A.).
In his Flore du Kouy-Tchéou Léveillé enumerates Cavalerie’s no. 2514 under Castanopsis Cavaleriei and also under Sloanea Hanceana. Castanopsis Cavaleriei had been already referred to Sloanea by Handel- Mazzetti (Symb. Sin. vir. 29, 1929) as Sloanea sp.; by A. Camus, l. c., it is enumerated as a doubtful species. Sloanea Hanceana Hemsl. = S. Hemsleyana (Ito) Rehd. & Wils. is very close to S. sinensis from which it differs chiefly in its larger leaves, serrate to near the base and in the shorter, slenderer and denser spines of the fruit.
I am unable to distinguish Sloanea chenfengensis from S. sinensis. Hu says it differs chiefly in the glabrous leaves and in the 4-celled fruits, but in S..simensis the number of locules varies as in most species between 4 and 6 and the leaves are called by Hance in the original description ‘“‘glaberrima” and so they are in all the specimens examined.
Ban inued from p. 27; for ee parts see Vol. x. 108-132, 184-196; xm. 5-281; xt. 299-332; xXIv. 223-2
92 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
TILIACEAE Tilia tuan Szyszylowicz in Hooker, Icon. Pl. xx. t. 1926 (1890). Tilia tuan var. Cavaleriet V. Engler & Léveillé in Fedde, Rep. Spec. Nov. vi. 266 (1909),— V. Engler, Monog. Tilia, 124 (1909), — Schneider, Ill. Handb. Laubholzk. 11. 389 (1909). — Léveillé, Fl. Kouy-Tchéou, 420 (1915).
Cutna. Kweichou: Pin-fa, montagnes, au sud, Cavalerie no. 3227, May 30, 1907, “vu 7 ou 8 petits arbres, fl. blanc-jaune” (holotype of 7. tuan var. Cavaleriei; photo. in A. A.).
Tilia tuan var. Cavaleriei represents the type of T. tuan according to Engler who has it as “var. a,” and to Schneider. It differs only slightly from the type specimen in the somewhat narrower leaves, the upper ones measuring 9.5 by 4 cm. to 12 by 5.5 cm.
Tilia Miqueliana Maximowicz in Bull. Acad. Sci. St. Pétersb. xxvr. 434 (1880); in Mél. Biol. x. 687 (1880).— V. Engler, Monog. Tilia, 111 (1909).
Tilia Kinashii Léveillé & Vaniot in Bull. Sci. Bot. France, xii. 422 1904
CENTRAL JAPAN: Kinashi no. 2, (holotype of 7. Kinashii; photo. in A. A.).
Tilia Kinashii was already referred to T. Miqueliana by V. Engler and according to a note on the type specimen also by G. Koidzumi.
Grewia biloba G. Don, Gen. Syst. 1. 549 (1831).— Handel-Maz- zetti, Symb. vir. 612 (1933).
Grewia glabrescens Bentham, Fl. Hongk. 42 (1861). — Léveillé, FI. Kouy-Tchéou, 419 (1915).
Grewia Esquirolii Léveillé, FI. ies -Tchéou, 419 (1915), pro synon.
Celastrus euonymoides Léveillé, 1, (1915), pro synon.
Grewia parv iflora \ se glabrescens ‘Rehder & Ww ilson in Sargent, PI. Wilson. 11. 371 15).
CuHIna. Kweic . ou: Gou-réou, J. Esquirol, no. 3189, Dec. 1911 (probable type of G. Esquirolii; photo. in A. A.); Lo-fou riviere, J. Cavalerie, no. 3513, April 1909, “arb. 2 m.” (type of Celastrus euonymoides; photo. in A. A.); same locality, J. Esquirol, no. 2204, Sept. 1910, and J. Cavalerie, no, 3492, Aug. 1909.
Grewia Esquirolii and Celastrus euonymoides are apparently manu- script names published only as synonyms of G. glabrescens; the name C. euonymoides appears on the label of Cavalerie’s no. 3513, but on Esquirol’s no. 3189 no name appears except Euvonymus which is crossed out. Esquirol’s no. 2204 is named G. glabrescens Benth. in Léveillé’s handwriting, while Cavalerie, no. 3492, bears no name _ except ‘“Grewia?’’, but was placed in the cover of G. glabrescens.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 93
I am following Handel-Mazzetti in considering Grewia glabrescens identical with G. biloba. Don describes the leaves as smooth above, but pubescent beneath, which agrees with Bentham’s description of the leaves as glabrous or sparingly sprinkled with a few short hairs on the upper side and dotted with a minute stellate pubescence under- neath, while Bunge describes the leaves of his G. parviflora as “supra hispidulis, subtus stellato-canescentibus.” The latter, therefore, be- comes a variety of G. biloba. The two varieties, however, cannot be sharply separated and are closely connected by intermediate forms. There is also a difference in the shape; typical G. biloba has the leaves generally longer, oblong-oval to oblong-lanceolate, while those of var. parviflora are shorter and broader, ovate or oval, and usually smaller.
Grewia biloba var. parviflora (Bge.) Handel-Mazzetti, Symb. Sin. vir. 612 (1933). Grewia biloba Burret in Notizbl. Bot. Gart. Mus. Berlin, 1x. 708 y 927
Rubus eG Léveillé in Fedde, Rep. Spec. Nov. vr. 111 (1908) ; cad. aoe ape Bot. x1x. Mém. 16 (Ronces Chin. Tap.) (1909), — Symon Grewia Chanetii ae in "Fedde, Rep. Spec. Nov. x. 147 (1911). Cuina. Hopei: Kia-chan, L. Chanet, no. 571, Aug. 1910 (holotype of G. Chanetii; photo. in A. A.); mont de Pong-chan, L. Chanet, June 15-16, 1910 (in herb. Léveillé sub G. Chanetii; photo. in A. A.). Yunnan: prés Pin-tchouan, Jean Py in herb. Ducloux, no. 618, April 27, 1906, “‘fleurs jaunes” (holotype of Rubus umbellifer in Herb. Univ. Calif.; photo. in A. A.). The specimen from vonnat approaches the type, but is rather densely pilose above and most of the leaves are oval or elliptic. Grewia Chaneti had been already identified as G. biloba by Burrett (1. c.).
Grewia abutilifolia Ventenat apud Jussieu in Ann. Mus. Paris, Iv. 92 (1804).— Burrett in Notizbl. Bot. Gart. Mus. Berlin, rx. 723 (1926).
Sterculia tiliacea Léveillé in Fedde, Rep. Spec. Nov. xi. 185 (1913) ; Kouy-Tchéou, 406 (1915).— Synon. nov.
Cuina. Kweichou: canal d’eau en amont de Ouang-mou, J. Esquirol, no. 47, May 26, 1904, ‘“‘arbrisseau, fl. jaunes” (holotype of Sterculia tiliacea; photo. in A. A
The leaves of the specimen cited are very broad, some even broader than long, about 9 cm. long and 10 cm. broad, and some show a slight tendency toward lobing, as in Henry’s no. 9887p which was identified by Burrett (1. c.) with G. abutilifolia.
94 JOURNAL OF THE ARNOLD ARBORETUM [VOL. XV
MALVACEAE Abutilon indicum (L.) Sweet, Hort. Brit. 54 (1827). — Léveillé Cat. Pl. Yun-Nan, 175 (1916). Abutilon Cavaleriet Léveillé in Fedde, Rep. Spec. Nov. x1. 185 (1913) ; Fl. Kouy-Tchéou, 272 (1914). — Synon. nov. Cuina. Kweichou-:ouest de Lo-fou, J. Cavalerie, Nov. 1905 (holotype of A. Cavaleriei; photo. in A. A.).
Urena lobata Linnaeus, Spec. Pl. 692 (1753). — Léveillé, Fl. Kouy- Tchéou, 275 (1914); Cat. Pl. Yun-Nan, 176 (1916).
Abutilon Esquirolti Léeveillé in Bull. Géog. Bot. xxiv. 252 (1914); Fl. Kouy-Tchéou, 272 (1914).— Synon. nov.
CuHInA. Kweichou: Gou-réou, dans les herbages, rare, 700 m., J. Esquirol, no. 3668, Sept. 1912, “haut 1.5-2 m., fl. roses” (holo- type of Abutilon Esquirolit; merotype in A. A.).
A form with ovate to orbicular-ovate undivided only sparingly denticulate leaves and the flowers clustered at the apex of the branches, not axillary along the stems.
Hibiscus Labordei Léveillé in Fedde, Rep. Spec. Nov. xm. 184 (1913); Fl. Kouy-Tchéou, 274 (1914).
Cuina. Kweichou: montagne de Kao-po (Tsin-gay), J Laborde in herb. Bodinier, Sept. 10, 1899 “grand arbuste”’ (holotype; photo, in A. A.).
This species I have not been able to identify with any of the described species; it seem to be nearest to H. mutabilis L. from which it differs chiefly in the 3-lobed leaves, the short pedicels 1-1.5 cm. long, and in the dense ochraceous tomentum of the calyx which is about as long as the 6 linear-lanceolate bracts.
Hibiscus crinitus (Wall.) G. Don, Gen. Syst. 1. 480 (1831).
Hibiscus cancellatus wera Hort. Bengal. 51 (1814), nom. nud.: ‘l. Ind. 11. 201 (1832).— Hochreutiner in Ann. Jard. Bot. Geneve, iv. 149 oe veillé, Cat. Pl. Yun-Nan, 175 (1916).— Non Linnaeus f.
Hibiscus Cavaleriet Léveillé in Fedde, Rep. Spec. Nov. xr. 184 (1913); Fl. Kouy-Tchéou, 274 (1914). — Synon. nov.
Hibiscus Bodiniert Léveillé, 1. c. (1913); 1. ¢., 273 (1914); Cat. Pl. Yun-Nan, 175 (1916).—Synon. nov.
CHInA. Kweichou: montagnes de Lo-fou, J. Cavalerie, no 2584, Nov. 1905 (holotype of H. Cavaleriei; photo. in A. A.); Hoa- kiang, J. Cavalerie, no. 3959, July 1912 (cited in Fl. Kouy-Tchéou under H. Cavaleriei; dupl. in A. A.); Kouy-yang, descente du fleuve, E. Bodinier, no. 503, July 27, 1900, “fl. jaunes” (syntype of H.
1934] REHDER, LIGNEOUS PLANTS DESCRIBED BY LEVEILLE 95
Bodinieri; merotype in A. A.); without locality, J. Esquirol, no. 89, June