on 8 Ey) ttle oe oe Beem noe ‘ gabe id ae ig. Aine we MO yh See ree oe ee ee Phew OM ee PA eb we ed ee we ve Wrekin ew Ew Cw ewe Cad ee be tae Se Se ee er re ee rece SM EM to EFL A Ek uh Bot 8 otk Pegi CT oe ave vot a ba ee ee ae pe ete . Ce eC er ee ' Ce eg a 4 0 y ae De on tua tee ea a : peo Vd ro tee ye Fob de Ka d6 $A, By re ce ‘ . er ee a Fo oe ee ey eae ee ‘ ‘ wee . ere ee ” ' ee ere eee a 5 . tis 8 pve we te een ae eae ee yw pea Vem re ® Ga Baie pet wae hal . [oi Oboe Hy ag ox a \ . ‘ ot . he +a Pare te ted w be Fe ee oy ewe 8) ela eh be aw ” Peary ae cue eee : oa ‘ rer est ws ‘ ay ee en i i wae ee ee nt ° so 4 hore : wy Teg ran cue : ‘ ’ , ea Ce ee Ce Gah avi 2s car oan af hess Was ‘ ‘ a) di cab Ae queef. Ged ae po ‘ eee ae Sr ay ' stam ‘ 6 RA PD ee Pe BR we oe Par, : eat ht Par er ee ee ee ’ i cap que Dal Le a Ron hae eds ag, a yay Sant “ ’ . ' ran nang Pee eee ee ; PW gray on, Sethi tee Aa gh ae fa cheuey tree se aherd ule 2D on se wat ‘ eevee nea be Gee gt Ra ome an re eee Ou ” . oa heap et en bat Perec ree ee a eed Son apr oweae ba oh $0 a toon no wy Cr Ta er ase oe ons ‘ aa Sart ere eC ee ‘ . eit a 9s a ‘ Pi Re Ra a eg i an oe rey Ce ee eae OEM Ht Cad peak be deg , no he Perret) Se cee 6 he gh gp AN ae oy fn ca ede ra So a ee ne ae ee eee eae Ce ee) ' wre woe oan ' you “ eae ' 3 ie ‘ A ie ren 4 . ‘ hae @ : ' . ‘ Pees ae . fiavead Sey eae tere Ce ee oar See) neers a a a oT) ” Sagem etn yD Ams ” wo ria Sn “ ‘ ee y Ne er . . CWen deg Da eS " jos ‘ Tie ne ry ne . ene CeCe eT Pere eae) i smh eae ‘ ‘ ‘ Pa ae) ‘ tae . ’ ‘ res fae te ‘ soa te ee Bak oan yam vet vst i a ne ; tee * see ae oy at oad ete eee ‘ . i H : top bee eee 0 er espiprag. + & be 4 ‘ et Pan ante Po sh We Cree er ae ee we wpe - ee mae Ae et te sen Pret eT sone ie deat ares Peer ear ey ee Pe ee ce gee ae 5 Pear bm ae ons rem carer aT] eee sx 4 MADRONO. A WEST AMERICAN JOURNAL OF BOTANY VOLUME XIV 1957-1958 BOARD OF EDITORS HERBERT L. Mason, University of California, Berkeley, Chairman EDGAR ANDERSON, Missouri Botanical Garden, St. Louis LymMAN BENSON, Pomona College, Claremont, California HERBERT F. COPELAND, Sacramento College, Sacramento, California Joun F. Davipson, Department of Botany, University of Nebraska, Lincoln Ivan M. Jonnston, Arnold Arboretum, Jamaica Plain, Massachusetts Mivprep E. Marurias, Department of Botany, University of California, Los Angeles MARION OWNBEY, State College of Washington, Pullman Ira L. Wiccrns, Natural History Museum, Stanford University, Stanford, California SECRETARY, EDITORIAL BOARD ANNETTA M. Carter, Department of Botany, University of California, Berkeley BUSINESS MANAGER AND TREASURER Marco A. Noss, Carnegie Institution of Washington, Stanford, California (1957) WINSLow R. Briccs, Department of Biology, Stanford University, Stanford, California (1958) Published quarterly by the California Botanical Society, Inc. 2004 Life Sciences Building, University of California, Berkeley Printed by Gillick Press, Berkeley, California (IHSON]. NOV5 19588 Ww To ROXANA STINCHFIELD FERRIS, Assistant Curator of the Dudley Herbarium of Stanford University, the California Botanical Society dedi- cates this, the fourteenth volume of Madrono. Known affectionately to all of your many friends as “Roxy,” it has been your lot to perform much “spade work” and to have reaped only that concealed reward of the personal satisfaction resulting from a job well done. We refer to the editing, and to the writing and rewriting and to the study of many groups of plants included in volumes IT and III, and to your present considerable contribution of actual manuscript to the forthcoming fourth and final volume of Abrams’ “‘Tllustrated Flora of the Pacific States.’ Few authors have had such loyal and well trained assist- ance; few assistants have been so willing to suffer uncomplainingly the drudgery and heartaches that go into labor of this kind. We bring to you the affectionate greetings of our membership and wish you many more years of gratifying activity in Botany. CALIFORNIA BOTANICAL SOCIETY PUBLISHERS OF MADRONO REPORT OF THE TREASURER FOR 1957 RECEIPTS: Balance on hand in commercial account, January 15, 1957........ $ 840.30 From memberships and subscriptions...................222-..----2:--.-200---2++ 1,987.06 From sales of back numbers of Madrono........................2...2..-.--- 58.00 IVECEMDUS DEOla. ATT GINO. <2 22 cce ..<0hecocsec ebcelcus ad, cess cdeeceeeeeee swe was 7s) Received as authors’ share of publication costs.........................--- 360.00 Contributions to memorial FUNG: iccoccciccit22cecctectescscecdececccesccessesseese 75.00 MO tale Ge Celes eee see ced ieee le ee $3,552.41 DISBURSEMENTS: Credited to endowment fund from sales of back numbers of NY COOH CCM eee pee eee ACEP SENS A 2 ee ee Se $ 58.00 Credited: 16 memorial fund 222 22x28 est eo 75.00 Wostxo tpanmual Ginmer. coscc8 seco 2th corde teen vaees2 en nee isis Corresponding Secretary’s EXPeNnsSes.......<...-c2-- Geena een Ra en ee ore De te Seer eerttey pee MIC nT Even eacndas tne Eero Ceo ef A Ue 276 ERRATA Page 3, following line 35 insert: of order Parietales, in the Englerian system. Dilleniaceae. Page 40, line 15: for Peirce read Pierce. Page 40, line 19: for Taris read Tario. Page 70, line 32: for (law-w) read (lav-w). Front cover, vol. 14, no. 3, line 6: for Minulus read Mimulus. Page 96, line 20: for Robert read Rupert. Page 115, line 5: insert Idaho in front of Elmore County. Page 147, line 1: transpose right and left. Page 153, line 34: for 3—dm. 7 read 3-7 dm. Page 218, line 35: for 1956 read 1957. Page 225, line 10: for saggitate read sagittate. Page 232, line 34: for no read so. ADRONO VOLUME 14, NUMBER 1 JANUARY, 1957 Contents PAGE FORECAST OF A SYSTEM OF THE DICOTYLEDONS, Herbert F. Copeland 1 A Botanical DisAsTER, Erwin F. Lange 9 RoBERT HiBBs PEEBLES, Thomas H. Kearney 11 THE CHROMOSOMAL AND DISTRIBUTIONAL RELATION- SHIPS OF LUPINUS TEXENSIS AND L. SUBCARNOSUS (LEGUMINOSAE), B. L. Turner 13 STUDIES ON THE CROSSABILITY OF LUPINUS TEXENSIS AND LUPINUS SUBCARNOSUS, Lawrence Erbe 17 New NortH AMERICAN ANDROPOGONS OF SUBGENUS AMPHILOPHIS AND A KEY TO THOSE SPECIES OcCUR- RING IN THE UNITED STATES, Frank W. Gould 18 CHROMOSOME NUMBERS IN LuPINus, Lyle L. Phillips 30 JEROME D. LAUDERMILK, Lyman Benson 36 ASTRAGALUS AGNICIDUS, A NEW LOCOWEED FROM HuM- BOLDT COUNTY, CALIFORNIA, R. C. Barneby 37 Notes AND NEws: ANEMOPSIS CALIFORNICA IN OrEGON, Russell Pengelly; NEws 40 A WEST AMERICAN JOURNAL OF BOTANY | PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO A WEST AMERICAN JOURNAL OF BOTANY Entered as second-class matter at the post office at Berkeley, California, January 29, — 1954, under the Act of Congress of March 3, 1879. Established 1916. Subscription price | $4.00 per year. Published quarterly and issued from the office of Madrofio, Herbarium, Life Sciences Building, University of California, Berkeley 4, California. BOARD OF EDITORS HERBERT L. Mason, University of California, Berkeley, Chairman EpGAR ANDERSON, Missouri Botanical Garden, St. Louis. LYMAN BENSON, Pomona College, Claremont, California. HERBERT F. COPELAND, Sacramento College, Sacramento, California. JouN F. Davipson, University of Nebraska, Lincoln. IvAN M. JoHNSTON, Arnold Arboretum, Jamaica Plain, Massachusetts. Mivprep E. MaAruias, University of California, Los Angeles 24. MarIon Owneey, State College of Washington, Pullman. IrA L. Wiccrns, Stanford University, Stanford, California. Secretary, Editorial Board — ANNETTA CARTER - Department of Botany, University of California, Berkeley. Business Manager and Treasurer—Matcotm A. Noss Carnegie Institution of Washington, Stanford, California CALIFORNIA BOTANICAL SOCIETY, INC. President: Wm. M. Hiesey, Carnegie Institution of Washington, Stanford, Cali- fornia. First Vice-president: Roger Reeve, Western Regional Research Laboratory, Albany, California. Second Vice-president: Lewis Rose, California Academy of Sciences, San Francisco, California. Recording Secretary: Mary L. Bowerman, De- partment of Botany, University of California, Berkeley, California. Corresponding Secretary: G. Thomas Robbins, Department of Botany, University of California, Berkeley, California. Treasurer: Malcolm A. Nobs, Carnegie Institution of Wash- - ington, Stanford, California. - ; FORECAST OF A SYSTEM OF THE DICOTYLEDONS HERBERT F. COPELAND! A standing problem of taxonomic biology — its importance and diffi- culty made known by the incompletely successful efforts of fully two cen- turies — is that of the natural and convenient organization of the families of dicotyledons in groups of the category of orders. De Candolle (1813; to him, what we call a family was an order, and what we call an order was a subclass, legion or cohort) remarked of this problem, “‘C’est la le probleme le plus important a résoudre qui se présent aujourd’hui dans |’étude des rapports naturels.” And Schnarf (1933) had still to say, “‘Dagegen ist die Gruppierung der Familien zu Ordnungen oder Reihen schon mit einer gewissen Unsicherheit behaftet, die darin zum Ausdruck kommt, dass die Abgrenzung der Ordnungen von verschiedenen Forschern vielfach in ver- schiedener Weise vorgenommen wird.” In his work just cited, De Candolle observed that there had been almost as many systems as systematists; which is not far from saying that every systematist has a right to his own system. Among the very many systems which systematists have produced, few have had much influence. During the nineteenth century, the system of De Candolle (of which that of Ben- tham and Hooker is a variant) overshadowed all others; subsequently, the system of Engler and Prantl has had the same effect. All this is as it should be. We need at every time to have an accepted system, by which we may know where to look for what concerns us in herbaria and manuals. The systems which have been offered as challenges to the accepted systems have brought about minor improvements in the latter. They have had the more important effect of keeping us aware that the accepted system is never the final truth. As the system of Engler and Prantl displaced that of De Candolle, so surely it will be displaced by one which is recognizably a more satisfactory representation of the system which exists in nature. Considering these things, I took a summons to appear in a symposium as occasion to try to predict the system of the future; this to the extent of formulating the skeleton of a system which is set forth below. This system gives much weight to microscopic characters, though I can scarcely claim mastery of the great mass of available data. Metcalf and Chalk (1950) give a bibliography of about twenty-five hundred titles, nearly all of them subsequent to the translation of Solereder by Boodle and Fritsch (1908). Schnarf (1931) listed about seventeen hundred con- tributions to embryology. Wodehouse (1935) listed some three hundred 1 Tt was a high honor, accepted with diffidence, to be asked to speak on “Anatomy and taxonomy” in a symposium commemorating the fiftieth anniversary of the Botanical Society of America, on August 28, 1956, at the meeting of the American Institute of Biological Sciences at the University of Connecticut. The present paper is a revision of the one given on that occasion. Maprono, Vol. 14, No. 1, pp. 1-40, January 29, 1957. FEB 5 2 MADRONO [Vol. 14 papers on pollen grains, and Erdtmann (1952) about eleven hundred in the same field. Where study of vascular anatomy, at Harvard and else- where, has yielded definite conclusions, it has been possible to formulate orders with considerable confidence; for the rest, the orders are those of Engler and Prantl with certain amendments suggested by embryological or palynological data. Large orders are preferred to small: in maintaining an order of two or three families, one is not far from leaving the families unplaced. However, it has not been found possible to be consistent in this matter. It is not con- sidered necessary that the orders be definable by description. Names are applied to the orders according to the principles of priority and typification. The experience of very many taxonomists has shown it necessary to apply these principles in dealing with genera and species. Experience with the formulation and naming of higher groups has been comparatively scant. Perhaps for this reason, taxonomists in convention have taken the position that it is possible to make defensible choices among the fairly numerous names which have been applied to higher groups without recourse to the principles cited. Whether or not this position is sound, no novelties are here added to the synonymy of ordinal names. Subclass DICOTYLEDONES Lindley Synops. British Fl. 4 (1829). Order 1. MULTISILIQUAE L. Gen. Pl. ed. 6 (1764). Orders Piperitae and Coadunatae L. op. cit. Orders Piperinae (Brongniart, as class) and Poly- car picae (Endlicher, as class) Braun in Ascherson FI]. Brandenburg 1: 36, 47 (1864). Orders Piperales and Ranales Engler Syllab. 93, 106 (1892). Among names of the three natural orders in which Linnzus placed these plants, that of which the apparent typical genus is Ranunculus is preferred to those of which the apparent typical genera are respectively Piper and Magnolia. This order includes the generality of dicots with apocarpous flowers, as well as some of their immediate derivatives. They have been studied extensively by Bailey and his associates (Bailey and Nast, 1943, 1944, 1945; Bailey, Nast, and Smith, 1943; Bailey and Smith, 1942; Bailey and Swamy, 1949; Smith, 1943, 1945, 1946, 1947; Swamy, 1949; Swamy and Bailey, 1949). These scientists are authority for the assemblage here of a large number of families, including the Piperacee and their immediate allies. To current botanical opinion, this is definitely the primitive order of flowering plants. Some families are homoxylous, 7. e., having wood without vessels, in contrast to the heteroxylous condition which is characteristic of flowering plants. In some families the pollen grains are monocolpate, 7. e., marked by a single groove; this character they share with the gener- ality both of the lower seed plants and of the monocots, while in typical dicots the pollen grains are tricolpate or of more elaborate types derived from this. The two primitive characters mentioned, and the peculiar 1957] COPELAND: DICOTYLEDON SYSTEM 3 anatomical character of oil cells in the tissues, are distributed among the families of Multisilique each one in seeming independence of the others. This means that, quite as one would expect of a primitive group, the families are isolated or fall into isolated blocks. It is necessary as a matter of convenience to maintain the order as a whole; it is not possible to divide it into a small number of natural orders. The families are as follows: a. Homoxylous, pollen monocolpate: Winteracee. b. Heteroxylous, pollen monocolpate: Degeneriacee, Himantandracee, Magnoliacez, Anonacee, Eupomatiacee, Myristicacee, Canellacee, Mon- imiacee, Gomortegacee, Lauracee, Hernandiacee, Lactoridacee, Caly- canthacee, Chloranthacez, Piperacee, Saururacee. c. Wood degenerate, pollen monocolpate: Nymphzacez (pollen tricol- pate in subfamily Nelumbonoidez). d. Homoxylous, pollen tricolpate: Trochodendracee, Tetracentracez. e. Heteroxylous, pollen tricolpate: Eupteleacee, Cercidiphyllacee, Illi- ciacee, Schisandracee, Berberidacee, Menispermacee, Lardizabalacee, Ranunculacee. Order 2. JULIFLORAE (Endlicher as class) Braun in Ascherson FI. Bran- denburg 1: 62 (1864). Order Amentacee of Linnzus and Jussieu: one would not maintain a name in -ace@ as that of an order. Tippo (1938), on the basis of studies of the anatomy of the wood, assembled the fami- lies Hamamelidacee, Platanacee, Myrothamnacee, Stachyuracee, Betu- lacee, Fagacee, and Casuarinacee, as a natural group derived immediately from Multisilique. It is an annoyance to have no definite opinion as to the natural place in the system of the familiar families Salicaceze and Juglandacee. Gundersen (1950) grouped Juglandacee with Myricacee and Rhoipteleacee, which may well be sound; but there is not much to tell where the group belongs. Order 3. SCABRIDAE L. Order Urtice Jussieu, the mere plural of a gen- eric name. Order Urticine (Bartling as class) Braun. Order Urticales Engler Syllab. 95 (1892). Ulmacee, Eucommiacee, Moracee, Urticacee. Study of the woods by Tippo (op. cit.) was held to confirm this generally- accepted group as natural, and to show that its origin was from the lower Juliflore. Order 4. GUTTIFERAE Jussieu Gen. Pl. 225 (1789). Suborder Theinee, Theacee, Marcgraviacee, Caryocaracee, Medusagynacee, Clusiacee, Hypericacee, Quiinacee, Eucryphiacee, Ochnacee, Dipterocarpacee. Vestal (1937) found the anatomy of the woods to confirm as natural this generally accepted assemblage of families. They show nice transitions from primitive vessels with barred perforations to advanced vessels with porous perforations. Similarly in the flowers, there are transitions from spiral parts of indefinite numbers to whorled parts of definite numbers, while the endosperm varies from nuclear to cellular (Schnarf on Saurauia, 1924; Swamy on Marcgravia, 1948). The basic family Dilleniacee is 4 MADRONO [Vol. 14 needed to bind together this assemblage; it might otherwise as well be included in Multisilique. Order 5. Bicornes L. Orders Rhododendra and Erice Jussieu Gen. Pl. 158, 159 (1789). Order Ericales Engler Syllab. 151 (1892). The vessel perforations vary in the Bicornes from barred to porous. The flowers are characteristically sympetalous (there are both primitive and derived ex- amples with separate petals) ; the stamens are free of the corolla, with no ribbed endothecium (except in the primitive family Clethraceze), the anthers opening through pores, the pollen grains united in tetrads. The endosperm is cellular: the first two divisions of the endosperm mother cell are transverse, producing a row of four cells, among which the terminal members give rise to haustoria. Nearly all authorities agree that this group is immediately related to Saurauia, which belongs in or next to Actini- diacee. The families are Clethracee, Ericacee, Empetracee, and Epacri- dacee. In many Epacridacez, the stamens are epipetalous and the anthers open through slits, and the pollen grains are solitary; but these plants are linked to Ericacez by clear lines of transition. The families Lennoaceze and Diapensiacee, which have been placed in this order, do not belong to it, and are for the present left unplaced. Order 6. GUIACANAE Jussieu Gen. Pl. 155 (1789). Order Diospvyrine (Brongniart as class) Braun in Ascherson FI. Brandenburg 1: 37 (1864). Order Ebenales Engler Syllab. 155 (1892). Styracacee, Sapotacee, Sym- plocacee, Ebenacez, and other families. De Candolle condemned tatonne- ment (fumbling!) as a method of recognizing the natural system; yet it was the accident that I have Styrax in near-by foothills and a plant of Camellia in my back yard that enabled me to see that Styracacee is im- mediately related to Theacez. Order 7. PASSIFLORINAE (Brongniart as class) Braun in Acherson FI. Brandenburg 1: 50 (1864). Order Rotacee L., in part. Order Cistz Jussieu, the mere plural of a generic name. Order Parietales (Endlicher) Braun op. cit. 49. Cistiflore Eichler. Vestal (1937) assembled the families Fla- courtiacee, Bixaceze, Cochlospermacee, and Cistacee as a natural group descended directly from Multisilique. Whether the herbaceous families Passifloracee, Caricacee, Cucurbitacee, and Begoniacee, usually placed with these, belong with them or belong together, and whether Violacez and Resedacee belong with them, is apparently as yet uncertain. Order 8. SENTICOSAE L. op. cit., the evident standard genus being Rosa. Orders Papilionacew, Lomentacee, and Pomacee L. Order Rosiflore (Endlicher) Braun. Order Rosales Engler Syllab. 115 (1892). Tippo (1938) showed that Saxifragacee sens. lat., Brunelliacee, Cunoniacee, and Rosacee belong together. Presumably Crassulacee, Pittosporacee, and Leguminose belong with them. They are derived directly from woody Multisilique. 1957] COPELAND: DICOTYLEDON SYSTEM 5 Order 9. RHOEADEAE L., including orders Corydales, Putaminee, and Siliquose L. Order Rhoeadine (Bartling as class) Braun. Order Rhoea- dales Engler Syllab. 111 (1892). Papaveracee, Tovariacee, Fumariacee, Capparidacee, Crucifere; Moringacee and Violacee have been placed here. They are believed to be derived directly from Multisilique. To this point in this presentation, it is believed that the truth as to the main outline of the phylogeny of the dicots has been perceived. The order Multisilique is primitive; Juliflore, Guttiferz, Passiflorine, Senticose, and Rhoeadez are immediate derivatives; Scabridz, Bicornes, and Guia- cane are secondary derivatives. A considerable number of further orders are recognizable, but their connections are less clear. Gundersen (1950) took note that most compound pistils with axile placentation pass during development through a stage in which the placentation is parietal. Consid- ering this fact in connection with the principle that ontogeny recapitulates phylogeny, he thought it probable that the group here called Passiflorinz is an important secondary center of variation, ancestral to most of the re- maining dicots. It is arguable, on the contrary, that parietal placentation is not usually a primitive character, but a result of pedogenesis, that is, of courses of evolution by which the immature condition of a relatively primitive organism becomes the mature condition of its derivatives. Hal- lier (1905) would have derived many of the more advanced dicots from Sterculiacez. Purely as a speculation, it is here suggested that more orders than Bicornes and Guiacane may be derived from Guttifere. Order 10. PrectAE L., including orders Rotacee, Caryophvilei, and Holoracee and Succulente L. in part. Orders Curvembrve, Centros perme, Polygonales, Opuntales, Primulales, and Plumbaginales of Engler and others. A matter of fifteen families, decidedly varied in gross characters. The bulk of the families are characterized by curved or coiled embryos. Schnarf (1931, 1933) found Opuntiacee definitely linked to these, per- haps in the neighborhood of Aizoacez, by embryological characters. The embryos are straight or nearly so in Polygonacee (in which the anatomy of the stem is anomalous, as it is also in Amaranthacee and Chenopodi- acee), Primulacee (notably similar to Caryophyllacee in gross features) and Plumbaginacee (distinguished by embryological peculiarities). In the middle of the Englerian system of the dicots, there is a long series of families, from Pandacee to Cynomoriacez, of which the majority are mere names to European and North American field botanists. Engler assembled many of these as two orders distinguished by the position of the ovules. In Geraniales, the ovule is epitropous, ‘“‘turned up,” either erect with the micropyle turned in or pendant with the micropyle turned out. In Sapindales, the ovule is apotropous, ‘“‘turned down,” either erect with the micropyle turned out or pendant with the micropyle turned in. The extensive study of woods principally of these families by Heimsch (1942) appears to have revealed a more natural grouping than the Eng- lerian: the next three orders represent it. 6 MADRONO [Vol. 14 Order 11. PoLYGALINAE (Brongniart as class) Braun in Ascherson FI. Brandenburg 1: 36 (1864). Order Polygalales (Bessey as suborder, 1897) Hallier (1905). Wood with rays hetrogeneous I or II A; parenchyma in apotracheal bands or scantily paratracheal; fiber-tracheids with conspicu- ously bordered pit-pairs. Humiriacee, Linacee, Erythroxylacee, Poly- galacee, Krameriacee, Diclidantheracee, Trigoniacee, Tremandracee, Zygophyllacee, Malpighiacee, Vochysiacez. Order 12. TRIHILATAE L. Order Terebinthine (Bartling) Braun. Order Terebinthales Wettstein. Pinnate Hutchinson. Rays heterogeneous II B or homogeneous; parenchyma banded or paratracheal; wood fibers with simple pits. Rutacee, Cneoracee, Simarubaceez, Meliacee, Sapindacee, Hippocastanacee, Aceracee, Bretschneideraceze, Connaracee, Burseracez, Terebinthacee. Julianiacee is to be reduced to Terebinthacez. The action of botanical congresses in conserving numerous names of families was scarcely duly considered, since practically all of these names are valid by the letter of the code. As an exception, the name Tere- bint (h) acez (Jussieu as order) appears to have been applied definitely to a family before Anacardiacee was. Order 13. GRUINALES L. Order Geraniales Engler. The wood (of woody examples) exnibits advanced characters, absence of scalariform perfora- tions and presence of libriform fibers. Geraniacez, Oxalidacee, Tropzola- cee, Balsaminacee. Limnanthacee, a small family of herbs of western North America, are in gross structure closely similar to Geraniacee. They are embryologically peculiar (having a 4- or 2-sporic embryo sac of unique type; Mason, 1951; Mathur, 1956), but may as well be placed here. Order 14. TrrcoccAaE L. Euphorbiacee, an enormous family, chiefly tropical, grossly varied in every character as though not a natural group. Order 15. COLUMNIFERAE L. Order Malvales Engler. Tiliacee, Mal- vacee, Bombacacez, Sterculiacez, and other families; a thoroughly nat- ural group. Order 16. CALYCANTHEMAE L. Orders Calyciflore and Hesperidee L., in part. Order Myriflore (Endlicher) Braun. Loasacee, Thymelzacee, Eleagnacee, Lythracee, Onagracee, Melastomacee, Myrtacee, and many others; needing further study. Order 17. UMBELLATAE L. Order Hederacee L., in part. Order Umbelli- flore (Bartling as class) Braun. Cornacee, Araliaceez, Umbelliferze. The suggestion that Garryacee also belongs here has been confirmed by a recent thorough study by Moseley and Beeks (1956). Order 18. SANTALINAE Grisebach. Like Calycanthemae and Umbellate, these have choripetalous flowers with inferior ovaries; the characters may be obscured by reduction. Olacacez, Santalacee, Loranthacee, Balano- phoracez, etc. Among sympetalous dicots, the Bicornes, Guiacane, and Primulales 1957] COPELAND: DICOTYLEDON SYSTEM 7 have already been given places. As to the remainder, the evidence, par- ticularly from the embryology, inclines one to treat them as a natural group of four orders, as follows. Order 19. LuripaE L., including orders Campanacee L. (Convolvulacee and Polemoniacee), Personate L. (Scrophulariacee, etc.), Aspertfolie L. (Hydrophyllacee and Boraginacee) and Verticillate L. (Labiate). Order Tubiflore (Bartling as class) Braun. About twenty-three families. In Bicornes and Guiacane, the cellular endosperm is inherited from cer- tain Guttiferz. In the present order it has evolved separately, the lower families, Convolvulacee and Polemoniacee, having nuclear endosperms. Whereas in Bicornes the endosperm mother cell undergoes two transverse divisions and produces a linear four-celled endosperm, in the present order it undergoes usually a transverse division followed by a longitudinal division, producing a T-shaped stage. The apetalous family Callitrichaceze has the embryogeny of this group. Order 20. ContTorTaE L., including Sepzarie L. Seven families, a very natural group, apparently a minor offshoot of the preceding. Order 21. STELLATAE L. Caprifoliacee and Rubiacee. The Adoxacee have the pollen of this group (Erdtmann, 1954). Order 22. AGGREGATAE L. Eight families, including Campanulacee, Lobeliacee, Valerianacee, Dipsacacee, and Composite. In the Englerian system, part of these are in Rubiales, but the embryological characters (and, indeed, the characters in general) place them as here. A large number of families remain unplaced. Places could be given to many of them by recognizing such orders as Juglandales, Aristolochiales, Sarraceniales, and Celastrales; but these are either small and themselves not certainly placed, or else not evidently natural. Surely, by sufficient study of anatomy, embryology, and palynology, we will eventually learn their true positions. At the same meetings and on the same day on which this paper was given oral presentation, Dr. Arthur Cronquist presented a system of the dicots which is expected soon to reach publication. Cronquist places all of the families in orders, which, for the sake of definition by description, are made considerably smaller and more numerous than the ones here main- tained. The orders are arranged in a phylogenetic pattern with which the one here presented is in essential agreement so far as it goes. The differ- ences between his system and mine are as though we were artists repre- senting the same tree under the conventions of different schools, and as though he had seen many more details than I (I have every reason to believe that he has perceived most of them correctly). The points of agree- ment allow us to believe that we are actually approaching the system of the future. Sacramento Junior College, Sacramento, California 8 MADRONO [Vol. 14 LITERATURE CITED ASCHERSON, P. 1864. Flora der Provinz Brandenburg der Altmark und des Herzog- thums Magdeburg. 3 parts. Berlin. Baitey, I. W., and C. G. Nasr. 1943-1945. The comparative morphology of the Winteracee. 7 parts. Jour. Arnold Arb. 24: 340-346, 472-481; 25: 97-103, 215-221, 342-348, 456-466; 26: 37-47. ; , and A. C. Smiru. 1943. The family Himantandracee. Jour. Arnold Arb. 24: 190-206. and A. C. SmirH. 1942. Degeneriacez, a new family of flowering plants from Fiji. Jour. Arnold Arb. 23: 356-365. and B. G. L. Swamy. 1949. The morphology and relationships of Austrobaileya. Jour. Arnold Arb. 30: 211-226. and . 1951. The conduplicate carpel of dicotyledons and its initial trends of specialization. Am. Jour. Bot. 38: 373-379. De Canpo_tieE, A. P. 1813. Théorie élémentaire de la botanique. . . . Paris. ENGLER, A. 1892. Syllabus der Vorlesungen iiber specielle und medicinisch-pharma- ceutische Botanik. Berlin. ERDTMANN, G. 1952. Pollen morphology and plant taxonomy. Angiosperms. Waltham, Mass. . 1954. Pollen morphology and plant taxonomy. Bot. Notiser 65-81. GUNDERSEN, A. 1950. Families of Dicotyledons. Waltham, Mass. Ha vier, H. 1905. Provisional scheme of the natural (phylogenetic) system of flow- ering plants. New Phytol. 4: 151-162. Hetmscu, C., Jr. 1942. Comparative anatomy of the secondary xylem in the “Gruin- ales” and “Terebinthales” of Wettstein with reference to taxonomic grouping. De Lilloa 8: 83-198. DE Jussieu, A. L. 1789. Genera plantarum secundum ordines naturales disposita. ... Paris. LINDLEY, J. 1829. A synopsis of the British flora. ... London. LINNAEUS, C. 1764. Genera plantarum. ... ed. 6. Stockholm. Mason, C. T., Jr. 1951. Development of the embryo sac in the genus Limnanthes. Am. Jour. Bot. 38: 17-22. Matuur, N. 1956. The embryology of Limnanthes. Phytomorphology 6: 41-51. Metcatr, C. R., and L. CHartx. 1950. Anatomy of the dicotyledons. . . . 2 vols. Oxford. MoseLeEy, M. F., Jr., and R. M. Beexs. 1956. Studies of the Garryacee — I. The comparative morphology and phylogeny. Phytomorphology 5: 314-346. ScHNARF, K. 1924. Bemerkungen zur Stellung der Gattung Saurauia im System. Sitz- ber. Akad. Wiss. Wien Math.-nat K]. Abt. 1, 133: 17-28. . 1931. Vergleichende Embryologie der Angiospermen. Berlin. . 1933. Die Bedeutung der embryologischen Forschung fiir das nattrliche System der Pflanzen. Biol. Gen. 9: 271-288. SmitH, A.C. 1943. The American species of Drimys. Jour. Arnold Arb. 24: 1-33. . 1945. A taxonomic review of Trochodendron and Tetracentron. Jour. Arnold Arb. 24: 119-164. . 1945. Geographic distribution of the Winteracee. Jour. Arnold Arb. 26: 48-59. . 1945. A taxonomic review of Trochodendron and Tetracentron. Jour. Arnold Arb. 26: 123-142. . 1946. A taxonomic review of Euptelea. Jour. Arnold Arb. 27: 175-185. . 1947. The families Illiciacee and Schisandracee. Sargentia 7: 1-244. SOLEREDER, H. 1908. Systematic anatomy of the dicotyledons. Translated by L. A. Boodle and F. E. Fritsch, revised by D. H. Scott. 2 vols. Oxford. Swamy, B. G. L. 1948. A contribution to the embryology of the Marcgraviacee. Am. Jour. Bot. 35: 628-633. 1957] LANGE: BOTANICAL DISASTER 9 . 1949. Further contributions to the morphology of the Degeneriacez. Jour. Arnold Arb. 30: 10-38. and I. W. Bartey. 1949. The morphology and relationships of Cercidiphyllum. Jour. Arnold Arb. 30: 187-210. Trprpo, O. 1938. Comparative anatomy of the Moracez and their presumed allies. Bot. Gaz. 100: 1-63. VesTaL, P. A. 1937. The significance of comparative anatomy in establishing the relationship of the Hypericacee to the Guttifere and their allies. Philipp. Jour. Sci. 64: 199-256. WobeHowsE, R. P. 1935. Pollen grains. ... New York and London. A BOTANICAL DISASTER ERWIN F. LANGE A chance meeting of two botanical explorers along the shores of the Columbia near The Dalles, Oregon, on November 13, 1843, left a story of disaster which completely altered the life of a German botanist, Fred- erick George Jacob Lueders. For on that day Lueders stood by helplessly as he watched the turbulent Columbia swallow up his entire botanical collection and collecting equipment. In a matter of seconds the results of three years’ labor in the wild and mountainous parts of the United States were washed away. All that Lueders was able to rescue from the water was a treasured copy of his Torrey and Gray Flora.! This event would probably have gone unrecorded in the pages of Northwest science history had it not been for the famed United States explorer, Captain John C. Fremont, who witnessed the event and noted it in his journal. Concerning the calamity he wrote: A gentleman named Lueders, a botanist from the city of Hamburg, arrived at the bay I have called by his name, while we were bringing up the boats. I was delighted to meet at such a place a man of kindred pursuits; but we had only the pleasure of a brief conversation, as his canoe, under the guidance of two Indians, was about to run the rapids; and I could not enjoy the satisfaction of regaling him with breakfast, which after his recent journey, would have been an extraordinary luxury. All his few instruments and baggage were in the canoe, and he hurried around to meet it at Grave Yard Bay; but he was scarcely out of sight when, by the carelessness of the Indians, the boat was drawn into the midst of the rapids, and glanced down the river, bottom up, with a loss of everything it contained. In the natural concern I felt for his misfortune, I gave to the little cove the name of Lueders’ Bay. Fremont’s note aroused but little interest until Leslie L. Haskins came to Brownsville, Oregon, as a photographer and botanist. As a small boy in 1 This book is today a part of the library collection of the Oregon Historical Society. Augusta Lueders, a daughter of the German botanist, sent it to Leslie L. Haskins, author of “Wild Flowers of the Pacific Coast,” who presented it to the Historical Society library. 10 MADRONO [Vol. 14 Sauk City, Wisconsin, around 1890 he had often seen and heard of an old German gardener and botanist, Frederick G. J. Lueders, a resident of that community. On coming to Oregon, Haskins was surprised to find that no one could give him information regarding the life and activities of the German botanical explorer. Only as the result of an intensive correspond- ence was Haskins able to uncover the story of Lueders’ life. The material was supplied by Miss Augusta Lueders, daughter of the botanist. Frederick Lueders was born in Hamburg, Germany, on October 3, 1813, the son of a gardener. In Hamburg he attended a private school and later studied botany at the Hamburg Botanical Gardens. Then he entered the large seed house of Haage in Erfurth, Saxony. On returning to Hamburg, the Society for Natural Science offered Lueders the opportunity for scien- tific exploration in the United States. To qualify for the opportunity, Lueders again attended school to learn navigation, and at graduation he qualified for first mate. Concerning the coming years of his life, Lueders wrote? a friend in 1861: I arrived in Sauk Prairie in July, 1841, in company with my friend ... We reached Green Bay by way of the lakes, and passed through the richly wooded country, which borders upon the Fox River and Lake Winnebago. Leaving the forest and entering the open country, we were much surprised at the beauty of the natural park. At that early time a few farms only guided the stage road. In almost every house where we stopped, the hospitable people invited us to spend a few days at their new home, and share what their humble plantation could afford, of course without pay. Arrived at Fort Winnebago, the terminous of the stage. The fortification was still garrisoned; there was besides a store, tavern and blacksmith’s shop near the fortress. From there we went down the Wisconsin River by a boat of the French fur trader. I spent the rest of the season about Sauk prairie in collecting several hundred species, in part, very interesting plants. Although I did not intend to spend the winter there, I was surprised by it before I could find conveyance to the Mississippi. In March, 1842, I went to Galena, And from there to St. Louis. There I found an easy introduction in my pursuit, as a Dr. Asa Gray, of New York, had kindly furnished me with a letter to Dr. Engelmann, whose services for the development of Western horticulture are amply known. After a short stay in the city I proceeded to search the western part of Missouri, collecting plants and curiosities. On my excursion to that part of Missouri, I found opportunity to gather information about the Western country, and resolved to pur- sue the next spring a westerly course as far as terra firma would permit me to study and collect the flora of the mountainous country. In the meantime there had awakened a spirit of emigration to Oregon, and large bodies of emigrants were along the frontier of Missouri forming several companies. One of these I joined leaving the civilized world in May, 1843. In the course of the journey, I collected plants and noted down peculiarities as circumstances would permit. The loss of my baggage in the rapids below the Grand Cascades of the Columbia River, rests not only severe with the collection of plants, but perhaps more so with a good many valuable instruments and other collecting material, as I had fitted myself out to spend several years in that part of the country. 2 Letter in herbarium, University of Oregon, Eugene. 1957] KEARNEY: PEEBLES 11 The kindest assistance was offered me by the gentlemanly officers of Fort Van- couver, but could not lead me into the course which my enthusiastic mind had marked out, and from there all communication by letter was tedious and uncertain. I concluded to return to Europe and engage anew, after having gained some useful experience. In February, 1844, I left the mouth of the Columbia River for the Sandwich Islands, and proceeded from there to Chile, touching the Paradise of the Pacific (Otaheite), then in a state of siege. I arrived in Hamburg in November of the same year. In the short space of my absence, family circumstances had taken a change, that made my presence there, at least for several years, necessary. So the course nearest my heart, for future life, was beyond my individual control. At Christmas, 1844, I again hailed the Mississippi. I lived at St. Louis until 1851, and after that time in Sauk County (Wisconsin) tilling the soil and my mind. — F. G. J. Lueders He spent the rest of his life in Sauk City, Wisconsin, and engaged in many astronomical studies. In 1869 he had published in Hamburg “‘The Aurora Borealis and Law of Reciprocal Action in the Universe” and in 1884 the University of Wisconsin published his observations on a number of Auroras which he had studied. Privately, Lueders also published a pamphlet ‘‘“Memoirs on Physical Astronomy.” For many years he was city treasurer of Sauk City. He died December 21, 1904. Thus a botanical disaster prevented Lueders’ name from being associ- ated with Northwest botany. Had his specimens been properly reported, there is no doubt that his name would have been linked with species of western plants. Lueders’ Bay, named by Fremont, is also a lost geograph- ical name. One can only speculate on its location, and as the large dams on the Columbia are changing the topography along the river, Lueders’ Bay itself may no longer be in existence. Portland State College, Portland, Oregon ROBERT HIBBS PEEBLES! The death, last March, of Robert Hibbs Peebles, one of the world’s leading cotton breeders and an outstanding student of cotton genetics, was a great loss to southwestern agriculture. He was the originator of what had become, in recent years, the preferred variety of Pima (American-Egyp- tian) long-staple cotton, and carried on investigations of the inheritance of various characters of this very important crop-plant. Toward the end of his career he was working on the problem of how to insure the greatest possible degree of natural cross-pollination, in view of the fact that arti- ficial cross-pollination of cotton varieties usually increases the yield, as compared with that of either parent. The endeavor was to do, with cotton, what has been done so successfully with hybrid corn. 1 This paper was in press at the time of Dr. Kearney’s death on October 19, 1956. —Ep. 12 MADRONO [Vol. 14 Fic. 1. Robert H. Peebles at his home, Sacaton, Arizona, Christmas Eve, 1944. The University of Arizona, recognizing the great value of Peebles’ con- tributions to the agriculture of the state, conferred upon him an honorary D. Sc., less than a year before his untimely death, at the age of 55. He was also given, posthumously, the United States Department of Agriculture’s Superior Service award and medal. To taxonomic botanists, Dr. Peebles was well-known as an ardent stu- dent of the native flora and co-author with me of ‘Flowering Plants and Ferns of Arizona” (1942) and “Arizona Flora” (1951). For two decades he devoted many of his week-ends to plant collecting in all parts of the state and added substantially to the number of species known to occur there. He was especially interested in the Cactacez, which are such a conspicu- ous element of the Arizona flora, and he became a recognized authority on that family. The technique which he developed for preparing herbarium specimens of cacti has never been excelled. Some readers of this notice will recall his exhibition of specimens and photographs at a meeting of the California Botanical Society in April, 1953. His name is commemorated in that of the remarkable and very rare little cactus, Navajoa Peeblesiana Croizat. 1957] TURNER: LUPINUS 13 The many friends of Bob Peebles will remember him always for his vivid and lovable personality. He was so very much alive that we can scarcely realize, even yet, that he is no longer with us. He has left a void that will be very hard to fill—THomas H. Kearney, California Academy of Sciences, San Francisco. THE CHROMOSOMAL AND DISTRIBUTIONAL RELATIONSHIPS OF LUPINUS TEXENSIS AND L. SUBCARNOSUS (LEGUMINOSAE) B. L. TURNER The genus Lupinus is represented in Texas by several species (Shin- ners, 1953). Of these, the two most commonly encountered are L. texensis Hook. and L. subcarnosus Hook. The latter taxon is the official state flower of Texas, though L. texensis is sometimes mistaken for this species. Both species are endemic to the state and are known locally as bluebon- nets. They are probably the most important native rangeland legumes in central Texas, often occupying hundreds of acres of rolling hillsides dur- ing the early spring months. The roots of these species are highly nodu- lated and are undoubtedly important soil nitrifiers. In addition, L. tex- ensis has become a popular garden ornamental in many parts of the world. (Although many trade catalogues list L. subcarnosus as the Texas bluebonnet, most of the material on the open market appears to be L. texensis.) GEOGRAPHICAL DISTRIBUTION Lupinus texensis occurs naturally on open calcareous soils throughout central Texas. Lupinus subcarnosus is restricted to sandy soils of south- central Texas. The interfingered distribution of the two species (Fig. 1) can be related to alternating grassland — forest strips which occur on deep clay and sandy soils respectively. The ecotone between these vege- tative types is sharp, and consequently both species may be found growing in close proximity along many miles of the contact area. Lupinus texensis has a wide ecologic amplitude and may grow in a variety of disturbed soil types. As a result, the species has become established along road shoul- ders which cross the otherwise unoccupied sandy lands, particularly as a result of deliberate sowing by state highway workers and other wild- flower enthusiasts. Lupinus subcarnosus is rarely if at all sown along high- ways, and in no instance has the author seen the plant growing naturally on clay soils or along highways in such areas. In the numerous cases where both species were found growing together during the spring of 1955, no sign of morphologic intergradation, meiotic irregularity, or other evidence of hybridization could be detected. 14 MADRONO [Vol. 14 4th _ ae. a md SS 7 we a ea a Fe ay Pe a aes : Al — ftp po tj . r | ) se eee —-, > ' [ 73s ¢ = LUPINUS TEXENSIS lee (sae | aS ee LUPINUS SUBCARNOSUS i oe \ Fic. 1. Probable natural distribution of Lupinus texensis and L. subcarnosus. Based on herbarium records at The University of Texas and extensive field observa- tion. Further explanation in text. CHROMOSOME NUMBERS Previous to the present study, two different counts had been reported for L. subcarnosus. Savchenko (1935) reported 2n = 48 and Tuschnja- kowa (1935) reported 2n = 36 for this species. Because of the past con- fusion in the application of the names L. texensis and L. subcarnosus (Shinners, 1953), it was at first thought that the two differing counts might be for both species instead of L. subcarnosus alone. As a result, meiotic studies! of natural populations of these two taxa were undertaken. However, it was soon discovered that both L. texensis and L. subcarnosus had the same chromosome number of n = 18. In all instances, meiosis was completely regular, metaphase plates showing 18 bivalents and anaphase plates were without bridges. Counts obtained are given in Table 1. 1 Buds were killed and fixed in a mixture of 4 chloroform:3 absolute alcohol: 1 glacial acetic acid. Anthers were squashed in acetocarmine 2 to 3 days after collec- tion. Voucher specimens are deposited at the University of Texas Herbarium, Austin, Texas. 1957] TURNER: LUPINUS 15 TABLE 1. CHROMOSOME COUNTS OF LUPINUS SUBCARNOSUS AND L. TEXENSIS Species Collection n number L. subcarnosus. Bastrop County: Bastrop State Park. Turner 3703. 18 L. subcarnosus Bastrop County: 4 miles west of Bastrop. Turner 3704. 18 L. subcarnosus Gonzales County: near Palmetto State Park entrance. Turner 3708. 18 L. subcarnosus Fayette County: 2 miles west of Moulton. Turner 3712. 18 L. subcarnosus Lavaca County: Sublime. Turner 3719. 18 L. subcarnosus Colorado County: 5 miles west of Altair. Turner 3723. 18 L. subcarnosus Fort Bend County: 0.5 mile east of Fulshear. Turner 3727. 18 L. subcarnosus Austin County: San Felipe State Park. Turner 3730. 18 L. texensts Travis County: Austin. Turner 3699. 18 L. texensis Lavaca County: 2 miles west of Moulton. Turner 3713. 18 L. texensis Lavaca County: 1 mile southeast of Shiner. Turner 3718. 18 L. texensis Austin County: 3 miles east of Ulm. Turner 3732. 18 L. texensis Hays County: 10 miles west of San Marcos. Turner 3733. 18 L. texensis Llano County: 3 miles northwest of Buchanan Dam. Turner and Johnston 2523. 18 Savchenko’s number of 2n = 48 was apparently for some misnamed taxon, or else strains of L. subcarnosus and/or L. texensis exist in the ornamental trade as derived polyploids. Savchenko did not cite voucher material but merely indicated that the counts were made from seeds ob- tained from Germany. 2 3 3 Fics. 2-3. Metaphase chromosomes of Lupinus texensis and L. subcarnosus: 2, L. texensis, Nn = 18; 3, L. subcarnosus, n = 18. Camera lucida drawings, 2000. 16 MADRONO [Vol. 14 DISCUSSION Lupinus texensis and L. subcarnosus are apparently very closely related as shown by their external morphological characters and their similar chromosome complements. However, they are clearly separated ecologi- cally and in the field they are reproductively isolated. The reproductive isolation is perhaps partially due to the self-pollinating nature of the breeding populations; naturally occurring cross-pollinated individuals are probably rare. Experimental hybridization between these two spe- cies is being undertaken. The discovery that both L. texensis and L. subcarnosus have chromo- some numbers of n = 18 has certain phyletic implications. Senn (1938), on the basis of Tuschnjakowa’s reported number for L. subcarnosus, con- sidered the species to be triploid in origin and thus, along with 2n counts of 48 in other species, concluded the base number for the genus to be x — 12 instead of 8, 9, 10, etc., as has been indicated by other workers (Darlington and Janaki-Ammal, 1945). Senn considered species with n numbers of 20, 21, 25, etc. to be derived aneuploids. The only other num- ber of n= 18 reported for the genus Lupinus is that made by Eickhorn (1949) on L. tassilicus Maire. SUMMARY The distributional relationship of L. texensis and L. subcarnosus has been indicated. The former species is widespread throughout central Texas, occurring in calcareous soils; the latter is more restricted in range, occurring on sandy soils of south-central Texas. Meiotic counts from a number of localities in central Texas showed the chromosome number of both species to be n = 18. An earlier report of 2n = 48 for L. subcarnosus was probably erroneous. In spite of the morphological and chromosomal similarities of the two species, they do not hybridize in nature, even in habitats which permit their side-by-side occurrence. The Plant Research Institute, University of Texas, Austin, and The Clayton Foundation for Research. LITERATURE CITED DaruincTon, C. D. and E. K. JANAKI-AMMAL. 1945. Chromosome Atlas of Cultivated Plants. London. 397 pp. ErckHorn, A. 1949. A propos de la caryologie du Lupinus tassilicus Maire. Rev. Cytol. et Biol. Veg. 11: 333-350. SAVCHENKO, P. F. 1935. Karyology of some species of the genus Lupinus. Bull. Ap- plied Botany, of Genetics and Plant Breeding. Series II, No. 8: 105-112. (In Russian; English summary on pp. 197-198 of same journal.) Senn, H. A. 1938. Chromosome number relationships in the Leguminosae. Bibliog- raphie Genetica 12: 175-336. SHINNERS, L. H. 1953. The bluebonnets (Lupinus) of Texas. Field and Lab. 21: 149-153. Tuscunyakowa, M. 1935. Uber die Chromosomen einiger Lupinus-Arten. Der Ziich- ter 7: 169-174. 1957] ERBE: LUPINUS il STUDIES ON THE CROSSABILITY OF LUPINUS TEXENSIS AND LUPINUS SUBCARNOSUS LAWRENCE ERBE During the period from March 17 to April 7, 1956, attempts were made to hybridize Lupinus subcarnosus with L. texensis. The work was begun just after L. subcarnosus began flowering. This was about two weeks later than L. texensis under the experimental conditions. The plants were most vigorous at this time as they had recovered from the shock of trans- planting but had not yet started to decline in vigor. However, this vigor- ous condition could not be maintained. This factor doubtlessly was in part responsible for the low percentage of pods set from the various crosses and selfs. This is particularly true of the intraspecific crosses and selfings attempted on L. texensis, since these plants were the last to be worked upon and by that time they were no longer in optimum health. The same emasculation and pollination techniques were employed as had previously proved successful on hybridization studies on the genus Lotus (Erbe, Master’s Thesis, University of Vermont, 1955). Stock plants of L. subcarnosus were obtained from a population occur- ring on sandy soil about ten miles west of Bastrop, Texas. Plants of L. texensis were obtained from a population occurring on the University of Texas campus in black clay. On the basis of the present study, the tentative conclusion was that the two species are effectively isolated genetically. Certainly, they do not hybridize readily when subjected to experimental emasculation and pol- lination techniques. These experimental results as shown in Table 1 agree with the observation, presented by Turner (Madrono 14, p. 16), that there is no evidence of hybridization occurring when the two species grow sympatrically. TABLE 1. RESULTS OF ATTEMPTED CROSSES IN LUPINUS No. of florets No.of pods Per cent of pods Interspecific crosses: L. texensis & L. subcarnosus 52 0 0 L. subcarnosus & L. texensis 25 ) 0 Intraspecific crosses: L. subcarnosus X L. subcarnosus 11 3 14 L. texensis K L. texensis 10 0 0 Selfed: L. subcarnosus 45 18 40 L. texensis 44 6 14 During the course of the study the writer bagged several inflorescences that were not used in the hybridization studies. Not a single floret of any of these inflorescences produced a pod. In addition, only one pod devel- oped on the unbagged inflorescences of approximately fifteen other plants of L. texensis. Bee activity was almost non-existent; only one bee was 18 MADRONO [Vol. 14 observed ‘‘working” an inflorescence. That particular inflorescence sub- sequently produced the only pod that developed on a L. texensis plant without experimental manipulation. It appears evident that pods are not produced by florets of LZ. texensis unless they have been “worked” by bees or man. Several of the unbagged inflorescences of L. subcarnosus set a few pods. The Plant Research Institute, University of Texas, Austin, and The Clayton Foundation for Research, NEW NORTH AMERICAN ANDROPOGONS OF SUBGENUS AMPHILOPHIS AND A KEY TO THOSE SPECIES OCCURRING IN THE UNITED STATES* FRANK W. GouULD The Old and New World andropogons of the subgenus Amphilo phis comprise a relatively distinct group, recognized as a separate genus by many systematists including O. Stapf, C. E. Hubbard, A. Camus, J. T. Henrard, and S. T. Blake. Both Amphilophis Nash and Bothriochloa Kuntze have been proposed as generic names for the species comprising this subgenus, with Bothriochloa (1891) antedating Amphilophis (1901). The name Amphilophis was first used by Trinius as a section name under Andropogon. Included in the section were a number of species belonging to Vetiveria, Sorghastrum, and Sorghum, as well as Andropogon saccha- roides and its allies. Hackel, in his monograph (1889), took up Amphilo- phis as a subgenus name for the A. saccharoides group. Bothriochloa never has been officially published as a subgeneric name. The Amphilophis andropogons are distinguished primarily on the basis of inflorescence characters. The pedicels, and at least the terminal rachis joints, have thickened margins and a medial groove or a broad thin mem- branous central area. The inflorescence characteristically is a leafless ter- minal panicle, with several to numerous racemose branches. In a few species there are as few as two or three branches per inflorescence. The following new species and new name combinations are proposed in Andropogon rather than in Bothriochloa primarily to conform with the standard United States treatments of the genus (Hitchcock, 1951; Swal- len, 1951; Gould, 1951; Gleason, 1952; Harrington, 1954). The Latin diagnoses have been kindly supplied by Dr. Lloyd Shinners of Southern Methodist University. Mr. Jason R. Swallen of the United States Na- tional Museum has aided in clarifying the relationships of generic and subgeneric names. The writer is indebted to the curators of the herbaria of 1 Technical Bulletin No. TA 2384, Texas Agricultural Experiment Station, Col- lege Station, Texas. 1957] GOULD: ANDROPOGONS 19 the following institutions for the loan of specimens during the course of study: University of Arizona (ARIZ), University of California (UC), University of Michigan (MICH), University of Texas (TEX), Southern Methodist University (SMU), and Smithsonian Institution (US). Speci- mens of the Tracy Herbarium, A. and M. College of Texas, are cited as (TRACY ). Collections of the writer are indicated by number alone. Nurs- ery grown plants were made available through the facilities and seed intro- duction program of the Texas Agricultural Experiment Station, and the Plant Introduction Section of the United States Agricultural Research Service. Andropogon spring fieldii sp. nov. Perennis cespitosa 30-80 cm. alta; nodi dense barbati pilis patentibus 3-7 mm. longis; foliorum laminz 2—3 (—5) mm. late glabrz vel hispid# prope ligulam bicrinate; inflorescentia dense villosa pilis 5-10 mm. longis, ramosa ramis 2—8 in axi 1—4 cm. longo, inferioribus raro ramulosis; racheos segmenta et pedicelli subzquales sulcati marginibus crassulis villosis; spicula sessilis 6.0—-7.3 mm. longa, gluma inferiore acuta vel bifidula infra mediam pilosa interdum supra mediam glanduloso-punctulata, lemmatis arista 20-26 mm. longa; spicula pedicellata sterilis angusta 5 mm. longa, pedicello 1-2 mm. longior; mei- osis pollinis regularis; chromosomata somatica 120. Tufted perennial with culms 30—80 cm. tall; culm nodes densely bearded with spreading hairs, these usually 3-7 mm. long; leaf blades 2—3, occa- sionally to 5 mm. broad, glabrous or sparsely pustulate-hispid on the axial surface and with tufts of hair on each side of the ligule; panicle densely white-villous, with 2—8 racemose branches, these infrequently rebranched ; rachis joints and pedicels about equal, with thickened, densely villous margins and a broad thin membranous central area; hairs of the inflor- escence 5—10 mm. long; sessile spikelets mostly 5.5—7.3 (—8.5) mm. long, the first glume acute or minutely bifid at the apex, hairy on the lower third or half of the dorsal surface, occasionally with a faint glandular pit or depression above the middle; awn of the lemma 20-26 mm. long; pedi- celled spikelet sterile, narrow, averaging 4-5 mm. long and 1-2 mm. longer than the pedicel; pollen meiosis regular; chromosome number 2n== 120. Type: From plant grown in nursery of Texas Agricultural Experiment Station, College Station, Texas, F. W. Gould 6642 (type, TRACY; iso- types, US, UC, TEX, SMU, ARIZ). Original seed from near Scholle, Socorro County, New Mexico (Wayne Spring field, 20 August 1950). Distribution: Western Texas, New Mexico, and northern Arizona; Ar- gentina. Specimens examined: Unitrep States. Texas. Andrews County: Shafter Lake, Tharp et al. 43024 (TEX). Brewster County: Alpine, Bailey 29931 (TRACY) ; Chisos Mts., Lundell 13286 (UC); Glass Mts., Warnock W370 (TRACY) Culber- son County; southeast of Van Horn, Davis et al. 90 (TRACY); Guadalupe Mts., Lee 67 (TEX); base of El Capitan, Tharp and Gimbrede 51-1543 (TRACY) ; Pine Spring, Young in 1916 (TEX). Dawson County: Texas Soil Survey in 1922 20 MADRONO [Vol. 14 (TRACY). Jeff Davis County: 30 miles west of Fort Davis, Reeves and Morrow G-165 (P.1. 216668) (TRACY) ; northern part of county, Burnett 9 (TEX). Presidio County: Taylor in 1941 (TEX). NEw Mexico. Chaves County: Roswell, Hinckley in 1936 (MICH). Lincoln County: Carrizozo, Grassl 239 (MICH). Sierra County: Cuchillo, Archer 416 (MICH). Socorro County: Scholle, seed collected by Spring- field in 1950. Type collection from plants grown at College Station, Texas, Gould 6642 (TRACY, US, UC, TEX, SMU, ARIZ). Valencia County: Paguate, Weather- wax 2763 (TRACY). Without locality, Wright 2103 (TRACY). Arizona. Coconino County: Havasupai Canyon, Clover 7121 (MICH). ArcEeNnTINA. Buenos Aires: west of Argerich, Eyerdam et al. 23510 (UC). This plant most commonly has been identified as Andropogon bar- binodis Lag., from which it differs typically in the smaller habit, narrower blades, longer nodal hairs, panicles with fewer branches, shorter axis, and more densely white-villous pedicels and rachis joints, and the lower chro- mosome number. The genetical relationship of this species with other taxa of the Am- philophis section is obscure. The short panicle axis, few panicle branches, large spikelets, and occasional pitted glume of the sessile spikelet suggest relationship with Andropogon edwardsianus Gould. The latter has deeply pitted glumes and the minimum chromosome number of the New World taxa, 2n 60; The dense villous pubescence of pedicels and rachis joints of A. spring- fieldit is not equalled in any other North American species of the Amphilo- phis section. It is, however, very similar to the condition characteristic of South American plants referable to A. saccharoides var. ertanthoides Hack. The latter, undoubtedly specifically distinct from A. saccharoides, also has a chromosome number of 2n=60. Both A. saccharoides var. erian- thoides and A. edwardsianus are known to occur in Uruguay. Andropogon spring fieldii is known to the writer from Argentina and undoubtedly also is present in Uruguay. Andropogon reevesii sp. nov. Perennis cespitosa 30-80 cm. alta gen- iculata demum ramosa; nodi glabri vel puberuli; folia glauca scabra firma; vagina glabra; ligula membranacea 2—4 mm. longa; lamina 2—4 mm. lata longe acuminata glabra vel supra parce pilosa; inflorescentia 6—8 cm. longa subflabellata sat pilosa ramis 6—9 subzequalibus 3—6 cm. longis simplicibus vel inferioribus ramulosis in axi 1.0-3.5 cm. longo; racheos segmenta et pedicelli 12-34 spiculas sessiles equantes cum sulca media membranacea marginis crassulos ciliatos equante; spicula sessilis 4-5 mm. longa glauca late acuta, gluma inferiore plerumque infra mediam parce pilosa superne marginibus scabris vel puberulis non pertusa; lemmatis fertilis arista 12-15 mm. longa; spicula pedicellata sterilis angusta 3.0-4.5 mm. longa pedicello longior; meiosis pollinis regularis; chromosomata somatica 120. Perennial bunchgrass; culms mostly 30-80 cm. tall, geniculate and freely branching below in age; culm nodes puberulent to glabrous; leaves glaucous, scabrous, very firm in texture; sheaths glabrous; ligule mem- branous, 2—4 mm. long; blades mostly 2-4 mm. broad, relatively narrow and stiff, tapering to a long-acuminate tip, glabrous or with a few scattered 1957] GOULD: ANDROPOGONS 21 hairs on the axial surface; panicles 6-8 mm. long, somewhat flabellate, moderately hairy, with an axis 1.0—3.5 cm. long and usually 6—9 branches, these unbranched or the lowermost simply rebranched; panicle branches mostly 3-6 cm. long, the uppermost about as long as the lower; rachis joints and pedicels 12 to 34 as long as the sessile spikelets, with a mem- branous medial groove about as wide as the thickened ciliate margins; sessile spikelets mostly 4-5 mm. long, glaucous, broadly acute at the apex, the first glume usually with a few coarse hairs below the middle and scab- rous or minutely puberulent on the margins near the apex, pitless; awn of fertile lemma mostly 12-15 mm. long; pedicellate spikelets sterile, nar- row, mostly 3.0-4.5 mm. long, longer than the supporting pedicels; pollen meiosis regular; chromosome number 2n—120. Type: Collected 2 August 1954 from plant grown in nursery of the Texas Agricultural Experiment Station, College Station, Texas, F. W. Gould 6647 (type, TRACY; isotypes, US, UC, MICH, TEX). Original seed from Arteago, about 15 miles east of Saltillo, Coahuila, Mexico (R.G. Reeves and Judd Morrow G-640, altitude 6,000 feet, October 15, OS SF Pal 216183). Distribution: Known only from the Arteago collection. This plant is similar to Andropogon wrightu Hack. in general aspect but also appears close to A. saccharoides Swartz. From the former it dif- fers in the stiff blades, the smaller, more reduced pedicelled spikelets, the smaller, consistently non-pitted sessile spikelets, and in the uniformly regular pollen meiosis. From the latter it differs in the narrow stiff blades, and the inflorescence with a relatively short axis, few branches, and long terminal branches. Andropogon reevesii has a chromosome number of 2n=120, while typical A. saccharoides has 2n=60 chromosomes. Andro- pogon saccharoides var. longipaniculata Gould has 2n=120 chromosomes but has a much larger and longer panicle, broader blades, and is a larger plant in general. Andropogon hybridus sp. nov. Perennis cespitosa erecta 30-80 cm. alta; nodi glabri vel puberuli; foliorum vagine virides glauce glabre; lamine 2—4 (—5) mm. late plerumque basin versus parce cilite szpe in ambitu parce pilose; panicula 6—11 cm. longa non dense pilosa ramis 3-8 simplicibus vel inferioribus ramulosis in axi 0.6-3.5 (—4.5) cm. longo; racheos segmenta et pedicelli subaquales sulcati marginibus crassulis pilosis pilis apicem versus 5—7 mm. longis basin versus multo brevioribus; spicula sessilis 4.5-6.5 longa aristis 18-25 mm. longis, gluma inferior nitida luteo-viridi apicem versus quinquenervosa supra mediam glandu- loso-pertusa, infra mediam plerumque parce pilosa; spicula pedicellata diminuta sterilis 2.2-3.6 mm. longa pedicello brevior; meiosis pollinis regularis; chromosomata somatica 120. Perennial, with strictly erect culms in small to medium sized clumps; culms 30-80 cm. tall, moderately branched and leafy above the base; culm nodes glabrous or minutely puberulent; leaf sheaths green or glaucous, 22 MADRONO [Vol. 14 glabrous; blades mostly 2—4 rarely —5 mm. broad, usually sparsely ciliate with long hairs near the base, often with a few hairs on the surfaces; panicles hairy but not densely so, 6-11 cm. long, usually with 3—8 primary branches on an axis 0.6—3.5, occasionally —4.5 cm. long, the lower branches often simply rebranched; rachis joints and pedicels about equal, with a broad, membranous, often dark-colored, central groove and thickened hairy margins, the hairs mostly 5—7 mm. long near the apex and much shorter towards the base; sessils spikelet 4.5-6.5 mm. long, with an awn 18-25 mm. long; first glume of sessile spikelet shiny, yellowish-green, with usually five greenish nerves apparent on the upper half, with a moderately deep glandular pit above the middle, and usually with a few stiff hairs on the lower one-third or one-half of the back; pedicelled spikelets highly reduced, sterile, mostly 2.2—3.6 mm. long and shorter than the supporting pedicel; pollen meiosis regular; chromosome number 2n=120. Type: Texas, La Salle County, two miles east of Cotulla, F. W. Gould 6978, 10 November 1955 (type, TRACY; isotypes, US, UC, TEX, MICH, SMU, ARIZ). The dominant grass along a low flat graded road- side ditch in a mesquite area, growing with Pappophorum bicolor and weedy forbs in gravelly red-brown clay. Type and isotypes from one plant. Distribution: South-central Texas to northeastern Mexico. Specimens examined: TExas. Atascosa County: Jourdanton, 6223 (TRACY) ; Pleasanton, 6283 (TRACY). Bee County: 6.6 miles west of Beeville, 6051 (TRACY). Bexar County: San Antonio, Higdon in 1936 (TEX). Burnet County: Marble Falls, 5959 (TRACY). Cameron County: near San Benito, Faulkner 94 (TRACY). Dimmit County: Asherton, 6004 (TRACY). Gillespie County: 8 miles east of Fredericksburg, 5354 (TRACY); 11 miles east of Fredericksburg, 6487 (TRACY). Gonzales County: 8.5 miles south of Smiley, 6066 (TRACY); 8.5 miles south of Smiley, 6067 (TRACY). Guadalupe County: 19 miles west of Sequin, 6939 (TRACY). Kerr County: Kerrville, 6484 (TRACY). Kleberg County: 11 miles west of Kingsville, 6043 (TRACY); Kingsville, 6034 (TRACY). La Salle County: Cotulla, 6978 (TRACY). Live Oak County: George West, 6047 (TRACY) ; Cum- mins 12 (TRACY). Matagorda County: Palacios, Richmon 26 (TRACY). Mav- erick County: 10.5 miles south Quemando, 5997 (TRACY); Eagle Pass, 6473 (TRACY); west of La Pryor, near county line, 6475 (TRACY). Nueces County: western Nueces County, Tharp 47428 (TRACY). Real County: Camp Wood, 6952 (TRACY); Leaky, 6482 (TRACY). Travis County: 20 miles northwest of Oak Hill, 5961 (TRACY). Uvalde County: Uvalde, 6479 (TRACY); Uvalde 6226b (TRACY). Val Verde County: 30 miles west of Del Rio, Rose 71 (TRACY). Webb County: Tharp 5255 (TEX). Mexico. CoAHulItra. Sabinas, 6477. This plant is intermediate between Andropogon edwardsianus Gould and A. barbinodis Lag. both in morphological characteristics and in chro- mosome number. From A. edwardsianus it differs in having broader blades, better developed upper culm leaves, more branched culms whose nodes often are puberulent, glume of sessile spikelet hairy below, panicle usually larger and more branched, and a 2n chromosome number of 120 rather than 60. From A. barbinodis it differs typically in the shorter, more erect, less branched culms, glabrous or puberulent nodes, narrower leaf blades, smaller panicles with fewer branches, these often all unbranched, shorter pedicellate spikelets, less hairy glume of sessile spikelets, and fewer chro- mosomes. In relatively few areas do plants of A. barbinodis have all sessile 1957] GOULD: ANDROPOGONS 23 spikelets pitted while in A. hybridus the sessile spikelets consistently are pitted. Andropogon hybridus characteristically is a plant of moderately dis- turbed habitats. It is most frequent along low roadsides and fence-rows, often forming dense stands. Throughout its range it is consistently asso- ciated with A. barbinodis and either A. saccharoides var. torreyanus or A. saccharoides var. longipaniculata.v $ Inflorescences of plants assumed to be hybrids between A. hybridus (2n=120) and A. saccharoides var. longipaniculata (2n=120) have been collected at two Texas localities (Bee County, 6.6 miles west of Beeville, 6050a; Uvalde County, 1.5 miles north of Uvalde, 6226c). No indications of hybridization between A. hybridus and A. saccharoides var. torreyanus or A. barbinodis have been observed. It is very possible, however, that the first two taxa, with n=60 and n=30 chromosomes respectively, have produced fertile allopolyploids referable to A. barbinodis (n=90). Andropogon hybridus appears to have arisen from one or more hybrids between A. edwardsianus (n=30) and A. saccharoides var. torreyanus. Hybridization, followed by doubling of chromosome number, could well have produced this fully fertile species. Andropogon hybridus seems almost certainly to be a relatively “young” species whose success is correlated with man’s occupation and development of the region in which it occurs. Despite its present abundance along roads and railroad rights-of-way, the favorite collecting sites of taxonomists, this grass is poorly represented in herbaria. All but one of the collections studied and cited in this paper were made in the last 20 years. A complete search has not been made of the large herbaria for early collections, but these are certain to be few. Andropogon palmeri (Hack.) comb. nov. Andropogon saccharoides Swartz subsp. leucopogon var. palmeri Hack., in DC. Monogr. Phan. 6:496. 1889. Amphilophis palmerit Nash, Fl. N. Amer. 17: 126. 1912. Type: Palmer 305 ‘Mexico ad Rio Blanco.” 1886. Specimens examined: Specimens from plants grown at College Station, Texas, from the following seed collections. Mrxico. DURANGO: between Torreon and Durango City, Morrow and Merrill G705 (P.I. 216186), between Zacatecas and Durango City, Morrow and Merrill G736 (P.I. 216196). Identification of this ma- terial was made by Jason R. Swallen of the United States National Museum. Andropogon palmeri is similar to A. barbinodis in growth habit and in- florescence characteristics. It differs from this species primarily in having densely villous blades and sheaths. The first glume of the sessile spikelet may or may not be glandular-pitted. Although no accurate chromosome count has been obtained, it is most likely that this species has 2n=180 chromosomes, the same as 4. barbinodis. ANDROPOGON BARBINODIS Lag. var. perforatus (Trin.) comb. nov. Andropogon perforatus Trin. ex Fourn. Mex. Pl. 2:59. 1886. A. sac- charoides subsp. leucopogon var. perforatus Hack., in DC. Monogr. Phan. 6:496. 1889. Amphilophis perforatus Nash, in Small, Fl. Southeast U. S. 24 MADRONO [Vol. 14 66. 1903. Bothriochloa perforata Herter, Rev. Sudamer. Bot. 6:135. 1940. Type: Berlandier 641 “Envir. de Mexique.”’ Distribution: South-central U. S. and Mexico, Argentina, and Uruguay. Specimens examined: UNITED StaTEs. TEXAS. Aransas County: Rockport, Cates in 1946 (TRACY). Bee County: west of Beeville, 6049 (TRACY). Bell County: near Little River, Wolff 2260 (TRACY). Hidalgo County: 12 miles north of Edin- burg, 6024 (TRACY). Blanco County: west of Johnson City, 5963 (TRACY). Bosque County: 12 miles northeast of Walnut Springs, Shinners 10420 (SMU). Brooks County: 19 miles north of San Manuel, 6029 (TRACY). Brown County: Brownwood, 5681 (TRACY). Brewster County: Chisos Mts., Sperry 396 (TRACY). Caldwell County: Luling, 6938 (TRACY). Dimmit County: north of Carrizo Springs, McCully 32 (TRACY). Edwards County: Texas Agr. Exp. Station, Sub- station 14, Cory 52473 (SMU, UC), Sperry in 1947 (TRACY). Gillespie County: Fredericksburg, 5965 (TRACY) ; east of Seguin, 6069, 6070 (TRACY). Guadalupe County: 19 miles west of Seguin, 6942 (TRACY). Hays County: Kyle, Tharp in 1920 (MICH). Llano County: Llano, Wolff 3047 (TRACY); south of Valley Spring, Smathers 14 (TRACY); Enchanted Rock, Tharp 7699 (TEX). Maverick County: 16 miles east of Eagle Pass, 6973a (TRACY). McCulloch County: Lohn, Whitaker 50-16 (TRACY). McLenn@n County: east Waco, Smith 306 (TEX). Mitchell County: northeast of Colorado City, Pohl 4987 (SMU). Motley County: west of Matador, Duvall 52-102 (RF 556) (TRACY). Pecos County: Fort Stockton, Cory in 1924 (TRACY). Palo Pinto County: west of Mineral Springs, Whitehouse 19296 (MICH, SMU). Somervell County: north of Glen Rose, Evans in 1951 (TRACY). Tarrant County: Ruth 1065 (MICH). Uvalde County: Montell, 6951 (TRACY); Sabinal, 6945 (TRACY). Young County: Graham, Reverchon 3439 (SMU). Arizona. Cochise County: Douglas, Gould and Haskell 4548, in part (SMU). Pima County: Santa Rita Range Reserve, Culley 58 (ARIZ). Mexico. CHIHUAHUA. Agua Caliente, LeSueur mex 051, in part (SMU, UC) ; between Camargo and Parral, Reeves and Morrow G-493 (P.I. 216165), in part (TRACY); north of Chihuahua City, Reeves and Morrow G-362 (P.I. 216157) (TRACY). Coanuira. 40 km. west of El Oro, Harvey 1275, in part (MICH); 10 miles north of Mondora, Reeves and Morrow G-328 (P.1. 216121) (TRACY). Duranco. North of Durango City, Reeves and Morrow G-524 (P.I. 216122) (TRACY); between Durango City and Torreon, Morrow and Merrill G-841 (P.1. 216096) (TRACY); between Dur- ango City and Mazatlan, Morrow and Merrill G-789 (P.I. 216088) (TRACY) ; between Durango City and Parral, Morrow and Merrill G-760 (P.I. 216080) (TRACY); Ignacio Allende, Gentry 6917 (ARIZ). Hipatco. Ixtaccihuatl, Purpus in 1905 (UC); Pachuca, Purpus 1631 (UC, MICH); north of Zimapan, 7023 (TRACY). Mexico (or D. F.). “Envir. de Mexique,” Berlandier 641, fragment from type (US). San Luis Porosi. Charcos, Whiting 525 (TEX, ARIZ, MICH). VERA Cruz: “Region d’ Orizaba,” Bourgeau 2374 (US). Zacatecas. Zacatecas, Pringle 1701 COC, ARIZ). The writer’s concept of this taxon is based on examination of a panicle fragment from the type and the original description by Fournier. The fragment, consisting of three inflorescence branches, was obtained for the United States National Herbarium by Dr. A. S. Hitchcock in 1907. The type is in the Trinius Herbarium at Leningrad and unavailable for study. The following is a critical description of the fragment in the United States National Herbarium: longest raceme 6 cm.; rachis and pedicels ciliate with relatively long hairs; first glume of sessile spikelets averaging 5.8 mm. long, relatively narrow, rather thickly beset below the middle with stiff hairs, with a single moderately sized glandular pit or depression (averaging 0.23 mm. in diameter) about 2 mm. from the tip; awn of 1957] GOULD: ANDROPOGONS 25 lemma averaging 26 mm. long; pedicelled spikelets averaging 3.8 mm. long, slightly longer than the supporting pedicels. Fournier in the original description stated, ‘““Culmo ramiso, .. . nodis barbatis; panicula flabellata e fasciculatis divergentibus composita,...”’. The hybrid origin of Andropogon barbinodis with its high chromosome number has been previously hypothesized (Gould, 1953). It is believed that the factors for pitted spikelets, short inflorescence axis, and brittle rachis have been introduced into this taxon from A. edwardsianus or di- rectly from the Old World A. pertusus complex. The glume pit occurs in A. barbinodis in all possible gradations, from a faintly discernible glandular spot on one or two spikelets of the panicle to a large, deep pit on all sessile spikelets. From one or more centers, the pit character has become dispersed in A. barbinodis populations almost throughout the range of the species. In North America the pits are most consistently present in plants of central Texas and eastern Mexico. Glume pits have been observed in specimens from New Mexico, Arizona, and even southern Utah (Washington County, near Springdale, VM. E. Jones 6071, near St. George, Gould 1359). It is the intent of the writer to assign to A. barbinodis var. perjoratus all plants of the species in which the sessile spikelets predominantly are pitted. This criterion is somewhat arbitrary as there is no distinct break in the pitted-spikelet series, but it does conform with the established concept of the ‘‘perforatus”’ entity. Plants referable to A. barbinodis var. perforatus generally have been confused with those of A. kybridus and A. edwardsianus, the only New World species consistently having pitted sessile spikelets. The A. bar- binodis plants are most readily distinguished from those of these species by the taller, stouter, more freely branched and less strictly erect culms, broader blades, larger panicles, larger pedicelled spikelets, and more densely hairy glume of the sessile spikelet. ANDROPOGON SACCHAROIDES Swartz var. pulvinatus var. nov. Perennis cespitosa 70-130 cm. alta; nodi plerumque breviter barbati; folia glabra sepe glauca, laminis 5-10 mm. latis; panicula exserta 8-16 cm. longa ramis 8-15 in axi 7-12 cm. longo, axillis atro- vel brunneo-pulvinatis; rami inferiores 2-5 cm. longi ramulosi in anthesis patentes demum con- tracti sed basi curvati; racheos segmenta et pedicelli pilosi pilis apicem versus 6—9 mm. longis, sulcata sulca media lata membranacea plerumque atrata; spicula sessilis 3-4 mm. longa lato-oblonga late acuta; spicula pedicellata 3 mm. longa pedicello plerumque brevior; chromosomata somatica 60. Perennial bunchgrass; culms 70-130 cm. tall; culm nodes mostly short- bearded; leaves glabrous, often glaucous, the blades 5-10 mm. broad; panicle well exserted, 8-16 cm. long, with an axis mostly 7-12 cm. long and 8-15 primary branches; lower panicle branches 2—5 cm. long, mostly rebranched, spreading in anthesis but contracting in fruit, the bases re- 26 MADRONO [Vol. 14 maining bowed-out; panicle branches with blackish or brownish, usually hairy pulvini in their axils; rachis joints nearly as long as the sessile spikelets, the latter overlapping for ’% or less of their length; rachis joints and pedicels hairy, the hairs 6-9 mm. long near the apex, shorter below; groove of pedicel broad, membranous, usually dark-colored; sessile spike- lets typically 3-4 mm. long, broadly oblong, with a broadly acute apex; pedicillate spikelets about 3 mm. long, usually slightly shorter than the supporting pedicels; chromosome number 2n=60. Type: Mexico. Coahuila, Rancho Sierra Hermosa, 40 miles west of Monclava, fF. W. Gould 6467, 25 June 1952 (type, TRACY; isotype UC). On rocky, brushy slopes, at 6,700 feet altitude. Distribution and specimens examined: Known only from the type col- lection and from plants grown at College Station, Texas, from seed from the type collection. This variety differs from A. saccharoides var. torreyanus in the taller culms, larger panicles, spreading or loosely contracted inflorescence branches with axillary pulvini, and more widely spaced spikelets. From A. saccharoides var. longipaniculata it is distinguished by the spreading or more loosely contracted inflorescence branches, the pulvini, the more widely spaced and blunter spikelets, and the chromosome number of 2n=60 rather than 120. KEY TO THE NATIVE AND NATURALIZED SPECIES OF ANDROPOGON SUBGENUS AMPHI- LOPHIS IN THE UNITED STATES I. All or some sessile spikelets pitless* Pedicelled spikelets about as large and broad as the sessile ones. Sessile spikelets more than 5 mm. long. Native species . . 7. A. wrightii Sessile spikelets less than 5 mm. long. Introduced species. Panicle axis shorter than the branches; sessile spikelets never pitted -% == %. . . 9. A. ischaemum Panicle axis longer than the raenee casei spikelets without pits or irregularly pitted ... sun, toe ewe A antenmedius Pedicelled spikelets much narrower and usually ehorten than the sessile ones. Sessile spikelets 4.5-7.3 mm. long, awns 20-33 mm. long (18-20 in Cali- fornia A. barbinodis) ; spikelets pitless or both pitted and pitless on the same panicle. Panicle axis less than 5 cm. long; panicle branches 2-8; rachis joints and pedicels densely white villous; culms slender, not over 1 meter tall and usually much shorter; leaf blades rarely over 4 mm. broad; pollen grains averaging 36-41 microns in diameter; chromosome number 2n=120 6. A. springfieldi Panicle axis typically 5-15 cm. or more long; panicle branches typically numerous, rachis joints and pedicels villous but not densely so; culms typically stout, often over 1 meter tall; leaf blades, at least some, often 5-8 mm. broad. Panicles of the larger culms 14-25 cm. long; culms very stout, stiffly erect, 1.2—2.5 meters tall, bluish glaucous below the nodes; culm nodes bearded with spreading hairs 3-6 mm. long; pollen averaging 39-40 microns in diameter; chromosome number, 2n=120 . 5. A. altus *Refers to circular glandular depression on outer (first) glume. 1957] GOULD: ANDROPOGONS 4) Panicles mostly 7-13 cm. long; culms curving-erect, tending to become decumbent and much-branched below in age, mostly 0.7-1.1 meters tall, not bluish glaucous below the nodes; culm nodes bearded with appressed hairs less than 3 mm. long; pollen averaging 45-52 microns in diameter; chromosome number, 2n=180. Sessile spikelets all or mostly pitless . 8a. A. barbinodis var. barbinodis Sessile spikelets mostly pitted . . . 8b. A. barbinodis var. perforatus Sessile spikelets less than 4.5 mm. long, awn of lemma less than 19 mm. long; spikelets never pitted. Awns absent or not more than 6 mm. long... .. ~=1. A. exaristatus Awns present, 8-18 mm. long. Panicles 5—9, occasionally -13 cm. long; glumes ovate, dull green and most commonly with a whitish waxy bloom; pollen averaging 34-37 microns in diameter; chromosome number, 2n=60 oe 2a. A. saccharoides var. VOPPes Es Panicles of the larger culms 10-20 cm. long; glumes narrowly ovate, shiny green; pollen averaging 39-42 microns in diameter; chromo- some number, 2n=120; southeastern and southern Texas to northern Mexico... .. .. . .. 2b. A. saccharoides var. longipaniculata II. All sessile spikelets pitted Pedicelled spikelets about as large and broad as the sessile ones. Sessile spikelets more than 5 mm. long. Native species . . 7. A. wrightii Sessile spikelets less than 5 mm. long. Introduced species. Panicle axis shorter than the lower branches . . . . . 10. A. pertusus Panicle axis longer than the lower branches . . . .. 11. A. intermedius Pedicelled spikelets much narrower and usually shorter than the sessile ones. Panicle axis less than 5 cm. long; primary panicle branches mostly 2-7, rarely more than 8; culm nodes glabrous or minutely pubescent. Upper culm nodes glabrous; primary panicle branches never rebranched; first glume of sessile spikelet 5.5-7.0 mm. long, glabrous on back, with a relatively large and deep glandular pit; leaves mostly in dense basal tuft, the culm leaves reduced; leaf blades rarely over 2 mm. broad; chromosome number 2n=60 . .. . . 3. A. edwardsianus Upper culm nodes glabrous or puberulent; loan 1 or 2 panicle branches frequently rebranched; first glume of sessile spikelet 4.5-5.7 mm. long, usually sparsely hispid on back near base; glume pit relatively small and shallow; culm leaves well developed; blades 2-5 mm. broad; chromosome number, 2n=120 . . . . . . 4. A. hybridus Panicle axis typically 5-15 cm. or more long; Sele branches numerous. Panicles of the larger culms 14-25 cm. long; culms very stout, stiffly erect, 1.22.5 meters tall, bluish glaucous below the nodes; culm nodes beard- ed with spreading hairs 3-6 mm. long; panicle axis and branches often remaining “kinked” from compression in the sheath; pollen averaging 39-40 microns in diameter; chromosome number, 2n=120 .. _'‘'S. A. altus Panicles mostly 7-13 cm. long; culms curving-erect, tending to become de- cumbent and much branched below in age, mostly 0.7-1.1 meters tall, not bluish-glaucous below the nodes; culm nodes bearded with ap- pressed hairs less than 3 mm. long; panicle axis and branches not “kinked”; pollen averaging 45-52 microns in diameter; chromosome number, 2n=180 . . . . . . 8b.A. barbinodis var. perforatus 1. A. EXARISTATUS (Nash) Hitchc., Biol. Soc. Wash. Proc. 41:163. 1928. Andropogon saccharoides var. submuticus Vasey ex Hack. in DC., 28 MADRONO [Vol. 14 Monogr. Phan. 6:495. 1889. Not A. submuticus Steud., 1854. Amphilophis exaristatus Nash in Small, Fl. Southeast. U. S. 65. 1903. Bothriochloa exaristata (Nash) Henr. Blumea 4:520. 1941. Distribution: Along the Gulf Coast of Louisiana and Texas; coastal Brazil, Argentina. 2a. A. SACCHAROIDES Swartz var. TORREYANUS (Steud.) Hack. in DC., Monogr. Phan. 6:495. 1889. Andropogon glaucus Torr., Ann. Lyc. N. Y. 1:153. 1824. Not A. glaucus Retz., 1789. Andropogon torreyanus Steud., Nom. Bot. ed. 2. 1:93. 1840. Based on A. glaucus Torr. Andropogon jamesu Torr. in Marcy, Expl. Red River 302. 1853. Distribution: Alabama, Missouri and Colorado to northern Mexico. 2b. A. SACCHAROIDES var. LONGIPANICULATA Gould, Field and Lab. 23(1):17-19. 1955. Distribution: Southern and southeastern Texas to northeastern Mexico. 3. A. EDWARDSIANUS Gould, Field and Lab. 19:183-185. 1951. Distribution: Edwards Plateau of central Texas; Argentina and Uru- guay. 4, A. HYBRIDUS Gould, sp. nov. Distribution: Southern Texas and northeastern Mexico. 5. A. ALTUS Hitchc. Contr. U.S. Nat. Herb. 17(3) :308. 1913. Distribution: Western Texas and southern New Mexico to west-central Mexico; Bolivia and Argentina. 6. A. SPRINGFIELDII Gould, sp. nov. Distribution: Western Texas and New Mexico to northern Arizona; Argentina. 7. A. WRIGHTII Hack. Flora 68:139. 1885. Amphilo phis wrightu (Hack.) Nash. Bothriochloa wrightu (Hack.) Henr., Blumea 4:520. 1941. Distribution: New Mexico and northern Mexico. 8a. A. BARBINOpDIS Lag. [Gen. et Sp. Nov. 3. 1816] var. BARBINODIS. Amphilophis barbinodis (Lag.) Nash in Small, Fl. Southeast. U. S. 65. 1903. Bothriochloa barbinodis (Lag.) Herter, Sudamer. Bot. Rev. 6:135. 1940. Distribution: Texas, southern Colorado, Utah, and California, south to Argentina. 8b. A. BARBINODIS Lag. var. PERFORATUS (Trin.) Gould, comb. nov. Distribution: South-central U. S., Mexico, Argentina, and Uruguay. 9. A. ISCHAEMUM L. Sp. Pl. 1047. 1753. Amphilophis ischaemum (L.) Nash, N. Amer. Fl. 17:124. 1912. Bothriochloa ischaemum (L.) Keng. Contr. Biol. Lab. Sci. Soc. China Bot. Ser. 10:201. 1936. Distribution: Widespread in tropical and temperate regions of Asia, 1957] GOULD: ANDROPOGONS 29 Africa and Europe. Established as a pasture and wayside grass in Texas and occasional elsewhere in the United States from pasture plantings. 10. A. peERTUSUS (L.) Willd., Sp. Pl. 4:922. 1806. Amphilophis pertusa (L.) Stapf. in Prain, Fl. Trop. Afr. 9:175. 1917. Bothriochloa pertusa (L.) A. Camus, Ann. Soc. Lyon n. ser. 76:164. 1931. Distribution: Tropical and subtropical Asia, Africa, and Australia. Occasional in southern Texas as a seeded pasture grass. 11. A. INTERMEDIUS R. Br., Prodr. Fl. Nov. Holl. 202. 1810. Amphilo- phis intermedia (R. Br.) Stapf in Prain, Fl. Trop. Afr. 9:174. 1917. Bothriochloa intermedia (R. Br.) A. Camus, Ann. Soc. Linn. Lyon n. ser. 76:164. 1931. Distribution: China, India, the Indo-Malay region and Australia. Intro- duced as a pasture grass in Texas. SUMMARY Three new species, one new variety, and two new name combinations are proposed in Andropogon, subgenus Amphilophis. A key is presented to the eight indigenous and three naturalized species of this subgenus occur- ring in the United States. A unique feature of the key is the prominent use of the glume pit character. Of the species included in the key, the gland- ular glume pit is consistently present in three, consistently absent in four, and of variable occurrence in five. Department of Range and Forestry, Texas Agricultural Experiment Station and A. and M. College of Texas, College Station, Texas LITERATURE CITED Gieason, H. A. 1952. The New Britton and Brown Illustrated Flora of the North- eastern United States and Adjacent Canada. Vol. 1. The Lancaster Press, Lan- caster, Penna. GouLp, F. W. 1951. Grasses of southwestern United States. Univ. Ariz. Biol. Sci. Bull. No. 7. Univ. Ariz. Press, Tucson. . 1953. A cytotaxonomic study in the genus Andropogon. Am. Jour. Bot. 4:297-306. HacxkeE1, E. 1889. In A. De Candolle and C. De Candolle, Monographiz Phanero- gamarum. Paris. Harrincton, H. D. 1954. Manual of plants of Colorado. Sage Books. Denver, Colo. Hircucock, A. S. 1951. Manual of the grasses of the United States. U. S. Dept. Agr. Misc. Publ. 200. Washington, D.C. Second ed., revised by Agnes Chase. SWALLEN, J. R. 1951. Graminez. In T. H. Kearney and R. H. Peebles, Arizona Flora. Univ. Calif. Press, Berkeley. 30 MADRONO (Vol. 14 CHROMOSOME NUMBERS IN LUPINUS LYLE L. PHILLIPS The genus Lupinus, a member of the sub-family Papilionoidez of the Leguminosz, is a group of world-wide distribution with population centers in western United States, Europe, and South America. The present cyto- logical study was undertaken in conjunction with a taxonomic revision of the perennial lupines of North America (Phillips, 1955) in which sixteen species and sixteen infra-specific taxa are recognized. Chromosome num- bers are listed below for twenty-six of these taxa. gt ~ Vyas - . eins 8 ° ye t ~ $9 cnso@ 2 is : to ey? "a4 4¥@ ty og inte ostiy? be” Fics. 1-5. Lupinus meiosis and mitosis: 1, L. laxiflorus var. laxiflorus, M;, n=24; 2, L. sulphureus subsp. sulphureus, Diak., n—24; 3, L. saxosus, late Diak., n=48; 4, L. humicola, T;, n=24; 5, L. sericeus subsp. sericeus, c-mitotic metaphase, n=24. The camera lucida drawings of the chromosomes were made at a magnification of 1940 and reduced to 970. The chromosome number determinations were made either at diakinesis or metaphase I of microsporogenesis or metaphase of root mitosis. The meiotic material was fixed in Carnoy III (3 parts ethanol, 4 parts chloro- form, and 1 part acetic acid) and smeared in aceto-carmine or propiono- carmine. Root tips were treated in oxyquinoline according to Tjio and Levan (1950) and smeared in aceto-orcein. Pollen fertility analyses were made with cotton blue lacto-phenol. 1957] PHILLIPS: LUPINUS 31 The present study on the perennial lupines and several previous reports on the chromosome numbers of Old World species (Kawakami, 1930; Savchenko, 1936; Tuschnjakowa, 1935; and Maude, 1940) make it ap- parent that the basic number of the genus is 12. Diploid (n=12) and tetraploid (n—=24) species as well as several taxa that deviate from the basic number (n=21, 25, 26) are cited for Europe and Africa. Of the Bi oe age ene Intersterile = Sterile F, , pollen apparently normal Fertile F,, backcross or self successful Fic. 6. Breeding behavior of eight species of Lupinus of the Northwestern United States. All taxa n=24: alb. = L. albicaulis; arg. = L. argenteus subsp. argenteus ; cau. = L. caudatus subsp. caudatus,; lax. = L. laxiflorus var. laxiflorus; lep. = L. lepidus subsp. lepidus; leu. = L. leucophyllus; pol. = L. polyphyllus var. poly- phyllus; ser. = L. sericeus subsp. sericeus. 32 MADRONO (Vol. 14 twenty-six North American taxa examined cytologically twenty are tetra- ploid, two are octaploid (n=48), and four are both tetraploid and octa- ploid. The octaploid chromosome level has been heretofore unreported for the genus. In none of the four taxa which contain two chromosome “races”’ is this difference in chromosome number correlated with morphological dissimi- larities. Apparently the genetic isolation created by chromosome doubling in these taxa has not been operative long enough to permit divergence into morphologically definitive types. In a few instances populations of octa- ploids are somewhat unique as compared with related tetraploids, but these unique individuals or populations fall within the variation pattern of the taxon as a whole and cannot justifiably be given specific or infra- specific recognition. Figure 6 presents a summary of a hybridization study involving eight tetraploid species native to Northwestern United States. It can be seen that all of the interspecific crosses were successful except those crosses involving L. lepidus. The F, hybrids exhibited nearly regular meioses (occasionally lagging chromosomes were seen at metaphase I), and a fairly high degree of pollen fertility (75-85 per cent). Attempted crosses of L. lepidus X L. argenteus and L. lepidus X L. leucophyllus resulted in the production of normal seed pods containing aborted ovules. Since no such stimulatory effect on pod development was observed with other crosses involving L. lepidus, this is interpreted to mean that, of the species studied, L. Jepidus is most closely related to L. argenteus and L. leuco- phyllus. The apparent lack of genetic barriers between these species, demon- strable under experimental conditions, is also evident in the field where hybrid individuals often result wherever two or more species are sympatric. Occasionally hybrids and introgressants in such a sympatric association will completely blur species boundaries, but more often the discernible intermediates are relatively few in number. Presumably, the plants of hybrid nature are not able to compete with parental species except where there are uncolonized habitats available for which they are better adapted than the parents. The low level of genetic differentiation between the species utilized in this study supplies a reasonable explanation for the extreme variability within species and for the overlapping variation pattern between many species. Species that can exchange genetic material readily are bound to be variable and difficult to separate taxonomically. Hence Lupinus has become known to taxonomists as a “difficult”? genus. For some of the wide-ranging taxa (L. laxiflorus var. laxiflorus, L. sert- ceus subsp. sericeus, L. polyphyllus var. polyphyllus) the citations listed below constitute only a portion of the collections counted. In these taxa the collections cited have been selected to reflect the geographical range from which cytological analysis has been made. The collections listed below are deposited in the Washington State College Herbarium. CHROMOSOME 1957] TAXON NuMBER (n) L. albicaulis 24 24 24 L. argenteus 24 subsp. argenteus 24 24 24 24 subsp. parviflorus 24 L. caudatus 24 subsp. caudatus 24 subsp. argophyllus 48 L. humicola 24 24 24 L. laxiflorus 24 var. laxiflorus 24 24 24 24 24 24 48 48 var. pseudoparviflorus 24 PHILLIPS: LUPINUS 33 COLLECTION Seattle, King County, Washington, Phillips 690. 5 miles south of Kelso, Cowlitz County, Washington, Phillips 667. Mollala, Clackamas County, Oregon, Phil- lips 720. 5 miles west of Bridgeport, Baker County, Oregon, Phillips 634. Pierce, Clearwater County, Idaho, Phillips 780. Alberton, Missoula County, Montana, Phillips 860. 10 miles east of Livingston, Park County, Montana, Phillips 855. 12 miles west of Custer, Custer County, South Dakota, Phillips 846. 5 miles east of Soda Springs, Bear Lake County, Idaho, Phillips 792. Baker, Baker County, Oregon, Phillips 635. 2 miles south of Madras, Jefferson County, Oregon, Phillips 627. Monticello, Summit County, Utah, J. Nish- itani 7-1952. Leavenworth, Chelan County, Washington, Phillips 733. Near Manhattan, Broadwater County, Montana, Phillips 858. Acme, Sheridan County, Wyoming, Phillips S36 Omak, Okanagan County, Washington, Phillips 606. Winton, Chelan County, Washington, Phil- lips 728. Selah, Yakima County, Washington, Phil- lips 627. Near Mount Hood, Hood River County, Oregon, Phillips 621. 6 miles east of Sisters, Deschutes County, Oregon, Phillips 707. 2 miles south of White Bird, Idaho County, Idaho, Phillips 800. 12 miles north of Boise, Ada County, Ida- ho, Phillips 803. Lyle, Klickitat County, Washington, Phil- lips 683. Underwood, Skamania County, Washing- ton, Phillips 610. Near Priest River, Bonner County, Idaho, Phillips 852. 34 TAXON L. lepidus subsp. lepidus subsp. lyalli L. leucophyllus L. littoralis L. polyphyllus var. polyphyllus var. prunophilus L. perennis subsp. perennts CHROMOSOME NUMBER (n) MADRONO [Vol. 14 COLLECTION 24 24 24 24 24 24 24 24 24 24 24 24 24 24 48 48 24 24 24 24 24 24 24 St. Regis, Mineral County, Montana, Phil- lips 805. Spanaway, Pierce County, Washington, Phillips 582. 3 miles south of Goldendale, Klickitat County, Washington, Phillips 612. Dayville, Grant County, Oregon, Phillips 630. Near Ukiah, Umatilla County, Oregon, Phillips 715. 6 miles north of Modoc Point, Klamath County, Oregon, Phillips 892. Toll Gate, Umatilla County, Oregon, Phil- lips 699. Near Thorpe, Kittitas County, Washington, Phillips 642. 2 miles north of Spangle, Spokane County, Washington, Phillips 876. Goldendale, Klickitat County, Washing- ton, Phillips 658. Near Pullman, Whitman County, Washing- ton, Phillips 842. La Grand, Umatilla County, Oregon, Phil- lips 636. Dixie, Baker County, Oregon, Phillips 633. 5 miles north of Boise, Ada County, Idaho, Phillips 804. Near Goldendale, Klickitat County, Wash- ington, Phillips 678. Wapato, Yakima County, Washington, Phillips 620. Hecata Beach, Lane County, Oregon, Kruckeberg 3315. Montsanto, Thurston County, Washington, Phillips 596. Mission Peak, Kittitas County, Washing- ton, Phillips 676. 2 miles east of Livingston, Park County, Montana, Phillips 851. Oswego, Clackamas County, Oregon, Phil- lips 646. Near Viola, Garfield County, Washington, Phillips 902. Wawawai, Whitman County, Washington, Phillips 869. 4 miles east of Plymouth, Marshall Coun- ty, Indiana, Phillips 822. 1957 | TAXON subsp. Jatifolius subsp. plattensis L. saxosus L. sericeus subsp. sericeus subsp. asotinensis subsp. sabinii L. suksdorfu L. sulphureus subsp. sulphureus subsp. kincaidit subsp. subsaccatus subsp. whithamii PHILLIPS: LUPINUS 30 CHROMOSOME NUMBER (n) COLLECTION 24 24 24 48 24 48 24 24 24 24 24 24 24 48 24 24 24 24 Amboy, Lee County, Illinois, Phillips 815. Near Hanover, Lebanon County, Pennsyl- vania, Phillips 830. Mt. Rainier, Pierce County, Washington, Phillips 613. Zigzag, Clackamas County, Oregon, Phil- lips 628. 5 miles east of Kimball, Kimball County, Nebraska, Phillips 809. 10 miles south of Liberty, Kittitas County, Washington, Phillips 689. Maryhill, Klickitat County, Washington, Phillips 687. Big Timber, Sweetgrass County, Montana, Phillips 851. Gillette, Campbell County, Wyoming, Phil- lips 849. Orofino, Clearwater County, Idaho, Phil- lips 890. Indian, Whitman County, Washington, Phillips 792. 10 miles west of Clarkston, Asotin County, Washington, Phillips 811. Elgin, Union County, Oregon, Phillips 736. Glenwocd, Klickitat County, Washington, Phillips 679. Kooskooskie, Walla Walla County, Wash- ington, Phillips 696. 2 miles east of Viola, Garfield County, Washington, Phillips 903. Silverton, Polk County, Oregon, Phillips fod. 10 miles south of Wenatchee, Kittitas County, Washington, Phillips 746. Cle Elum, Kittitas County, Washington, Phillips 688. Ellensberg, Kittitas County, Washington, Phillips 674. 6 miles south of Coulee City, Grant Coun- ty, Washington, Phillips 882. Butch Creek, Pend Oreille County, Wash- ington, Rumely & Phillips 453. Near Nordman, Bonner County, Idaho, Rumely & Phillips 455. West shore of Priest Lake, Bonner County, Idaho, Rumely & Phillips 456. 36 MADRONO [Vol. 14 SUMMARY Chromosome number determinations for 26 taxa of North America indi- cate twenty of these to be tetraploid (n=24), two to be octaploid (n=48), and four taxa to be both tetraploid and octaploid. A hybridization study involving eight species of Northwest United States shows genetic incompatibility barriers to be poorly developed be- tween these species, thus supplying a possible reason for the overlapping patterns of morphological variation found in the genus Lupinus. Department of Field Crops, North Carolina State College, Raleigh LITERATURE CITED KAWAKAMI, N. 1930. Chromosome number in Leguminose. Bot. Mag. Tokyo 44: 319. Maupeg, M. 1940. Chromosome numbers in some British Plants. New Phytol. 39: 17. Puiiies, L. 1955. A revision of the perennial species of Lupinus of North America. Res. St. State Coll. Wash. 23: 161. SAVCHENKO, N. 1936. Karyology of some species of the genus Lupinus. Bull. Appl. Bot. Select. II 8: 105. Ty10, J. and A. Levan. 1950. The use of oxyquinoline in chromosome analysis. An Aula Dei 2: 21. Tuscunyakowa, M. 1935. Uber die chromosomen einiger Lupinus-Arten. Ziichter T2169: JEROME D. LAUDERMILK Mr. Jerome D. Laudermilk, who passed away in January, 1956, was a general scientist. The originality of his inquisitive mind impressed those who knew him well. He read widely and probed deeply as he read. Char- acteristically he was not satisfied to accept Leeuwenhoek’s account of his microscope until he had ground lenses and made a microscope of his own exactly according to Leeuwenhoek’s formula. The structure of ancient weapons was a special field of research, and he lectured and demon- strated his models publicly and for the Pomona College Department of Military Science and Tactics. He was interested so deeply in the opera- tions of those who deal in the occult that at one time he was kidnapped, taken to an obscure house, and convinced that his life would be longer if he did not write on the subject. Jerry Laudermilk was a graduate of Kansas State College of Pharmacy, and he served in the United States Army in World War TI. Being in ill health he spent several years in the desert near Wickenberg, Arizona, where he developed a deep interest in and knowledge of desert vegetation. He came to southern California thirty-five years ago, and he lived for the last thirty years in Claremont, where he was Research Associate in Geo- chemistry and Paleobotany at Pomona College. There, in association with Dr. Philip A. Munz, he investigated the food habits of extinct giant sloths by study of the dung of the animals in the caves they inhabited in the deserts near the Colorado River. This has provided knowledge of the past vegetation of the area. 1957] BARNEBY: ASTRAGALUS 37 His interests carried him into many problems concerned with plants, minerals, fossils, and other natural objects. In his investigation of plants Mr. Laudermilk was never satisfied with the statements in books. He went directly to nature and drew or wrote from what he found there. He was an excellent illustrator of books and scientific papers and a painter of ability. He wrote many popular articles presenting science and espe- cially botany and geology as a layman would enjoy it, and in these he brought knowledge from many fields to bear on matters commonly ap- proached by a single avenue. His illustrations and manuscripts found their way into such journals as Natural History, Desert Magazine, and Westways. His last work was the principal series of illustrations for the writer’s textbook entitled “Plant Classification,” scheduled to be pub- lished in February, 1957. Mr. Laudermilk clung to life for many months in the hope of seeing these illustrations in print, and it is a great regret to the author that he was not able to do so— LYMAN BENSoNn, Department of Botany, Pomona College, Claremont, California. ASTRAGALUS AGNICIDUS, A NEW LOCOWEED FROM HUMBOLDT COUNTY, CALIFORNIA R. C. BARNEBY The known history of the Astragalus described below goes back about twenty-five years, when Mr. Henry Tosten, the original discoverer, moved with his family to a ranch situated high in the outer North Coast Range near the divide between the South Fork of the Eel and the Mattole River in southern Humboldt County. Suffering great losses among his sheep, Mr. Tosten quickly identified this species as the culprit. In the summer of 1931 he prepared herbarium material and sent it to the late Mr. J. P. Tracy of Eureka, the outstanding authority on the flora of the region, and the specimens passed in due course to the University of California Herbarium at Berkeley, where I came across them in the winter of 1949. In May, 1954, I was able to visit the Tosten place and the genial owner obligingly took me up to the ridge above the ranch-house, the original Station, where the locoweed still survived in sparing quantity. I am in- debted to Mr. Tosten for the following information. No sooner was the Astragalus recognized as poisonous than vigorous steps were taken to root it out. It was restricted to a wooded ridge, where the natural vegetation had been disturbed by logging, and was so abun- dant in early years that it was possible to collect great piles of stems for burning. Since then intermittent but never wholly successful attempts were made to control or exterminate it, and plans were afoot in 1954 to clear off the hilltop and plough it out. Mr. Tosten early assumed that the plant was an introduced weed; and it is said to be unknown to other ranchers in the community or county. A company of bark-strippers was 38 MADRONO [Vol. 14 encamped on the site for some years before the Tosten occupancy, and this circumstance lends color to the view that it might have been brought in accidentally, possibly, as suggested by Mr. Tosten, from the Sacra- mento Valley. Yet it can be stated emphatically that it is not any species as yet recognized or described from the Americas. All attempts to identify the specimens with some Old World species having proved vain, I am forced to conclude that, in spite of its weedy occurrence, it is most prob- ably native to California and requires a name. The native status of A. agnicidus finds considerable support when we consider its probable relationships. It appears not to fit easily into any group described from Europe or Asia. On the typesheet Mr. Tracy re- marked, in my opinion correctly, that it was near A. umbraticus Sheld., but differed in the pubescence and the pods. While A. agnicidus and A. umobraticus are readily distinguished at the specific level, as shown in the key below, they are much alike in general facies and organization, and together with A. Congdoni Wats. and A. Paysonu (Rydb.) Barneby form a small but homogeneous and presumably natural group in the genus. Leading technical characters common to the four species are: free stip- ules; thin-textured, often visibly penninerved leaflets; nodding or de- clined flowers and fruits; white petals; and pods shortly stipitate or subsessile, continuous with the receptacle (and hence falling naturally with the disjointing pedicels), in form obliquely linear- or lance-oblong, more or less incurved, compressed-triquetrous, grooved dorsally and fully bilocular. The three species known hitherto are all rare or highly localized. Astragalus Congdonit is endemic to the cismontane foothills of the Sierra Nevada, where it ranges in disjunct and scattered stations from the Mokelumne south to the Tule River, and is seemingly confined to areas of old metamorphic, sometimes partly serpentinized, bedrock. As- tragalus umbraticus is known from only seven or eight stations, four of which lie in the Klamath Highland in southwest Oregon, on the Coquille, Rogue and Illinois rivers; from there it extends south to the lower Trinity River and Redwood Creek in Humboldt County, California; and it was collected long ago at an unspecified locality in the Coast Ranges of Yam- hill County in northwest Oregon. The somewhat less closely related 4. Paysontu is known as yet only from two records, one from the Snake- Green River divide in western Wyoming, the other at a point over three hundred miles distant to the northwest in the Clearwater Mountains of central Idaho. The distributional pattern of these species, considered in- dividually and collectively, suggests that the section is a relatively old one, very likely composed of homogenic depleted species as defined by Stebbins (in Madrono 6:241—258, 1942). Stebbins has pointed out the alternative consequences of a change in environment on rare or relic species consisting of few biotypes: eventual extinction where the change is detrimental to their welfare, or regained vitality when conditions are altered to their advantage, or where competition is, even temporarily, re- duced. It seems possible that disturbance of the highly competitive climax 1957] BARNEBY: ASTRAGALUS 39 woodland and the sudden weedy abundance of A. agnicidus are related phenomena. Further exploration of the more inaccessible parts of the North Coast Range may yet provide a definite answer. The species discussed above may be distinguished by the following key: 1. Leaflets (except in a few early leaves) 15-35; flowers relatively large, the banner 9-i6.5 mm., the keel 7-12.5 mm. long; cismontane Oregon and California. 2. Stems and herbage nearly glabrous, the few scattered hairs strictly appressed and not over 0.6 mm. long; ovary and pod glabrous; Coast Ranges, from Yamhill County, Oregon to central Humboldt County, California. Stipe of the pod 0.8-1.9 mm. long, the body 1.4—2.4 cm. long, 2.6-3.6 mm. in diameter, OST ON AML tee ace see chan ten) Lott PW Gc cec ease cate ites ota A. umbraticus 2. Stems (at least above the base) and herbage villous or pilose, the longest hairs at least 0.9 mm. long; ovary and pod pubescent; southern Humboldt County and Sierra Nevada, California. 3. Raceme compact and dense, the axis little-elongating, 2—4.5 cm. long in fruit; calyx-teeth linear or lance-acuminate, 3.3-5 mm. long; pod thinly villous- pilose, the stipe 0.3-0.4 mm., the body 11-15 mm. long, 3-3.4 mm. in diam- eter, 8-9-ovulate; Humboldt County.....002... 00000002222 A. agnicidus 3. Raceme open, becoming loosely secund and mostly 5-20 cm. long in fruit; calyx-teeth subulate, 1-2.5 mm. long; pod more densely strigulose-villosu- lous, the stipe 1-2.5 mm., the body (1.5) 2-3.5 cm. long, 2.3-3.2 mm. in diameter, 23—29-ovulate; Sierra foothills............022...00202222222eeeeeeee ee A. Congdoni 1. Leaflets 7-15; flowers small, the banner about 7 mm., the keel about 5 mm. long; western Wyoming and central Idaho. Stipe of the pod 1-1.5 mm., the body 10-17 mm. long, 2.5—3.5 mm. in diameter, 8-10-ovulate...........0000222002... A. Paysoni Astragalus agnicidus sp. nov. Herbae elatae foliosae e radice verti- cali ramosa perenni, pilis debilibus patulis subsinuosis rectisque parce villosae, foliolis bicoloribus inferne pallidis ad nervum medianum barba- tis, superne saturatius viridibus glabris, ciliatis, inflorescentia nigro- villosula; caules erecti et adscendentes fistulosi striati straminei (3) 4—9 dm. longi, ad medium ramosi vel subsimplices; stipulae membranaceae 4-15 mm. longae, imae ovato-triangulares amplexicaules inter se liberae, superiores lanceolato-acuminatae vel lineari-caudatae dimidium caulem amplectentes decurrentes deflexae; folia (3.5) 5-12 (16) cm. longa, supe- riora subsessilia, foliolis (6) 9—13-jugis petiolulatis ovato-vel lanceolato- oblongis obtusis vel emarginatis rarius acutiusculis mucronulatis (3) 5-22 mm. longis, majoribus penninerviis; pedunculi saepius 8-12 supra medium caulem emissi 5—13 cm. longi, folio subaequilongi; racemi dense (10) 15—40-flori, floribus mox patulo-declinatis, axi fructifero vix elon- gato (1) 2-4.5 cm. longo; bracteae hyalinae lanceolatae vel lineari- caudatae 2—6 mm. longae reflexae; bracteolae minutae vel 0; calycis nigro-villosuli tubus campanulatus pallide membranaceus 3.2—4.2 mm. longus, 2.4-3 mm. latus, dentes firmiores virides lineares vel lineari- acuminati 3.3-4.9 mm. longi; petala alba immaculata; vexillum per 45° recurvum, oblanceolato-subrhombicum emarginatum, 9.1-11 mm. lon- gum; alae 8.3-9.2 mm. longae, laminis oblanceolatis obtusis vel oblique obovatis emarginatis subrectis 5.4-6.5 mm. longis, 1.5—2.6 mm. latis; carina 7—7.4 mm. longa, laminis semi-obovatis 3.9—4.2 mm. longis, 2—2.4 mm. latis, per 90—95° in apicem obtusum deltoideum incurvis; legumen 40 MADRONO [Vol. 14 patulo-declinatum subsessile, stipite vix 0.4 mm. longo calyce persistenti occultato, de visu laterali anguste lanceolatum paullo incurvum 11-15 mm. longum, 3—3.4 mm. latum, basi obtusum, apice in rostrum anguste triangulari-acuminatum cuspidatum angustatum, triquetro-compressum, sutura ventrali prominula concave arcuata carinatum, dorso anguste sul- catum, valvulis tenuibus piloso-villosulis demum chartaceis reticulatis stramineis, late inflexis, septo completo 1.5—2.2 mm. lato; ovula 8-9; semina (vix matura) brunnea laevia 1.7—2.1 mm. longa. Astragalo umbratico Sheld. affinis, sed caulibus elatis, pube magis copiosa patula villosa multo longiori, dentibus calycinis elongatis, necnon legumine breviori villosulo 8-9 (nec 10—15)-ovulato summopere distincta. The species name agnicidus is derived from agnus, lamb, and caedere, to kill; the species was first brought to notice by its reputedly poisonous qualities. Specimens examined. CALIFORNIA. “Local on Tosten & Peirce Ranch, near Bear Buttes, 4 miles s. of Miranda, Humboldt County, alt. about 2500 ft., June 7, 1931, Henry Tosten ex herb. J. P. Tracy. Said to be a sheep poison and attempted to be eradicated, fall of 1931” (type, UC, two sheets, 502991, 502992). Topotypes: August 20, 1931, J.C. Taris Jr., UC; May 19, 1954, just coming into flower, on brushy logged-over ridge, Barneby 11570 (CAS, RSA, author’s coll.). Loan of the material at the Herbarium of the University of California, above cited, is hereby gratefully acknowledged. Wappingers Falls, New York. NoTEs AND NEws ANEMOPSIS CALIFORNICA IN OrEGON. An apparently well-established clump of Anemopsis californica was found in an roadside irrigation ditch, along Crystal Springs Road about one mile southwest of the bridge that crosses Lost River, Klamath Falls, Oregon, on August 15, 1955 (Pengelly 743). The plant was associated with cattails (Typha) and arrow-leaf (Sagittaria) ; however, the ditch passes through typical Artemisia tridentata association. This appears to be the first record of the occurrence of this species in Oregon, the nearest known locality to the south being near the mouth of the Sacramento River.—RvussELL PENGELLY, Klamath Falls, Oregon. Some publications of interest follow: Responses of Vegetation to Fire, by James R. Sweeney. University of California Publications in Botany 28 (4): 143-250, pls. 12-27, 10 figs. in text. 1956. $2.00. Uni- versity of California Press, Berkeley 4, California. A study of the effects of chaparral fires upon herbaceous vegetation. The Genus Clarkia, by Harlan Lewis and Margaret Ensign Lewis. University of California Publications in Botany 20 (4): 241-392, 28 figs. in text. 1955. $2.00. Uni- versity of California Press, Berkeley 4, California. A monograph of the genus based upon a many-faceted, biosystematic approach to the problem. Variation and Genetic Relationships in the Whitlavia and Gymnobythus Phacelias, by George Willson Gillett. University of California Publications in Botany 28 (2): 19-78, pls. 3-5, 16 figs. in text. 1955. $1.00. University of California Press, Berkeley 4, California. A genetic analysis and systematic treatment of two of the seven sub- genera of the genus Phacelia. se INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication should not exceed an estimated 20 pages when printed unless the author agree to bear the cost of the ad- ditional pages at the rate of $15 per page. Illustrative materials (includ- ing “typographically difficult”’ matter) in excess of 30 per cent for papers up to 10 pages and 20 per cent for longer papers are chargeable to the author. Subject to the approval of the Editorial Board, manuscripts may be published ahead of schedule, as additional pages to an issue, provided the author assume the complete cost of publication. Shorter items, such as range extensions and other biological notes, will be published in condensed form with a suitable title under the general heading, ‘Notes and News.” Institutional abbreviations in specimen citations should follow Lanjouw and Stafleu’s list (Index Herbariorum. Part 1. The Herbaria of the World. Utrecht. Second Edition, 1954). | Articles may be submitted to any member of the Editorial Board. MADRONO A WEST AMERICAN JOURNAL OF BOTANY A quarterly journal devoted to the publication of botanical re- search, observation, and history. Back volumes may be obtained from the Secretary at the following rates: Volume I,1916-1929 . . . . =. $5.00 Volume IT, 1930-1934. . . . . 5.00 Volume ITI, 1935-1936 . . . . 5.00 Volume IV, 1937-1938 . . . . 5.00 Volume V, 1939-1940. . . . . 5.00 Volume VI, 1941-1942 . . . . 5.00 Volume VII, 1943-1944 . . . . 5.00 Volume VIII, 1945-1946 . . . 5.00 Volume IX, 1947-1948 . . . . 6.00 Volume X, 1949-1950. . . . . 7.00 Volume XI, 1951-1952. . 8.00 Volume XII, 1953-1954 . . . . 8.00 Volume XIII, 1955-1956 . . . 8.00 Single numbers, Volumes 1 and) 2). (hee ee .50 Succeeding numbers, (except Volume XI, No.2) . .. 1.00 Volume XTNo.2.° 2s ed 1.50 The subscription price of MapRONo is $4.00 per year. If your institution does not now subscribe to MApDRONO, we would be grateful if you would make the necessary request. Since there is no extra charge for institutional subscriptions, an individual may hold membership in the California Botanical Society on the basis of his institution’s subscription. Address all orders to: G. THomas Rosstns, Corresponding Secretary Department of Botany University of California, Berkeley 4, California ADRONO VOLUME 14, NUMBER 2 APRIL, 1957 Contents PAGE NOTES ON PActIFIC MARINE Atcak, Paul C. Silva 41 A NEw YUCCA FROM SONoRA, MEXICco, Howard Scott Gentry 51 A REVISION OF THE LUPINUS ARBUSTUS COMPLEX OF THE LAXIFLoRI, David B. Dunn 54 SOME ADDITIONS TO THE CALIFORNIA Moss FtLora, Howard Crum 74 Review: T. H. Goodspeed, The Genus Nicotiana (Richard W. Holm) 79 A WEST AMERICAN JOURNAL OF BOTANY ‘/PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO A WEST AMERICAN JOURNAL OF BOTANY Entered as second-class matter at the post office at Berkeley, California, January 29, 1954, under the Act of Congress of March 3, 1879. Established 1916. Subscription price $4.00 per year. Published quarterly and issued from the office of Madrofio, Herbarium, Life Sciences Building, University of California, Berkeley 4, California. BOARD OF EDITORS HERBERT L. Mason, University of California, Berkeley, Chairman Epcar ANDERSON, Missouri Botanical Garden, St. Louis. LyMAN BENSON, Pomona College, Claremont, California. HERBERT F. COPELAND, Sacramento College, Sacramento, California. Joun F. Davipson, University of Nebraska, Lincoln. Ivan M. JoHNsTON, Arnold Arboretum, Jamaica Plain, Massachusetts. MILpreD E. MArutias, University of California, Los Angeles 24. Marion OwnBEY, State College of Washington, Pullman. Ira L. Wiccins, Stanford University, Stanford, California. Secretary, Editorial Board — ANNETTA CARTER Department of Botany, University of California, Berkeley. Business Manager and Treasurer—MaAtcotm A. Noss Carnegie Institution of Washington, Stanford, California CALIFORNIA BOTANICAL SOCIETY, INC. President: Rimo Bacigalupi, Jepson Herbarium, Department of Botany, Univer- sity of California, Berkeley, California. First Vice-president: Richard W. Holm, Natural History Museum, Stanford University, Stanford, California. Second Vice- president: Lyman Benson, Department of Botany, Pomona College, Claremont, Cali- fornia. Recording Secretary: Mary L. Bowerman, Department of Botany, University of California, Berkeley, California. Corresponding Secretary: G. Thomas Robbins, Jepson Herbarium, Department of Botany, University of California, Berkeley, Cali- fornia. Treasurer: Malcolm A. Nobs, Carnegie Institution of Washington, Stanford, California. ——— 1957] SILVA: PACIFIC MARINE ALGAE 41 NOTES ON PACIFIC MARINE ALGAE! PAUL C. SILVA For the past ten years I have been engaged in taxonomic and phytogeo- graphic studies of the marine algae of the Pacific coast of North America. Although it is planned to present the results ultimately as a floristic trea- tise, the completion of this project is so far from sight as to suggest the usefulness of publishing certain nomenclatural, taxonomic, and distribu- tional notes at this time. All specimens cited in this paper are in the Her- barium of the University of California, Berkeley. Collections are mine unless otherwise indicated. Rosenvingiella nom. nov. Gavyella Rosenvinge (1893, p. 936). Non Gayella Pierre (1890, p. 26). Lectotype: G. polyrhiza Rosenvinge. Although unanimity of opinion regarding the validity of this genus of Prasiolaceae is lacking, many phycologists prefer to regard the terete members of this family as constituting a genus separate from Prasiola. Unfortunately, the name Gavyella had been used previously (in the Sa- potaceae) by Pierre. Therefore I have proposed the substitute name, Rosenvingiella. Two species have been reported from the Pacific coast of North America. Rosenvingiella polyrhiza (Rosenvinge) comb. nov. Gavyella polyrhiza Rosenvinge (1893, p. 937, figs. 45, 46). Rosenvingiella constricta (Setchell et Gardner) comb. nov. Gavyella constricta Setchell et Gardner in Gardner (1917, p. 384, pl. 33, figs. 5-9; i oz, ie. 5). Rosenvingiella constricta and Prasiola meridionalis Setchell et Gardner (1920) occupy a unique ecological niche: they are restricted to rocks (usually offshore) covered with guano and pounded by heavy surf, upon which they form patches or distinct bands in the spray zone. Rosenvin- giella constricta inhabits only the lower part of this association, where it may grow intermixed with Prasiola meridionalis or in nearly a pure stand. The green patches or bands may be seen from a considerable distance, but only occasionally is one able to get close enough to discern the two com- ponent growths, which differ slightly in color. Still more rarely is one able to collect from these rocks, so that the range of these associated species as indicated by documented records probably is smaller than the actual range. Heretofore these two species have not been recorded south of Car- mel Bay. Although I have observed probable stands along the central California coast south of Carmel and on the Channel Islands, only once did circumstances permit collections to be made OB Cove, Santa Cruz Island, 12 March 1950). 1 Grateful acknowledgment is made to the National Science Foundation and to the Research Board of the University of Illinois for grants in support of this study. Maprono, Vol. 14, No. 2, pp. 41-80, May 1, 1957. BRAY -« MAY 4 42 MADRONO [Vol. 14 Ectocarpus dimorphus nom. nov. Ectocarpus variabilis (Saunders) G. M. Smith (1942, p. 647, figs. 1-4). This distinctive species is commonly found growing on various Lami- nariales. Since £. variabilis G. M. Smith is a later homonym of E. varia- bilis Vickers (1905, p. 59) from Barbados, and since no other epithet is available, I have proposed a name which refers to the two kinds of pluri- locular organs found in this species. DICTYOPTERIS JOHNSTONE! Gardner (1940, p. 270, pl. 35). This spe- cies, which was published posthumously, is based on a “single, apparently fragmentary, part of a plant’ dredged from 25 fathoms at Lone Cove, Santa Cruz Island, California (G. R. Johnstone 96, 28 November 1928, UC 472507). It is supposedly distinguished from D. zonartoides Farlow (1899) by the narrow segments and inconspicuous midrib. Hollenberg (1948) reported this species from “‘a large tide pool at medium high tide level on exposed rocky shore several miles south of Redondo Beach.”’ Hol- lenberg’s plant differs from the type, however, in being larger, coarser, and with well-developed percurrent axes. Material agreeing with the type was dredged by an Allan Hancock Foundation Expedition (Station 1431- 41, 26 September 1941) off White Cove, Santa Catalina Island (Dawson, 1949, p. 21). Abundant material of Dictyopteris was dredged from 40 meters at the same locality by the M/V Orca of the J. W. Sefton Founda- tion Expeditions (42317, 8 February 1949), providing a series of specimens that strongly suggests the conspecificity of D. johnstonei and D. zonar- ioides. Although many specimens are typical D. zonarioides, others show a diversity of segment size, having some branches typical of D. zonar- 1oides, other branches typical of D. johnstonei, and still other branches intermediate. Entire plants of the narrow D. johnstonei form were col- lected on the north shore of the west island, Islas San Benito, Baja Cali- fornia, Mexico (6409, 4 February 1950). The significance of this form remains to be demonstrated. Laminaria setchellii nom. nov. Hafgygia andersonu Areschoug (1883, p. 3). Laminaria andersonu (Areschoug) Farlow ex Anderson (1891, p. 220). Non L. andersonu Eaton ex Hervey (1881, p. 98). The name Laminaria andersonu has been applied to two species, and unfortunately the earlier application was to the species currently called L. sinclairu rather than to the plant which has been assumed to be L. an- dersonii. Doty (1947, p. 40) remarked that from Hervey’s description of the blade “‘one would be inclined to think that he was referring to L. Sin- clairi.”” However, from Hervey’s statements that the plant grew on rocks with Pterygophora and that it had “the usual branching hold-fast’? Doty concluded that L. andersonti as usually interpreted was definitely indi- cated. Hervey described a specimen (which unfortunately I have not been able to locate) sent by Dr. C. L. Anderson from Santa Cruz, California, with a stipe one-sixth of an inch thick and about eighteen inches long bearing a blade about an inch wide and eighteen inches long. This descrip- 1957] SILVA: PACIFIC MARINE ALGAE 43 tion clearly refers to L. sinclairii. Indeed, on the label of specimen no. 118 of Farlow, Anderson and Eaton’s Algae Exsiccatae Americae Borealis, “Laminaria Andersonii Eaton mscr in Rep’t. U.S. Fish Comm. 1875,” is listed as a synonym of L. simclaira. Further clarification is provided by Farlow (1881): ‘‘Since it was ascertained that the species called L. Ander- sonii was the same as Lessonia Sinclairu the manuscript name of L. An- dersonii has been applied by Prof. Eaton and myself to a second species from the California coast which belongs to the digitate division of the genus and not to the section Saccharinae.”’ Hervey apparently was not aware that the plant to which the manuscript name L. andersonu was originally applied has been recognized as L. sinclairu, and he thus un- wittingly created a nomenclatural problem by effecting valid publication. The epithet anderson in its revised application was not validated until 1883, when Areschoug published it under Hafgygza. It was transferred to Laminaria by Anderson (1891), but the resulting combination is a later homonym and hence illegitimate. Inasmuch as no legitimate epithets are available for this species, | have proposed a new name in honor of W. A. Setchell, one of the foremost contributors to our knowledge of the Lami- nariales. Heretofore the known range of this species has been Whidbey Island, Washington, to Carmel Highlands, Monterey County, California (Smith, 1944, p. 137). There are no herbarium specimens to authenticate the statement of Setchell and Gardner (1925, p. 605) that this species occurs as far north as Sitka, Alaska. The following collections extend the range northwest to Vancouver Island, British Columbia, and south to Santa Barbara County, California, and to certain of the Channel Islands. CANADA. British CoLtumBia. Vancouver Island: Point No Point, 16 miles west of Sooke, 7056. UNITED STATES. Catrirornia. Monterey County: mouth of Mal- paso Creek, 3622; Kasler Point, 874; Point Sur, 7101; Partington Point, 2256; Lucia, 2844. San Luis Obispo County: Point Piedras Blancas, 1359; Cayucos, 2317. Santa Barbara County: Point Sal, 5136; Point Pedernales, 2402; Point Arguello, 2541; Point Conception, 2468. San Miguel Island: Cuyler Harbor, 3743. Santa Rosa Island: Sandy Point, 4128. Santa Cruz Island: Fraser Point, 6050. San Nicolas Island: first rocky outcropping west of sand spit, 45417. Laminaria setchellu is abundant in all surf-swept areas throughout the range on the mainland, forming extensive stands at lowest lower low water level (LLLW). On the Channel Islands it is an excellent indicator of localized cold water. LAMINARIA SINCLAIRII (Harvey ex Hooker f. et Harvey) Farlow, An- derson et Eaton (1878, no. 118). Lessonta sinclair Harvey ex Hooker f. et Harvey in Hooker (1846, p. 460). Hafgvgia sinclair (Harvey ex Hooker f. et Harvey) Areschoug (1883, p. 6). Laminaria anderson Ea- ton ex Farlow (1876, p. 715, nomen nudum). Laminaria andersont Eaton ex Hervey (1881, p. 98). The original description of Laminaria sinclairiu (as Lessontia sinclairit) is meager, but sufficient to validate the name. Fail- ure to accept this description as valid publication would cause L. ander- 44 MADRONO [Vol. 14 soni Eaton ex Hervey to be the correct name for this species, thereby intensifying the nomenclatural confusion between L. sinclairii and L. set- chellit. Heretofore the known range of Laminaria sinclairui has been from Van- couver Island, British Columbia, to San Luis Obispo County, California. The following collections extend this range into Santa Barbara County, California: Point Sal, 5735; Point Pedernales, 2403; Point Conception, 2423; Government Point, 5451, 5519, 5577; Gaviota, 5275. This species apparently is absent on the Channel Islands. DICTYONEUROPSIS RETICULATA (Saunders) G. M. Smith (1942, p. 651, figs. 9-13). Two new localities can be recorded for this strikingly beauti- ful kelp, previously known only from sublittoral Monterey Bay. Three juvenile plants were collected by an abalone survey team of the California Department of Fish and Game from the Fort Ross cove area of Sonoma County at a depth of 3 to 12 meters, 20 September 1951. Extensive sub- littoral stands were discovered off the Channel Islands by the Sefton Foundation Expeditions. One collection was dredged from 31 meters be- tween Fraser Point and Kinton Point, Santa Cruz Island (5884) and an- other stand was found at a depth of 24 meters 1'4 miles southwest of Ford Point, Santa Rosa Island (5981), both in March, 1950. At Pacific Grove an occasional plant grows in the littoral zone at lowest lower low water level (3044). The blade of the largest plant dredged off Ford Point mea- sured 19 by 107 cm., and this was clearly an incomplete specimen. It is surprising how steadfastly Setchell refused to recognize the dis- tinctness of this taxon, considering that he himself pointed out the differ- ences between it and Dictyoneurum and even illustrated an unmistakable specimen (1896, p. 46, pl. 1). EcreciaA J. E. Areschoug. Of the nine genera of Laminariales re- stricted to the west coast of North America, Fgregia alone exhibits much variation. Heretofore, with collecting largely confined to the mainland of California north of Carmel and south of Santa Barbara, there have ap- peared to be two well-defined species: L. menziesi (Turner) Areschoug (1876), extending from Vancouver Island, British Columbia, to Point Conception in Santa Barbara County, California; and E. laevigata Set- chell (1896), ranging from Point Conception to Baja California, Mexico, with isolated occurrences in Carmel, Monterey County, and Port Harford (Port San Luis), San Luis Obispo County. A study of a long series of specimens from the Channel Islands and the mainland between Carmel and Santa Barbara reveals a more complex picture: in these areas the two species overlap and intergrade, suggesting hybridization. Typical Egregia menziesii has a tough rachis closely beset with simple, entire or dentate, spatulate blades. Short blunt tubercles cover all or most of the rachis (the lowermost portion may be smooth) and frequently also cover the bladders and blades. Typical E. laevigata has a smooth brittle rachis, somewhat broader than that of /. menziesi, bearing relatively 1957] SILVA: PACIFIC MARINE ALGAE 45 large, elliptical to ligulate, simple and entire or frequently highly dissected blades. The foliar extension of the bladder, which in E. menziesii is stubby, is well developed and ultimately dissected. In the form with the blades dissected into filiform segments, this kelp is known as the “Feather Boa.” Despite the conspicuous differences between typical representatives of the two taxa, when a careful study is made of all available material as to the presence or absence of tubercles and the size, shape, degree of dissec- tion and distribution of blades, the two species are seen to be far less sharply delimited than was previously believed. Typical Egregia menziesii constitutes a remarkably uniform series of populations from Vancouver Island, British Columbia, to Government Point (just south of Point Conception), Santa Barbara County, Califor- nia. At both ends of the range, however, deviation occurs. On Vancouver Island there is a population with highly dissected blades (Tofino, Clayo- quot Sound, Setchell & Parks, 23 June 1930; Point No Point, 16 miles west of Sooke, 7053, 17 June 1955). In the vicinity of Point Conception (from Point Pedernales to Government Point) there are plants (2447), putatively hybrid, that show certain characteristics of E. laevigata of that area, namely, a smooth rachis (except at the tips of branches) and a wider spacing of blades. No plants referable to E. menziesii occur on the mainland south of Government Point, but in the northern Channel Islands plants of Egregia consistently exhibit characteristics of the two species combined with intermediate features apparently at random. In the ma- jority of specimens the rachis is tough (not brittle as in E. laevigata) and covered in part with tubercles that are longer and more slender than those of typical E. menziesii. The blades are variable in size and shape, but in older fronds they become highly dissected, as in EF. laevigata. Although this series of populations seems to be of hybrid origin and is morphologi- cally heterogeneous, its geographical distinctness supports its recognition as a subspecies. Assignment to one or the other species, however, seems to me to necessitate an arbitrary decision. Thus I arbitrarily refer the Channel Islands subspecies to E. menziesit. EGREGIA MENZIESII (Turner) Areschoug subsp. insularis subsp. nov. Axis lentus haud fragilis plerumque partim attamen tuberculis gracilibus obtectus paginae frondium maturiorum profunde dissectae. Rachis tough, not brittle, usually covered at least in part with slender tubercles; blades of older fronds highly dissected. Type. North shore, West Anacapa Island, Ventura County, Channel Islands, California, 14 March 1950, 6130 (UC 981571). Additional records. CaLtrorniA. Channel Islands: San Miguel Island: Cuyler Har- bor, 3744. Santa Rosa Island: Sandy Point, 4124; mouth of Garafion Cafion, 4104; Bechers Bay, 3235; East Point, 3972. Santa Cruz Island: Fraser Point, Hubbs 47-81; Willows Anchorage, 3809, 6047; Smugglers Cove, 6207; Prisoners Harbor, 4140, 5823; Frys Harbor, 3352; Twin Harbors, 3331. Anacapa Islands: F. H. Elmore. Santa Barbara Island: landing, 4415; southeast reef, 4354. San Nicolas Island: first rocky outcropping west of sand spit, 4441. 46 MADRONO [Vol. 14 EGREGIA LAEVIGATA Setchell subsp. borealis (Setchell) stat. nov. Egre- gia laevigata forma borealis Setchell (Phyc. Bor. Amer. no. XL, 1901). Typical Egregia laevigata occurs from Goleta, California, to Punta San Eugenio, Baja California, Mexico. Plants clearly referable to E. laevt- gata, yet differing slightly but consistently, are to be found in California from Gaviota, Santa Barbara County, northward to Santa Cruz, Santa Cruz County. The rachis is brittle and usually smooth, although at times the terminal blade, or the young rachis, or both, may bear short blunt tubercles. The blades are similar to those of E. laevigata, elliptical to ligu- late, but are usually more widely spaced and never dissected. This series of populations as represented at Carmel was described by Setchell as E. laevigata {. borealis. However, its morphological uniformity and geo- graphical distinctness would seem to warrant its recognition as a sub- species. In an herbarium annotation, Setchell designated as the type of Egregia laevigata a specimen collected by him (1659) at Carmel Bay, Monterey County, California, 17 May 1897 (UC 96743). Inasmuch as this speci- men was collected subsequent to the publication of EF. laevigata, its selec- tion as lectotype is untenable. In the original publication Setchell men- tioned plants from Santa Cruz, Port Harford, and San Pedro, all California localities, but only those from San Pedro, Los Angeles County (Setchell 1157, December 1895) have dissected blades, a character emphasized by Setchell. Therefore, it seems reasonable to propose Setchell 1157 as the lectotype collection, and, of several specimens available, UC 96758 as lectotype specimen. Representative specimens examined. CALIFORNIA. Santa Cruz County: Santa Cruz, C. L. Anderson. Monterey County: Monterey, Setchell 1412 & 3080; Pebble Beach, 1051, 1659, 1799, 3053. San Luis Obispo County: Piedras Blancas, 1376, 5019; Cam- bria, H. L. Mason; Cayucos, 2313; Port Harford, Setchell 1135; Oilport, Reed 57; Pismo Beach, 5293. Santa Barbara County: Point Arguello, 2517; Government Point, 5454, 5520, 5575; Gaviota, 5229. A case might well be argued for recognizing in Egregia only one species comprised of several subspecies. It seems of greater taxonomic usefulness, however, to continue the recognition of two species. Bossiella nom. nov. Bossea Manza (1937a, p. 46). This genus of Corallinaceae unfortunately bears a name that is a later homonym of Bossea in the Geraniaceae, which was proposed by Reichen- bach (1841, p. 201) asa substitute name for Cynosbata (DC.) Rchb. and commemorates J. F. W. Bosse. The following species seem worthy of rec- ognition. Bossiella californica (Decaisne) comb. nov. Amphiroa californica De- caisne (1842, p. 124). As interpreted by Manza (1937b) and Smith (1944), this species is known only from the Monterey Peninsula, Califor- nia. Yendo (1902b) reported it (as Cheilosporum californicum) from Vancouver Island, British Columbia, but the photograph of this collection indicates that most likely Yendo had a sparsely branched specimen of 1957] SILVA: PACIFIC MARINE ALGAE 47 B. corymbifera. The type must be restudied to eliminate the persistent un- certainty as to the identity of this taxon. Bossiella cooperi (Dawson et Silva) comb. nov. Bossea cooperi Dawson et Silva in Dawson (1953, p. 158). Bossiella corymbifera (Manza) comb. nov. Bossea corvmbifera Manza (1937b, p. 562). Heretofore this species has been reported only from the Monterey Peninsula, California (Manza; Smith), and Coos County, Ore- gon (Doty). The following records amplify the known range. CANADA. BritisH CoLtumpBra. Vancouver Island: Point No Point, 16 miles west of Sooke, 7047; Victoria, 1875, R. Middleton. UNITED STATES. WasutncrTon. Whidbey Island, Gardner 918. CaLirorniA. Mendocino County: Mendocino, Brown 754a. Sonoma County: 3 miles north of Fort Ross, L. Miles; Horseshoe Cove, near Bodega Bay, L. Miles. Marin County: Tomales Head, W. Hartman. Bossiella dichotoma (Manza) comb. nov. Bossea dichotoma Manza (1937b, p. 562). This species, previously reported only as far north as Moss Beach, San Mateo County, California, has been found at Shelter Cove, Humboldt County (7000). Bossiella gardneri (Manza) comb. nov. Bossea gardnert Manza (1937b, p. 563). Bossiella insularis (Dawson et Silva) comb. nov. Bossea insularis Daw- son et Silva in Dawson (1953, p. 159). Bossiella interrupta (Manza) comb. nov. Bossea intcrrupta Manza (OS; pr 503): Bossiella ligulata (Dawson) comb. nov. Bossea ligulata Dawson (1953, p. 156). Bossiella orbigniana (Decaisne) comb. nov. Amphiroa orbigniana De- caisne (1842, p. 124). Bossiella pachyclada (Taylor) comb. nov. Bossea pachvclada Taylor (1945, p. 194). Bossiella plumosa (Manza) comb. nov. Bossea plumosa Manza (1937a, p. 46). Heretofore this species has been reported only from San Mateo County (Manza) and the Monterey Peninsula (Smith) in California and from southern Oregon (Doty). The following records amplify the known range. CANADA. BritisH Co_umesta. Vancouver Island: Point No Point, 16 miles west of Sooke, 7048. UNITED STATES. WasuinctTon. Jefferson County: Ruby Beach, 7104. OREGON. Lincoln County: Yachats, 7135. CALIFORNIA. Del Norte County: Cres- cent City, 6902. Humboldt County: Trinidad, 6827. Mendocino County: mouth of Jughandle Creek, 6720. Sonoma County: Shell Beach, 4 miles south of Jenner, 6671; 2 miles north of Bodega Bay, 3496. Marin County: Bolinas, Gardner 1026. San Fran- cisco County: Lands End, Setchell 5747. Santa Cruz County: Santa Cruz, 3551. Mon- terey County: mouth of Malpaso Creek, 3632; Kasler Point, 820, 849; Partington Point, 2288; Lucia, 2817. San Luis Obispo County: Cayucos, 2312. Santa Barbara County: Point Pedernales, 2419; Government Point, 5541. 48 MADRONO. [Vol. 14 Bossiella sagittata (Dawson et Silva) comb. nov. Bossea sagittata Dawson et Silva in Dawson (1953, p. 157). PACHYARTHRON CRETACEUM (Postels et Ruprecht) Manza (1937a, p. 45). This North Pacific species extends southward to Vancouver Island, judging from Yendo’s photograph (1902b, pl. 51, fig. 1) of a plant col+ lected at Port Renfrew (as Amphiroa cretacea f. tasmanica). CALLIARTHRON REGENERANS Manza (1937b, p. 565). The following records outline the known range of this species, which heretofore has been reported only from San Mateo County in California (Manza) and Coos and Curry counties in Oregon (Doty). CANADA. BritisH CoLumpBiA. Vancouver Island: Point No Point, 16 miles west of Sooke, 7045. UNITED STATES. Wasuincton. Whidbey Island, Gardner 83. CALIFORNIA. Del Norte County: Crescent City, 6904. Humboldt County: Shelter Cove, 7002. Mendocino County: mouth of Jughandle Creek, 6717. Sonoma County: Shell Beach, 4 miles south of Jenner, 6670; 2 miles north of Bodega Bay, 3495. Marin County: Bolinas, Gardner 1027. Monterey County: Partington Point, 2278; Lucia, 2837. San Luis Obispo County: Piedras Blancas, 5040; Pismo Beach, 5367. Santa Barbara County: Government Point, 2195, 5467, 5511. San Miguel Island: Cuyler Harbor, 3756. Santa Rosa Island: Sandy Point, 4123; Canada Lobos, 2763; East Point, 4071. Santa Cruz Island: Fraser Point, 5967; Willows Anchorage, 3823, 6039; Smugglers Cove, 6221. Santa Barbara Island: landing, 4405; southeast reef, 4365. MEXICO. Baja cCAtirorniA. Islas Todos Santos: south island, 4829. CALLIARTHRON SCHMITTI Manza (1937b, p. 566). This rare species, heretofore known only from the San Diego region, has been found on lit- toral rocks at Eagle Point, San Juan Island, Washington (Papenfuss & Scagel, 7 July 1952). The fronds are prostrate, as inferred correctly by Manza from fragmentary dredged material. Serraticardia (Yendo) stat. nov. Cheilosporum sect. Serraticardia Yendo (1905, p. 2). Cheilosporum subgen. Serraticardia Yendo (1902c, p. 193, nomen nudum). Type species: Serraticardia maxima (Yendo) comb. nov. Cheilosporum maximum Yendo (1902a, p. 22, pl. 2, figs. 18, 19; pl. 6, fig. 9). Serraticardia macmillanii (Yendo) comb. nov. Cheilosporum mac- millani Yendo (1902b, p. 718, pl. 52, figs. 4, 5; pl. 56, figs. 11-14). Cal- liarthron pinnulatum Manza (1937b, p. 565). This plant, heretofore known only from the original collection made by Yendo in the summer of 1901 at the Minnesota Seaside Station (Port Renfrew), Vancouver Island, has been discovered growing on surf-swept rocks of the lower littoral at two localities in California, namely, three miles north of Fort Ross, Sonoma County (L. Miles, 19 July 1951) and Pescadero Point, Monterey Peninsula (3090, 10 July 1948). The reg- ular distichous opposite branching is suggestive of Corallina, but the conceptacles are borne laterally on the faces of intergenicula. On axial intergenicula the conceptacles are usually paired, one toward each lateral margin. Pinnae, being of approximately the same width as conceptacles, usually bear only one. According to Yendo, conceptacles may also occur terminally on pinnae, but my material does not show terminal concept- 1957] SILVA: PACIFIC MARINE ALGAE 49 acles nor are Yendo’s figures convincing. Conceptacles borne subtermin- ally on the face of a pinna might easily be misinterpreted as terminal. The most closely related species appears to be Cheilosporum maximum Yendo, from Japan, and both species were placed by Yendo in his section Serraticardia (Yendo, 1905, p. 26). Although Manza apparently was not familiar with Cheilosporum macmillani, he (1937b, p. 567) referred C. maximum to Joculator Manza (1937a, p. 47), a genus based on the character that conceptacles may be both lateral and terminal in the same plant. (Cheilosporum as circumscribed by Manza is characterized as hav- ing conceptacles only on the upper margins of intergenicula.) Joculator was merged into Corallina by Dawson (1953, p. 124) on the consideration that lateral conceptacles sometimes occur in Corallina officinalis L. Jocu- lator pinnatifolius Manza, the type of its genus, has a predominance of terminal conceptacles, and in this character as well as in general habit it seems more closely related to Corallina than to either Cheilosporum mac- millanit or C. maximum. In the two latter species lateral conceptacles are the rule rather than the exception, and in habit they seem intermediate between a Corallina with compressed intergenicula (e.g., C. chilensis De- caisne) and a pinnate Bossiella (e.g., B. plumosa). In fact, it appears that we are dealing with a group of species which tend to bridge the gap be- tween Corallina and Bossiella, two genera previously thought to be widely separated. No anatomical differences between these two genera have been demonstrated. The occurrence of species intermediate between two genera need not affect the taxonomic usefulness and validity of recognizing both genera. In the present case Cheilosporum macmillaniu and C. maximum seem sufficiently similar to one another and sufficiently different from both Corallina and Bossiella to warrant the recognition of a third genus. The intergenicula of Serraticardia differ from those of Bosszella in not being expanded to the point of differentiating wings from the midrib. On the other hand, they are expanded more than those of Corallina, resulting in a more closely knit pattern of branching. The type of Calliarthron pinnulatum Manza is a robust but poorly branched specimen of Serraticardia macmillanii from Moss Beach, San Mateo County, California (Manza, 2 January 1935). It is surprising that Manza referred this specimen to Calliarthron, a genus characterized by a unique anatomy of interwoven filaments. PLOCAMIUM OREGONUM Doty (1947, p. 177, pl. 14, fig. B). According to entries in his notebook, N. L. Gardner had prepared a manuscript for this species and had designated as the type a collection from Trinidad, Humboldt County, California (June 1934, Gardner 7837, UC 536732). The following records supplement those cited by Doty (from Curry, Coos, and Lincoln counties in Oregon and Marin and Humboldt counties in California). CANADA. British Cotumstia. Vancouver Island: Point No Point, 16 miles west of Sooke, 7036. UNITED STATES. Catrrornia. Mendocino County: mouth of 50 MADRONO [Vol. 14 Jughandle Creek, 6706. Sonoma County: Shell Beach, 4 miles south of Jenner, 6659; 2 miles north of Bodega Bay, 3510. Department of Botany, University of Illinois, Urbana, Illinois LITERATURE CITED ANDERSON, C. L. 1891. List of California marine algae, with notes. Zoe 2: 217-225. ARESCHOUG, J. E. 1876. De tribus Laminarieis et de Stephanocystide osmundacea (Turn.) Trevis. observationes praecursorias. Bot. Not. 1876: 65-73. . 1883. Observationes phycologicae. Particula quarta. Acta Reg. Soc. Sc. Upsal. ser. III, 12. 23 pp. Dawson, E. Y. 1949. Contributions toward a marine flora of the southern California Channel Islands, I-III. Allan Hancock Found. Publ. Occas. Paper no. 8. 57 pp., incl. 15 pls. . 1953. Marine red algae of Pacific Mexico. Part I. Bangiales to Corallina- ceae subf. Corallinoideae. Allan Hancock Pacific Exped. 17(1) :1-239, incl. 33 pls. DECAISNE, J. 1842. Mémoire sur les Corallines ou Polypiers calciféres. Ann. Sci. Nat. Bot. ser. II, 18: 96-128. Dory, M. S. 1947. The marine algae of Oregon. Part I. Chlorophyta and Phaeophyta. Farlowia 3: 1-65, incl. pls. 1-10. II. Rhodophyta. Jbid. 159-215, incl. pls. 11-14. FarLtow, W. G. 1876. List of the marine algae of the United States. Rept. U. S. Fish Comm. 1875: 691-718. . 1881. Note on Laminariae. Bull. Torrey Club 8: 67-68. . 1899. Three undescribed Californian algae. Erythea 7: 73-76. Fartow, W. G., C. L. ANDERSON, and D. C. Eaton. 1878. Algae Exsiccatae Americae Borealis. Fasc. 3. GarDNER, N. L. 1917. New Pacific coast marine algae. I. Univ. Calif. Publ. Bot. 6: 377-416, incl. pls. 31-35. . 1940. New species of Melanophyceae from the Pacific coast of North Amer- ica. Univ. Calif. Publ. Bot. 19: 267-286, 6 pls. Hervey, A. B. 1881. Sea Mosses. Boston: Cassino. xv, 281 pp., 20 pls. HOLLENBERG, G. J. 1948. Notes on Pacific coast marine algae. Madrofio 9: 155-162. Hooker, J. D. 1846. Flora Antarctica. Part 21. Pp. 445-468. Manzi, A. V. 1937a. The genera of the articulated corallines. Proc. Nat. Acad. 23: 44-48. . 1937b. Some North Pacific species of articulated corallines. Proc. Nat. Acad. 23: 561-567. PrerrE, J. B. L. 1890. Notes botaniques. Sapotacées. [Part 1]. Paris: Klincksieck. pp. 1-36. REICHENBACH, H. G. L. 1841. Repertorium herbarii sive nomenclator generum plan- tarum systematicus ... Dresden & Leipzig: Arnold. xcv, 213, 236 pp. ROSENVINGE, L. K. 1893. Grgnlands Havalger. Meddelelser om Gr¢gnland 3: 765-981, 2 pls. SETCHELL, W. A. 1896. Notes on kelps. Erythea 4: 41-48, 1 pl. SETCHELL, W. A. and N. L. GarpNneEr. 1920. Phycological contributions I. Univ. Calif. Publ. Bot. 7: 279-324, incl. pls. 21-31. . 1925. The marine algae of the Pacific coast of North America. Part III. Melanophyceae. Univ. Calif. Publ. Bot. 8: 383-898, incl. pls. 34-107. SmitH, G. M. 1942. Notes on some brown algae from the Monterey Peninsula, Cali- fornia. Am. Jour. Bot. 29: 645-653, 13 figs. . 1944. Marine algae of the Monterey Peninsula, California. Stanford Uni- versity Press. ix, 622 pp., incl. 98 pls. 1957] GENTRY: YUCCA Sy] Tavior, W. R. 1945. Pacific marine algae of the Allan Hancock Pacific expeditions to the Galapagos Islands. Allan Hancock Pacific Exped. 12: i-iv, 1-528, incl. 100 pls., 3 text-figs. Vickers, A. 1905. Liste des algues marines de la Barbade. Ann. Sci. Nat. Bot. ser. IX, 1: 45-66. Yenpo, K. 1902a. Corallinae verae Japonicae. Jour. Coll. Sc. Tokyo Univ. 16 (article 3). 36 pp., 7 pls. . 1902b. Corallinae verae of Port Renfrew. Minn. Bot. Stud. 2: 711-722, pls. 51-56. . 1902c. Enumeration of corallinaceous algae hitherto known from Japan. Bot. Mag. Tokyo 16: 185-196. . 1905. A revised list of Corallinae. Jour. Coll. Sc. Tokyo Univ. 20 (article - 12). 46 pp. A NEW YUCCA FROM SONORA, MEXICO HowarpD Scott GENTRY Yucca grandiflora sp. nov. Arbor 3—4 m. alta; folia 70-100 cm. longa, 4—5 cm. medio lata, viridia, laevia, margine brunneo sparse-filifero, spina terminali valida basi sulcata, 2—2.5 cm. longa; inflorescentia paniculata, ramis lateralibus dense pubescentibus; perianthium subglobosum, seg- mentis subaequalibus, 7-9 cm. longis, ovatis, mucronatis, glabris; fila- menta pubescentia, basi segmentorum adnata; ovarium elongatum, 4.5—6 cm. longum; stylus breviter 3-lobatus; fructum non vidi. Arborescent, 3-4 m. tall, branching from the base and toward the crown, with deep leaf crowns; leaves 70-100 cm. long, 4—5 cm. wide at mid-blade, slightly narrowed above base, dark green, smooth, ascending to descending, persisting dry and deflected in age on the trunk, the mar- gin narrow, brown, filiferate with long, fine, brittle threads, the terminal spine stout, brown, broadly grooved; inflorescence an irregular open pani- cle 70-100 cm. long; peduncle 10-30 cm. long, glabrate below; bracts and bractlets scarious, dull white, friable; lateral branches densely white- tomentose, flexuous, horizontal; flowers short-pedicellate to subsessile, glabrous, creamy white, divergent on horizontal, openly-spaced lateral branches (fig. 2); perianth 7-9 cm. long, the segments spreading, ovate, thin, bluntly mucronate, connate at base, the outer slightly smaller and thicker than the inner; filaments hyaline-pubescent throughout, the an- thers oblong; pistil slender, 4.5—6 cm. long, deeply sutured, shortly beaked below the lobate stigma; fruits not seen. Type. Above Tierra Negra, Cedros Range, east of Rio Cedros, Sonora, Mexico, February 14, 1952, Gentry 11601 (U. S. Nat. Herb. 2089433 and 2089434). This plant differs from all other known species of Yucca in the large subsessile flowers borne on tomentose lateral branches in an open ragged- appearing panicle. Although the fruits are not available for study, the obvious relationship is with the group having fleshy fruits, the Sarcocar pa. 52 MADRONO [Vol. 14 Fic. 1. Mature clump of Yucca grandiflora showing both basal and crownal branching. Above Tierra Negra on the Cedros Range in southern Sonora. In the rather flexible, long, wide leaves and the pubescent lateral branches of the inflorescence, it resembles Yucca schottw. The large flowers and elongate pistils or ovaries, however, relate it more closely to Y. arizonica, from which it differs by the 1) pubescent lateral branches of the inflores- cence as compared to glabrous branches, 2) subsessile flowers as com- pared to long-pedicellate flowers, 3) thin, finely veined, ovate perianth segments as compared to thicker, coarsely veined, lanceolate perianth seg- ments, and 4) erect or divergent flowers as compared to nodding flowers. In the type locality, where it formed a widely scattered colony, Y. grandiflora was found associated with species of Quercus and Acacia pen- natula in an extensive tract of Oak Woodland having volcanically derived calcareous soils with a grass cover. Elevations here ranged from about 2500 to 3500 feet. A similar-appearing Yucca observed to the southeast of this locality, both along the Arroyo Guajaray and the Rio Mayo in adjacent Chihuahua, may prove eventually to be this species. It is known locally by its Warihio Indian name, “‘sahuiliqui.” This Yucca is one of many collected to determine sapogenin content. The leaves proved to contain 1.4 per cent sarsasapogenin, as reported by 1957] GENTRY: YUCCA 23 ‘on ee Fic. 2. Type material of Yucca grandiflora before pressing. Note the divergent subsessile flowers and the horizontal, openly spaced, lateral branches. the Eastern Regional Research Laboratory at Philadelphia.' Sarsasapo- genin is one of the steroidal compounds regarded as precursor to cortisone and related drugs. The fairly high percentage value of sarsasapogenin, together with the fiber in the numerous leaves, indicates an economic potential for Yucca grandiflora. Plant Introduction Section, Horticultural Crops Research Branch, Agricultural Research Service, United States Department of Agri- culture, Plant Industry Station, Beltsville, Maryland 1 Wall et al. 1955. Steroidal Sapogenins XXVI. Supplementary table of data for steroidal sapogenins XXV. U.S. Dept. Agr., Agr. Res. Serv. Circ. ARS-73-4. 54 MADRONO * [Vol. 14 A REVISION OF THE LUPINUS ARBUSTUS COMPLEX OF THE LAXIFLORI Davip B. DUNN During and prior to the preparation of the manuscript of Lupinus for the Flora of Nevada (Dunn, 1956a), the type specimens of over 500 North American lupines, exclusive of species from Mexico, were exam- ined. For each of these types, notes were taken, photographs were made, and floral parts were dissected and mounted in plastic, as described by Dunn (1954). One of the most important groups of type specimens ex- amined was the Lindley collection of Douglas type specimens, Cambridge University (Dunn, 1956b). As a result of the extensive survey of these type specimens, together with careful study of the collection data recorded by Douglas in his journal (1914), it was deemed necessary to make sev- eral nomenclatural changes. Some of these were published earlier by me (Dunn, 1955) so that the names would be available for use in the Flora of Nevada. The opportunity now presents itself for discussion of the problems centering around L. arbustus and closely related taxa, and for a full explanation of the recombinations pertinent to this complex. In studying the Douglas type material from the Lindley Herbarium (Dunn, 1956b), I came to the conclusion that Lindley’s original descrip- tion and illustration of L. laxiflorus (Bot. Reg. t. 1140, 1828) did not match the specimen labeled L. laxiflorus (Douglas 297, fig. 1). Instead, his description and illustration match the specimen labeled L. tenellus (Douglas 277, fig. 2). It is only possible to conjecture upon how this error came about. Whether the mixing of the two names occurred as an error by Douglas in collecting ripe seed at a later date than his herbarium sample or as an error when the seeds were planted, it is now impossible to say. In my opinion, however, the description and illustration are critical and must go with the specimen they fit rather than with the specimen which bears the cogent name, particularly for early material collected prior to the adoption of the type concept. (See Appendix 1, paragraph 4a, and Articles 19, 21; International Code, 1952). This interpretation, however, has resulted in the necessity for retypifying L. laxiflorus, and for bringing this name (L. laxiflorus Doug]. ex Lindl., not Agardh) into synonymy under L. argenteus Pursh. subsp. argenteus var. tenellus (Dougl. ex G. Don) Dunn, since the taxon involved is considered as below specific rank. The reason for the use of the varietal name tenellus is simply prior- ity within a given rank (Articles 16, 66, 67, and 70; International Code, 1952). Fortunately L. argenteus was published by Pursh in 1814 so the name L. argenteus was not upset by recognizing that L. laxiflorus was a part of that species. Torrey and Gray were the first to apply a name below the rank of species to the taxon tenellus (see number 1 in taxonomic treat- ment). The “y” used by Torrey and Gray as a prefix to the trinomial 1957] DUNN: LUPINUS ARBUSTUS 29 Fic. 1 (left). Photegraph of the specimen in the Lindley Herbarium bearing the label L. laxiforus: I have designated this specimen as the type of L. arbustus subsp. neolaxiflorus Dunn. It does not match Lindley’s illustration or description in several important characters (see text). Fic. 2 (right). Photograph of the specimen in the Lindley Herbarium bearing the label L. tenellus. This specimen matches Lindley’s description and illustration of L. laxiflorus quite closely, with minor deviations (see text). This taxon is treated as L. argenteus subsp. argenteus var. tenellus (Doug. in G. Don) Dunn. name tenellus has generally been conceded to mean a variety since it was not the practice of the time to designate subspecies. This retypification of L. laxiflorus necessitates the redesignation of the taxon classically known as L. laxiflorus to subsp. neolaxiflorus (Dunn, 1955) and placing it under L. arbustus Dougl. ex Lindl. (Bot. Reg. t. 1230, 1829) which is the next specific epithet in line, thus retaining as much continuity of the name as possible. The name /axiflorus, then, is no longer spread over five other subspecific taxa, but is restricted to subsp. neo- laxiflorus specifically, the taxon to which Agardh and botanists since then have been applying the name /axiflorus in the strict sense and not in the broad sense employed in current manuals. The characters of most fundamental importance in interpreting the specimen labeled L. tenellus to be the taxon described by Lindley as L. laxiflorus are: 1) the short petioles throughout, about as long as the leaf- 56 MADRONO [Vol. 14 Fic. 3. Chart of the floral parts of the taxa within Lupinus arbustus, solid lines drawn to the typical shape and uniform scale of the mean measurements of 25 indi- viduals (numbers 2b, etc., refer to taxa in text and on Map 1). Dotted lines represent parts beneath the surface, folds under the outer surface or, as in the banner and wings, the range of shape. The horizontal rows from bottom to top represent 1) left side view of entire flower; 2) flattened dorsal view of the banner, the two halves showing the range in the amount of pubescence; 3) left wing; 4) right side view of keel; 1957] DUNN: LUPINUS ARBUSTUS =I7/ lets (fig. 2) and illustrated by the artist in Edwards’ Botanical Register (t. 1140, 1828); 2) the leaves appearing fasciculate due to the short peti- oles and to the occurrence of several leaves on the dwarf axillary branches present at most of the nodes at anthesis of the primary racemes; 3) leaf- lets linear to linear-elliptic-oblanceolate as illustrated; 4) the slender purplish stems referred to by Lindley; the specimen labeled tenellus is distinctly purplish and slender by contrast with the specimen labeled laxiflorus; 5) the obcordate shape of the banner described and illustrated; and 6) the floral parts (plastic-coated dissections) match the illustration, t. 1140. Lindley’s statements referring to an absence of bracteoles (in his de- scription of L. laxiflorus) and a beardless keel are both erroneous since both of the specimens labeled L. tenellus and L. laxiflorus have bracteoles and ciliate keels, although the former has the least ciliation. Both also have a spur about 0.5 mm. long at the base of the upper lip of the calyx. The shape and conformation of the floral parts (fig. 3c, solid lines) were drawn from Douglas 297 (fig. 1), which is considered typical for L. arbustus subsp. neolaxiflorus. The extent to which the flowers differ from the illustration in Edwards’ Botanical Register (t. 1140) should be apparent. It should be noted that more of the characters which indicate that the specimen labeled L. tenellus is the taxon described as L. laxiflorus are to be found in the illustration than are found in the description. I dare say this situation is not uncommon among taxa described in either Edwards’ Botanical Register or in the Curtis’ Botanical Magazine. That the pur- pose of an illustration is to clarify characters difficult to describe in words is recognized by the Recommendation 54F of Article 54, International Code (1952), where illustrations are recommended. I would contend that where such are provided with the original description that the illustra- tions are an integral part of the description. The extent to which illustra- tions are recognized by the Code is noted in Article 21 in which it is stated that an illustration or a description may become the type for a species without a type specimen. It should further be pointed out that the speci- mens in the Lindley Herbarium considered as types are more technically lectotypes. Thus, Article 19 is involved regardless of who wrote the name L. laxiflorus on the specimen (fig. 1). Note again the wording in the In- ternational Code (Appendix 1, paragraph 4a, Determination of types) ; wherein “recognizable figures” are specifically acknowledged as being a means of determining what material has been described. My interpretation, previously published in outline (Dunn 1955, 1956b), has been questioned by Phillips (1955, p. 196), whose paper came out 5) inside view of the unfolded calyx cup, slit from the left sinus to the pedicel. Since the floral parts for 2a and 26 are very similar in shape, only the wing is illustrated for 2a (within the 2b series) ; the wing of 2b is glabrous both laterally and margin- ally near the claw (not illustrated). Subspecific entities are indicated by the key figures: 2a—2f, 58 MADRONO [Vol. 14 2a. Lupinus arbustus subsp. arbustus var. arbustus A 2b. Lupinus arbustus subsp, arbustus vor. montanus @Q 2c, Lupinus arbustus subsp. neolaxiflorus @ 2d. Lupinus arbustus subsp. silvicola 2e. Lupinus arbustus subsp.calcaratus + (timate 2f. Lupinus arbustus subsp. pseudoparviflorus (€ ------ Intermediates or hybrids 9@¢}t™ Fic. 4. Distribution of the taxa within Lupinus arbustus. 1957] DUNN: LUPINUS ARBUSTUS 59 after I had submitted this present manuscript for publication in June, 1955. Phillips admits that Lindley’s description of L. /axiflorus does not fit the specimen labeled L. laxiflorus, but he believes the latter to be la- beled in Lindley’s handwriting, whereas I am convinced it is in Agardh’s. Phillips borrowed only the types designated as L. tenellus and L. laxt- florus. Phillips fails, however, to recognize the significance of the coinci- dence of morphological characters in Lindley’s description and illustration of L. laxiflorus with the characters shown in the specimen labeled L. ten- ellus, nor does he recognize the significance of the characters which differ- entiate the specimen labeled L. tenellus from the specimen labeled L. laxiflorus. Phillips, in other words, relies on what has been presumed to be the type specimen irrespective of the facts that the authorship of the hand- writing thereon is uncertain and that neither the type description nor the illustration fits this presumed type specimen. It should be emphasized that the type concept was not accepted in Europe until the 1930’s, after which time type labels were placed on specimens by current workers; sometimes sheets with type material were cut apart and the parts mounted on separate sheets; and, in one case of which I know, the type label was transferred to the wrong half of the sheet. There has been too much chance for error in the processing and reprocessing of the early specimens. With this early material it is not the specimen considered as the type it- self, but the description and illustration which are critical, as noted re- peatedly in the International Code (1952, Articles 19, 21, and Appendix 1). In the case of the lupines under consideration, I have made a consid- erable effort, as already stated, to clarify these matters and to coordinate the type localities as given in Douglas’ Journal (Dunn, 1956b) with the type specimens in the Lindley Herbarium and with their published de- scriptions. The taxonomic problems involved with the treatmeént of lupines have been multiplied considerably since settlement of western North America. Extensive cattle grazing undoubtedly created numerous openings, and in more recent years the great increase of roads has greatly multiplied the disturbed areas. Lupines are early invaders in areas very low in soil nitro- gen such as the exposed subsoils of road cuts or areas vacated by receding glaciers (Lawrence, verbal communication, 1950). Road cuts are com- monly inhabited by lupines, a site inadequate for most plants lacking a means of nitrogen fixation. Several species, normally perennial but ca- pable of maturing seed in one season, are now commonly found scattered through grainfields. Such extensions of suitable habitat have undoubtedly resulted in many range extensions during the past hundred years, thus bringing taxa into contact which may have appeared quite distinct in Douglas’ time, but which had not been segregated sufficiently long to de- velop adequate barriers to hybridization. The study of the taxa of Lupinus encompassed in the Flora of Nevada has revealed that there are occasional morphological intermediates between many of the closely related taxa. 60 MADRONO [ Vol. 14 Many of these intermediates have been collected since settlers and mech- anization have come to the country. While it is readily admitted that col- lecting conditions became simpler with the advent of roads, accounting for increased knowledge of the genus, another explanation must also be considered as possible, if not probable, for the occurrence of morphologi- cal intermediates. This is Edgar Anderson’s theory of introgressive hy- bridization (1949), which appears to fit the situation well. His explanation of the survival of hybrid swarms gives a plausible reason for the existence of these intermediates, although it is questionable if they do more than survive long enough to permit some gene flow between what are otherwise well-established taxa with genomes that have been selected by the envi- ronment over a long period of time. The question arises as to how to deal taxonomically with these intermediates. If all the taxa which have inter- mediates between them were placed in one species, the genus Lupinus would be reduced to a handful of species, but the keys to separate the subspecific taxa would indeed be complex. I believe it is far better to treat many of the taxa as ecospecies in the sense advanced by Anderson. At this point I wish to express my appreciation to the officials of the herbaria cited in the text below for the loan of the material and especially to Dr. P. A. Munz for the use of the facilities at the Rancho Santa Ana Botanic Garden where most of the work on this paper was done. KEY TO THE TAXA Petioles all short, 2-4 cm. long, the lower ones commonly not much longer than the leaflets, occasionally up to 8 cm. long; leaves cauline; banner obcordate to orbi- cular, sparsely pubescent in the dorsal grooves; the wings glabrous; calyx spur generally wanting, short but present in L. argenteus subsp. argenteus var. ten- ellus. (Lupinus argenteus and allies; not treated here except for the nomenclature of the. .var, Lenellirs) oe 1. L. argenteus subsp. argenteus var. tenellus Petioles gradated, the lower ones 8-15 cm. long, to 2-4 cm. long on the uppermost leaf; commonly mostly basal leaves but members of some taxa with the leaves all cauline, then with the longest lower petioles over 10 cm. long; banner obovate to orbicular; calyx spur generally well-developed, except in L. sulphureus and allies. Banner glabrous (a few individuals with some pubescence on the back); wings glabrous; the base of the calyx enlarged, gibbous above, not spurred. hts Ae Ss aN ODS Penske eel te eee L. sulphureus and allies, not treated here. Banner generally abundantly pubescent over the central area of the back, some- times only sparsely pubescent in the grooves or sometimes glabrous; wings generally with lateral pubescence near the tip but sometimes glabrous in mem- bers of two taxa; calyx spur generally well-developed. Leaflets glabrous above, or glabrate, oblanceolate, 6-10 mm. wide, banner and wings pubescent or often glabrous; in damp habitats; northeastern Wash- ington and the Rocky Mountains......2f. L. arbustus subsp. pseudo parviflorus Leaflets pubescent above, sparsely so in some, generally linear-elliptic, to oblan- ceolate, banner and wings pubescent, or often glabrous in subsp. neolaxi- florus; California, Nevada, Oregon, Washington and the southern half of Idaho. Flowers 6.5-9.5 mm. long, excluding the spur. 1957] DUNN: LUPINUS ARBUSTUS 61 Plants 2—4 dm. tall, stems clumped, generally with several long-petioled (8-12 cm.) basal leaves; wings and banner glabrous or pubescent ; Wenatchee Mountains, Washington, south-eastward into the Snake River drainage of southern Idaho (specimens from the southeastern part of the range exhibit a reduction in the number of basal leaves and an increase in pubescence and height of plant). SNL, ee ePID Sd SS Ne SRE 2c. L. arbustus subsp. neolaxiflorus Plants 4-5 dm. tall, stems few, with few or no basal leaves, the leaves oblanceolate or narrowly so, sparsely pubescent above; wings pubes- cent at the tip; spur somewhat slender; banner orbicular, pubescent in the dorsal grooves; northern Sierra Nevada, California, and Cas- cade Mountains, Oregon........................ 2d. L. arbustus subsp. silvicola Flowers 9.0-14 mm. long, excluding the spur; banner and wings pubescent, the banner rarely glabrous. Flowers 11-14 mm. long, generally white or light yellow to lavender (see also var. montanus), intermediates bright purplish blue; calyx spur slender, 1-3 mm. long; upper lip generally 6-7 mm. long; wings and flower slender, viewed laterally; Great Basin. 2 eae eet cet ee pene TE Aen ee 2e. L. arbustus subsp. calcaratus Flowers 9-12 mm. long; purplish blue to pinkish lavender, drying blue; calyx spur blunt and broader, 1-1.5 mm. long; upper lip 5 mm. long; wings 4-5 mm. wide, the flowers appearing wider; mountains of northeastern Oregon and southward into the southern Sierra Nevada, California. Wings with pubescence laterally near the apex only. ee eee Se en 2b. L. arbustus subsp. arbustus var. montanus Wings with pubescence as above but also with cilia above or below the claws ocr with lateral villi on the veins near the claws, or with hairs at all three locations......2a. L. arbustus subsp. arbustus var. arbustus 1. LUPINUS ARGENTEUS Pursh subsp. ARGENTEUS var. TENELLUS (Dougl. in G. Don) D. Dunn, Leafl. West. Bot. 7:254. 1955. (G. Don cites Dougl. mss. but the description is not a copy of anything in Douglas’ Journal; some other mss?). Type. Douglas 277, 31 May 1825 (CGE), p. 125 in Douglas’ Journal, collected from the vicinity of the Grand Rap- ids of the Columbia River, Washington or Oregon (+40 miles ‘‘?”’, Doug- las’ hiking range). Lupinus laxiflorus Dougl. ex Lindl. not Agardh, Bot. Reg. t. 1140. 1828. Lupinus tenellus Doug). in G. Don, Gen. Hist. Dichl. Pl. 2:367. 1832. Lupinus laxtflorus var. (“y”) tenellus (Dougl. in G. Don) Torr. & Gray, Fl. N. Am. 1:377. 1840 (L. foliosus var. (“3”) stenophyllus Nutt. mss. zbid. in syn.). Lupinus stenophvllus Nutt. ex Rydb. Bull. Tor- rey Club 34:42. 1907. Lupinus argenteus var. stenophyllus (Nutt. ex Rydb.) Davis, Flora Idaho, 492. 1952. Lupinus lanatocarinatus C. P. Smith, Sp. Lup. 317. 1942 (an intermediate with L. caudatus Kell.). Type. East of Fort Hall, Bingham County, Idaho, Davis 137-35 (IDS). Lupinus fremontensis C. P. Smith, Sp. Lup. 320. 1942. Type. Sand dunes 6 miles northwest of St. Anthony, Fremont County, Idaho, Davis 326 (DS, isotype at IDS). Lupinus edward-palmeri C. P. Smith, Sp. Lup. 572. 1946. Type. Big Butte Station, Idaho, Palmer 558 (US). Lupinus carci- formes C. P. Smith, Sp Lup. 574. 1946. Type. Four miles south of Macks Inn, Fremont County, Idaho, Christ & Ward 14899 (DS, isotype at NY). 62 MADRONO [Vol. 14 Lupinus hullianus C. P. Smith, Sp. Lup. 573. 1946. Type. From a burn, Clark County, Idaho, Hull 235 (USFS 91141, co-type DS-Sm.).1 Lupi- nus montis-cooku C. P. Smith, Sp. Lup. 726. 1952. Type. Cook Mountain, Clearwater Forest, Idaho, Sutton 70 (USFS 38580). A detailed description is not presented here, since it is considered that such properly belongs in a monographic treatment of the ‘“‘argenti’’ which will be presented at a later date. The name tenedlus has priority at the rank of variety. The name stenophvllus may require recognition at some rank below the level of species when the type specimen is located. It is not possible to separate the two as Rydberg did, since the type of tenellus has a distinct but short spur, which Rydberg used to characterize steno- pAyllus. Hence, the treatment of var. stenophyllus as a synonym of var. tenellus as given in Torrey and Gray’s Flora of North America appears correct. The short spur, however, led Torrey and Gray to place the variety tenellus in the species L. laxiflorus sensu Agardh. The taxon tenellus (L. laxiflorus of Lindl.) belongs in the species argenteus, as pointed out by Davis (FI. Idaho, 1952), because of the shape of the banner. There is also a light patch of ciliation on the back of the banner under the lip of the calyx, a character which occurs consistently throughout the L. argenteus complex (except in L. rubricaulis). The pubescence on the back of the banner in the L. arbustus group ranges from none to an extensive area on the back. 2a. Lupinus arbustus Doug]. ex Lindl. Bot. Reg. t. 1230. 1829 (subsp. arbustus var. arbustus). Lindley wrote “... local .. . gravely soils in North California, ... common near Fort Vancouver ..”’ This statement is in error. Douglas went to California on December 22, 1830. The notes in Douglas’ Journal (1914) matching this taxon place the type region near Falls of the Columbia (Celilo Falls, Klickitat County, Washington or Wasco County, Oregon), Douglas 296, 20 June 1825 (CGE). Lupinus laxiflorus var. arbustus (Dougl. in Lindl.) M. E. Jones, Contrib. West. Botany 14:33. 1912. Lupinus laxiflorus var. laxiflorus sensu Phillips pro parte, Res. St. State Col. Wash. 23:196. 1955. Plants 3—6 dm. tall; dry stems 3—4 mm. in diameter, finely sericeous to densely subappressed-sericeous; lower petioles 8—13.5 cm. long, present at flowering or the leaflets fallen; leaflets of the largest leaves 8-12, 4—5.5 cm. long, 5-8 mm. wide, linear-oblanceolate, tending to be conduplicate, arcuate, the tips acute; peduncles 3—5 cm. long; flowers 10-12 mm. long; base of the upper calyx lip developed into a distinct spur 1—1.5 mm. long; banner reflexed above the midpoint, pubescent over the central area of the back and pubescent in the ventral sulcus near the umbos; wings with lateral pubescence near the tip and also with lateral villi on the veins near the claws and ciliation on the edges above and below the claws; keel 1DS-Sm.=C. P. Smith Herbarium in the Dudley Herbarium, Stanford Univer- sity. All other abbreviations of herbaria follow Lanjouw & Stafleu, Index Herbari- orumy 2.1952. 1957 | DUNN: LUPINUS ARBUSTUS 63 densely ciliate near the acumen and sparsely so above the claws, also with lateral villi near the claws; ovules 4-5; mature pods not seen. The variety arbustus in its typical form appears to be of rare occur- rence. Unfortunately the specific name was applied by Douglas to a plant (Douglas 296), which appears to be of hybrid origin or perhaps ancestral to, or at least lying between, L. caudatus Kell. and the rest of the complex treated here as L. arbustus, of which L. arbustus subsp. neolaxiflorus Dunn represents an extreme. The floral characters of the type specimen, Douglas 296, in the Lindley Herbarium combine traits of L. caudatus (which has lateral villi near the claws of the wings and keel and marginal cilia above and below the claws of both the wings and keel, the hairs occa- sionally absent at two or three of the six locations) with those of the vari- ous subspecies treated here under L. arbustus (which have a patch of pubescence laterally near the tip of the wings, commonly absent in a per- centage of two subspecies). The vegetative affinity and nature of the pu- bescence of the plant are both very close to var. montanus which grades into subsp. neolaxiflorus in one direction and subsp. calcaratus in the other direction. The specimens of more recent collection which match the specimen Douglas 296 are from what appears to be a local endemic population in the mountains of Klickitat County, Washington. This suggests that Douglas collected both 296 and 297 in that area, but it is possible that a similar population may exist or have existed on the south side of the Columbia River in Wasco County, Oregon. The taxon L. caudatus was observed to be sympatric with L. arbustus subsp. calcaratus (Kell.) Dunn, but there appears to be a barrier to hy- bridization since no intermediates were observed in herbarium material. Hence L. caudatus was retained in its present status, rather than reducing it to a subspecies of L. arbustus. Lupinus caudatus appears to hybridize with several of the other subspecific taxa within L. arbustus, however, and thus may form one end of an overlapping ring of subspecies as discussed by Goldschmidt (1952, p. 90). Since the present evidence suggests that L. caudatus may be an ecospecies, it was considered that breeding studies would be necessary before reaching a decision involving any change in status of L. caudatus. Judging from the number of morphologically inter- mediate herbarium specimens seen, L. caudatus also appears to hybridize with several other less closely related taxa than those mentioned above. Some anomalies have resulted from what appears to have been hybri- dization between L. caudatus and L. arbustus subsp. silvicola and second: ly, L. caudatus and L. arbustus subsp. neolaxiflorus, such anomalies re- sembling L. arbustus subsp. arbustus var. arbustus, but being well beyond the range of the typical material as well as showing vegetative affinities with the suggested parents. Five such specimens are: 1) Congdon (MIN), July 25, 1891, Independence Mountains of Cassia County, Idaho; 2) Christ 14486 (NY), from the same place; 3) Lyle (DS), August, 1930, 7 miles south of Lick Ranger Station, North Fork, Wallowa River, Wal- 64 MADRONO [ Vol. 14 lowa County, Oregon; 4) Congdon (MIN), near Inspiration Point, Yo- semite, Mariposa County, California; 5) Shoop 94 (UMO), Majic Dam, Hailey, Blaine County, Idaho. Distribution. WasHINGTON. Klickitat County: vicinity of the Falls of the Colum- bia, Douglas 296 (see type citation); hillsides at top of grade from Roosevelt to Bickelton (flowers deep wine-red), Pickett, McMurray, & Dillon 1441 (WS) ; west- ern Klickitat County (near Bingen?), Suksdorf 39 (WS); Columbus, Suksdorf 1792 (WS). 2b. LUPINUS ARBUSTUS subsp. ARBUSTUS var. MONTANUS (Howell) Dunn, Leafl. West. Bot. 7:254. 1955. Lupinus laxiflorus var. montanus Howell, Erythea 3:33. 1895. Type. Mount Hood, Oregon, T. Howell 1494 (iso- type, UC, MO). Lupinus laxiflorus var. cognatus C. P. Smith in Jepson, Man. FI. Pl. Calif. 527. 1925. Type. Wallowa Mountains, Oregon, Cusick 3187 (DS). Lupinus lutescens C. P. Smith, Sp. Lup. 235. 1940. Type. Badger Mountain, Douglas County, Washington, Thompson 14626 (WTU). Lupinus proteanus Eastw. Leafl. West. Bot. 4:190. 1945. Type. Emigrant Pass, Nevada, Eastwood & Howell 231 (CAS). Lupinus per- confertus C. P. Smith, Sp. Lup. 738. 1952. Type. Lemhi Nat. Forest, Horse Heaven Pass, Idaho, Cusick 59 (USFS 56696). Lupinus laxiflorus var. laxiflorus sensu Phillips pro parte, Res. St. State Col. Wash. 23:197. 1955. Plants 3—6 dm. tall; dry stems 2.5—3.5 mm. in diameter, clustered, from a woody caudex, sericeous; longest lower petioles 9-16 cm. long, present at anthesis; leaflets of largest leaves 9-13, 3-4.5 cm. long, 5—7 mm. wide, linear-elliptic to linear-oblanceolate, densely or finely sericeous on both sides, tending to be conduplicate, arcuate, on drying; peduncles 3—5 cm. long; flowers 10-12 mm. long; spur at base of calyx 1—-1.5 mm. long; ban- ner reflexed above the midpoint, pubescent over the central area of the back and in the ventral sulcus near the umbos; wings with lateral pubes- cence near the tip but glabrous on the basal half; keel ciliate along the upper margins but otherwise glabrous; ovules 4-5; pods 20-30 mm. long, 6—8 mm. wide, sericeous to villous. The variety montanus is morphologically very close to the variety arbustus. The variety montanus is the more common of the two, and the vegetative appearance is more nearly that of a robust subsp. neolaxi- florus. The var. montanus is found on rocky or gravelly montane slopes with yellow pines or at higher elevations. The smallest specimens of var. montanus are from Kittitas County, Washington, where the vegetative characters resemble those of subsp. neolaxiflorus, but the flower confor- mation is typical of var. montanus. The specimens cited from Douglas and Chelan counties, Washington (L. lutescens C. P. Sm.), morphologically belong with var. montanus but apparently there has been introgression from L. sulphureus. The white color makes this population look very much like subsp. calcaratus, which blends into var. montanus across the northern part of Oregon. Distribution. (partial citation, see Map 1). CaLirornta. Kern County: Shirley Creek, Greenhorn Mountains, Griffith 464 (POM). Los Angeles County: Swartout 1957] DUNN: LUPINUS ARBUSTUS 65 Valley, San Gabriel Mountains, Munz 4607 (POM). Mariposa County: 8 miles west of Glacier Point, Wiggins 9224 (POM). Siskiyou County: north of Weed, June 12, 1940, Lewis (LA). IpaHo. Adams County: Smith Mountain, Davis 2539 (WS). Nez Perce County: Culdesac, May 20, 1932, Warren (WS). County ?: Craig Mountains, Sandberg, Mac- Dougal & Heller 226 in part (POM). ‘ Nevapa. Douglas County: Spooner, June 23, 1902, Baker (POM). Esmeralda County: Emigrant Pass, Eastwood & Howell 231 (CAS). OrEeGoN. Baker County: Cornucopia, slopes of Wallowa Mountains, Thompson 13343 (RSA, WS). Powder River Mountains, August, 1896, Piper (WS). Benton County: near Corvallis, June 19, 1898, Kincaid (WS). Clackamas County: (both the following are probably hybrids with subsp. szlvicola) opposite Oswego, May 24, 1888, Millard (WS) ; Oswego, May, 1889, Drake & Dickson (WS). Gilliam County: 3 miles south of Olex, Hitchcock 19218 (RSA & WS). Jefferson County: Cove Palisade, 13 miles southwest of Madras, Cronquist 6942 (RSA). Morrow County: between Spray and Hardman, Blue Mountains, Cronquist 6618 (RSA). Multnomah County: Wil- lamette River, below Portland, Skeldon S10858 (POM, MO, WS). Sherman County: DeMoss, Hill 7; 49 (WS). Union County: source of Two Color Creek, Wallowa Mountains, Cusick 3187 (MO, WS), 3668 (WS), 3690 (RSA, WS) ; vicinity of Union, Cusick 3714 (WS). Wallowa County: Lookout Mountain, Hell’s Canyon, Constance & Jacobs 1424 (WS). Wasco County: 5 miles east of Oregon Skyline Trail, Highway 50, Martin 4815 (RSA, LA, MO, & WS) (an intermediate with subsp. silvicola) ; 25 miles south of Maupin, Peck 26162 (WS); The Dalles, Suksdorf 1959, 1961, 1962, 1968, 2185 (WS). WasHINGTON. Asotin County: south of Puffers Butte, Cronquist 5818 (RSA & WS) (probable hybrid). Chelan County: Wenatchee Mountains, Hitchcock 17285 (RSA & WS), Griffiths & Cotton 127 (WS). Wenatchee, Whited 41, 1103 (WS) ; 25 miles southeast of Wenatchee, Pickett 1271 (WS). Douglas County: slopes of Badger Mountain, Thompson 14626 (WTU). Kittitas County: Table Mountain, Thompson 14887a in part (RSA); Bald Mountain, Thompson 14806 (RSA). Klickitat County: Columbus, Suksdorf 6498 (MO, WS), 6512 (WS); Rockland, Klickitat Mountains, Suksdorf, June 12, 1893 (WS), 2306 (MO). Yakima County: Yakima Indian Res- ervation, Medicine Valley, Heidenreich 94 (WS). 2c. LUPINUS ARBUSTUS subsp. NEOLAXIFLORUS Dunn, Leafl. West. Bot. 7: 254. 1955. Type. Vicinity of the Falls of the Columbia River (Celilo Falls, probably in Klickitat County, Washington, but also possibly on the south side of Columbia River, Wasco County, Oregon, perhaps as much as 20—40 miles inland, Douglas’ hiking range), Douglas 297, 20 June 1825 (CGE). Lupinus laxiflorus sensu Agardh, pro parte, Syn. Gen. Lup. 1835 (not L. laxiflorus Dougl. ex. Lindl.). Lupinus laxtflorus sensu Torrey and Gray, pro parte, Fl. North America, 1840, likewise for the other authors covering western North America (1840-1951). Lupinus invoensis var. demissus C. P. Smith, Bull. Torrey Club 51:304. 1924. Type. Wallowa Mountains, Baker County, Oregon, Peck 5329 (WILL. U.; cotype, DS.). Lupinus caudatus var. submanens C. P. Smith, Sp. Lup. 106. 1939. Type. Antone Creek, 2 miles east of Anthony Lake, Wallowa County, Oregon, August 10, 1930, Lyle (DS). Lupinus lyleianus C. P. Smith, Sp. Lup. 107. 1939. Type. Seven miles east of Pearson Ranger Sta- tion, Umatilla National Forest, Oregon, July 14, 1930, Lwle (DS) (an intermediate with L. caudatus). Lupinus vakimensis C. P. Smith, Sp. Lup. 238. 1940. Type. Cleman Mountains in alpine sagebrush, 25 miles north- 66 MADRONO [Vol. 14 west of Yakima, Washington, Thompson 14572 (WTU). Lupinus we- natchensis Eastw., Leafl. West. Bot. 3:174. 1942. Type. Alpine slopes of Wenatchee Mountains, Thompson 14242 (CAS). Lupinus henrysmithu C. P. Smith, Sp. Lup. 566. 1946. Type. Minidoka National Forest, Idaho, H. L. Smith 119 (USFS 42539). Lupinus amniculi-putoru C. P. Smith, Sp. Lup. 575. 1946 (an intermediate with L. caudatus Kell.). Type. Mink Creek, Bannock County, Idaho, July 30, 1935, Crane (DS). Lupinus mackew C. P. Smith, Sp. Lup. 725, 1952. Type. Clifty Block Mountain Range, Kaniksu National Forest, Idaho, Mackey 65 (USFS 47532). Lupinus augusti C. P. Smith, Sp. Lup. 733. 1952 (an intermediate with L. caudatus Kell.). Type. Head of Slater Creek, Boise National For- est, Elmore County, Idaho, Pearce 23 (USFS 64136). Lupinus stipaphilus C. P. Smith, Sp. Lup. 733. 1952. Type. North Star Lake, Boise National Forest, Elmore County, Idaho, Pearce 165b (USFS 67739). Lupinus festucasocius C. P. Smith, Sp. Lup. 738. 1952 (an intermediate with L. caudatus Kell.). Type. Copper Basin Potholes, Lemhi National Forest, Idaho, Johnson 20 (USFS 56432). Lupinus stocku C. P. Smith, Sp. Lup. 743. 1952. Type. Bostetter Ranger Station, Minidoka National Forest, Idaho, Stock 186 (USFS 33836). Lupinus standingi C. P. Smith, Sp. Lup. 749. 1952. Type. Deep Creek, near Malad, Cache National Forest, Idaho, Standing 18 (USFS 44296). Lupinus laxiflorus var. laxiflorus sensu Phil- lips pro parte, Res. St. State Col. Wash. 23:197. 1955. Plants 2—4 dm. tall, the stems clumped from a woody caudex, the upper nodes branching later, finely sericeous throughout, the petioles of the basal leaves 8-13 cm. long, slender, present at anthesis; leaflets 8-10, linear-elliptic to linear-oblanceolate, the largest 2.5—5 cm. long, 3—7 mm. wide, pubescent on both sides, the tips acute; peduncles 2—8 cm. long; racemes 7—12 cm. long, lax or dense; bracts subpersistent to caducous; verticels 8-20 mm. distant; pedicels 3-4 mm. long; flowers 8-10 mm. long; upper lip of the calyx 3—4.5 mm. long, with a gibbous base or a short spur 0.2—1.4 mm. long at the base, the lip exposed or partially cov- ered by the sides of the banner, with bracteoles 0.2—-1 mm. long; banner obovate to suborbicular, sparsely pubescent in the dorsal grooves and under the calyx lip, or glabrous in about one-fourth of those studied; wings pubescent laterally near the tip in about one-third of those seen, more commonly glabrous; keel minutely and sparsely ciliate along the upper edges; ovules 3-5; pods 24-28 mm. long, 6-7 mm. wide, silky- sericeous. Included in L. arbustus subsp. neolaxiflorus is the main mass of mate- rial considered by authors since Agardh as L. laxiflorus. Its center of dis- tribution is in the mountains of central Washington, extending southward into the northern edge of Oregon and eastward in the Snake River drain- age of southern Idaho. It is found in meadows and gravelly valleys and in rolling hills with Artemisia tridentata and Tetradymia, and on up to rocky slopes in the Ponderosa Pine zone. Some specimens appearing to be hy- brids with subsp. pseudoparviflorus were collected among aspens and lodgepole pines. 1957] DUNN: LUPINUS ARBUSTUS 67 The smaller vegetative habit and the smaller flowers, as well as the fact that the flowers commonly have glabrous wing tips, less pubescence on the back of the banner, and sparse ciliation on the upper edges of the keel, all suggest a close relationship with L. lepidus. The vegetative stat- ure of two specimens from Kittitas County, Washington (Thompson 14806, 14887, cited under var. montanus) and the wing pubescence on the wing tips of some of the individuals from the Wenatchee Mountains (Hitchcock 17285), both suggest gene flow between L. arbustus var. montanus and subsp. neolaxiflorus. Distribution (partial citation). InpaHo. Blaine County: Corral Creek, 15 miles up Morgan Creek, Hitchcock 14125 (RSA). Smoky Mountains, Macbride & Payson 3754 (POM). Cassia County: 22 miles east of Rogerson, Goose Creek Mountains Division, Christ 18490 (NY). Custer County: Salmon River Mountains near Bonanza, Macbride & Payson 3392 (POM); 28 miles southeast of Patterson, Christ 17809 (NY). Franklin County: Franklin Basin, head of Cub River, Christ 16366 (NY). Gooding County: Gooding, Shoop 117 (UMO) (with some traits of subsp. pseudo- parviflorus). Idaho County: Heaven’s Gate, Seven Devil Mountains, Q. Jones 215 (RSA). Jerome County: 14 miles west of Eden, Christ 15420 (NY). Power County: Crystal, 19 miles southwest of Pocatello, Christ 18559 (NY, an intermediate with L. caudatus). Twin Falls County: 12 miles east of Rogerson, Christ 18485 (NY). Washington County: 17 miles northwest Mann Creek Store at 4th of July Creek, Christ 17943 (NY). OrEGON. Baker County: Eagle Creek, Wallowa Mountains, Cusick 2331a (WS) ; ridge south of Anthony Lake, Elkhorn Range, Maguire & Holmgren 26904 (POM & WS); 12 miles northwest of Unity, Hitchcock 19464 (RSA). Grant County: Indian Springs Road to Strawberry Peak, Blue Mountains, Maguire & Holmgren 26857, 26858 (WS). Hood River County: 5 miles west of Hood River, Cooke 17342 (WS) ; Hood River, Suksdorf, June 15, 1883 (WS). Multnomah County: Bonneville, Suks- dorf 1793, 1797 (WS), Hill 70 (WS). Wasco County: Friend, Hill 29, 56 (WS) ; The Dalles, Suksdorf 2184 (WS). Union County: Two Color Creek, Wallowa Mountains, Cusick 3692 (WS). WasHINGTON. Asotin County: road to Blue Mountains, 6 miles from Anatone, Downen 100 (WS); Anatone, St. John 9561 (WS) ; opposite Zindel, St. John & Rex Brown 3225, 3691 (WS). Chelan County: Boulder Peak, Thompson 11787 (POM) ; Tronson Ridge, Thompson 9320 (POM, MO); Alpine Ridge near Mt. Stuart, Thomp- son 9513 (RSA). Columbia County: one mile east of Table Rock, Umatilla National Forest, Kruckeberg 2524 (RSA). Douglas County: Badger Mountain, northeast of Wenatchee, Hitchcock 17387 (WS). Kittitas County: Lion Lookout Station, Table Mountain, Hitchcock, Rethke & van Raadshooven 3616 (LA, RSA, POM & WS); ridge east of Virden, Thompson 11588 (POM, WS, MO); Salom La Sac, Thompson 10464 (POM, MO); Iron Mountain, Thompson 10039 (RSA) ; hillsides near Yakima River at Cle Elem, Benson 1257 (POM); Beverly Creek, Thompson 10039 (POM) ; Wenatchee Mountains, Thompson 14242 (WS). Klickitat County: Falls of the Co- lumbia, Douglas 297 (refer to citation of type); Klickitat, T. Howell, May, 1870 (WS); Bingen vicinity, Suksdorf 10519, 12387, 10491, 10492 (WS); Falcon Valley, Suksdorf 347; 2569; 7284; 8280 (WS). Skamania County: Dog Creek, Suksdorf 11664 (WS) ; “Zahnberg,” Suksdorf, July 8, 1896 (WS) ; Chenowith, Suksdorf 2568 (WS). Yakima County: North of Wenas, Thompson 14556, in part (WS). 2d. LUPINUS ARBUSTUS subsp. SILVICOLA (Heller) Dunn, Leafl. West. Bot. 7:255. 1955. Lupinus silvicola Heller, Muhlenbergia 6:81. 1910. Type. Placer County, California, near the summit, A. A. Heller 9857 (Ne- vada Agri. Exp. Sta.; isotypes MO, POM, MIN, NMC, WS, US). Lu- 68 MADRONO [Vol. 14 pinus laxiflorus var. silvicola (Heller) C. P. Smith, in Jeps. Man. FI. Pl. Calif. 527. 1925. Lupinus lassenensis Eastw. Leafl. West. Bot. 4:221. 1946. Type. Manzanita Creek, Lassen Volcanic National Park, Califor- nia, L. S. Rose 45262 (CAS). Lupinus laxiflorus var. laxiflorus sensu Phillips pro parte, Res. St. State Col. Wash. 23:197. 1955. Plants 4.5—5 dm. tall, with several stems from a woody caudex, the stems minutely appressed-pubescent to puberulent throughout; leaves cauline with the petioles gradated, 10 cm. long below and 3-4 cm. long above; leaflets 8—10, linear-elliptic, puberulent on both sides, bright green, obtuse or acute, mucronate; peduncles 2—4 cm. long; racemes 8-10 cm. long, flowers scattered or subverticils 8-15 mm. distant; bracts tardily deciduous; pedicels slender, 2—5 mm. long; flowers 6.5-10 mm. long; upper lip of the calyx 3-5 mm. long, the spur ca. 1 mm. long, the tip usually covered by the sides of the banner; banner pubescent dorsally and in the ventral sulcus; wings pubescent laterally near the tip; keel villous- ciliate near the acumen; ovules 4—5; pods 20-25 mm. long, 6—7 mm. wide, finely sericeous. Lupinus arbustus subsp. silvicola is centered in the northern Sierra Ne- vada of California, extending into the Cascade Range of Oregon. It is most commonly found from the arid transition zone, with yellow pine, into the upper Canadian zone, with spruce and white pine. Occasional specimens have been collected at lower elevations. This taxon grades into var. montanus in the north and subsp. calcaratus along the Sierra Nevada. Gene flow from subsp. slvicola was probably responsible for the blue- flowered specimens of subsp. calcaratus in northeastern California, which Eastwood called L. elegantulus, since these specimens resemble subsp. silvicola in being less hairy, but they have the flower proportions of subsp. calcaratus. Several of the specimens cited under var. montanus from the Mt. Hood area (Martin 4815; Oswego, Millard ; Oswego, Drake & Dick- son) show characters which suggest gene flow from subsp. sé/vicola. Christ 16868, cited below from Deschutes County, Oregon, is closer to the typi- cal subsp. sélvicola, but shows traits of var. montanus. Distribution (partial citation). CaLirornia. Alpine County: Winnemucca Lake, Woods Lake region, Peirson 12799 (RSA) ; Lake Alpine region, Pezrson 11579 (RSA). Butte County: Jonesville, Copeland 421 (LA, POM, RSA, MO). Eldorado County: trail to upper Echo Lake, Pezrson 6309 (RSA). Lassen County: west of Fredonyer Pass, Heller 15142 (RSA, POM, MO). Mariposa County: Inspiration Point, Yosemite, June 5, 1897, Congdon (MIN). Modoc County: Fandango Pass, Warner Mountains, Eastwood & Howell 8141 (CAS). Nevada County: Castle Peak, Howell 18531 (RSA) ; Donner Lake, Heller 6944 (POM, MIN, MO); Independence Lake, Hall & Babcock 4532 (POM); Soda Springs, M@. E. Jones 2406 (POM); Truckee, July, 1895, C. F. Sonne (POM). Placer County: Deerpark, Lake Tahoe region, Eastwood 405 (RSA, MO); Summit, Heller 9857 (see type citation). Plumas County: Eureka Peak, J. T. Howell 27699 (RSA); Warner Valley, Applegate 5781 (RSA) ; Silver Lake, July 13, 1929, Merrill (WS, MO). Shasta County: slope of Diamond Peak, Dunn 11812 (RSA, LA. & MIN); Lassen Peak, M. E. Jones 11658 (POM); trail, Drakesbad to summit of Lassen Peak, Peirson 6815 (RSA); two miles east of Hatchet Summit, Heller 15688 (WS, MO). Sierra County: between Gold Lake and Blaisden, Barker 766 (RSA). Siskiyou County: east of Deer Mountain, Heller 15260 (RSA, MO). Tehama 1957] DUNN: LUPINUS ARBUSTUS 69 County: Mineral, July 16, 1935, Epling & Robison (LA); two miles east of Chico Creek, Highway 32, Heller 15672 (WS, MO). OrEGON. County ?: Woodville, T. Howell 1341 (MO). Crook County: mouth of Canyon Creek, Ochoco National Forest, Kucera 19, 22 (WS) ; Tumale Ditch, Whited 624 (WS). Deschutes County: Santiam Pass northeast of Sisters, Christ 16868 (NY). Klamath County: near Fort Klamath, Applegate 4096 (RSA, WS). Wasco County: Marion’s Point Lookout, Mount Hood National Forest, G. N. Jones 4028 (POM). 2e. LUPINUS ARBUSTUS subsp. CALCARATUS ( Kell.) Dunn. Leafl. West. Bot. 7:255. 1955. Lupinus calcaratus Kell. Proc. Calif. Acad. Sci. 2:195, f. 60. 1862. The illustration becomes the type since the type was lost in the San Francisco fire (International Code, 1952, Art. 21). Type locality not given but probably western Nevada or adjacent California, specimens distributed by P. Train from Convict Creek, Mono County, California, May 30, 1937, may be considered as typical. Lupinus variegatus Heller, Muhlenbergia 8:89. 1912. Type. Ruby Mountains near Deeth, Elko County, Nevada, Heller 10551 (Nevada Agri. Exp. Sta.; isotypes, POM, MIN, NMC). Lupinus multitinctus A. Nels. Bot. Gaz. 53:221. 1912. Type. Big Willow Canyon, Idaho, J. F. Macbride 114 (isotypes MIN, WS, MO). Lupinus laxiflorus var. calcaratus (Kell.) C. P. Smith, Bull. Torrey Club 51:304. 1924. Lupinus laxiflorus var. villosulus C. P. Smith, Am. Jour. Bot. 13:530. 1926. Type. Clove Mountains near Deeth, Ne- vada, Heller 9098 (DS). Lupinus elegantulus Eastw. Leafl. West. Bot. 3:20. 1941 (an intermediate with subsp. sélvicola). Type. Fandango Pass, Warner Mountains, Modoc County, California, Eastwood & Howell 8141 (CAS). Lupinus noldekae Eastw. Leafl. West. Bot. 4:149. 1945. Type. Near Hot Creek, Mono County, California, July 1938, A. Noldeke (CAS). Lupinus geraniophilus C. P. Smith, Sp. Lup. 727. 1952. Type. Johnson Creek Ranger Station, Weiser National Forest, Idaho, H. J. Helm 30 (USFS 44172). Lupinus varneranus C. P. Smith, Sp. Lup. 730. 1952 (an intermediate with L. argenteus var. tenellus). Type. Boulder Lake, Idaho National Forest, Idaho, J. M. Varner 82 (USFS 17912). Lupinus multi- tinctus var. grandjeani C. P. Smith, Sp. Lup. 735. 1952. Type. Boise Na- tional Forest, Elmore County, Idaho, EL. Grandjean 460 (USFS 27399). Lupinus graciliflorus C. P. Smith, Sp. Lup. 739. 1952. Type. Fairview Ranger Station, Lemhi National Forest, Idaho, G. A. Miller M-86 (USFS 63045). Lupinus laxiflorus var. laxiflorus sensu Phillips pro parte, Res. St. State Col. Wash. 23:197. 1955. Plants generally 4—6 dm. tall with a cluster of stems from a woody caudex, branching above after the primary raceme reaches anthesis, pu- berulent to finely sericeous and occasionally strigose to somewhat villous; leaves all cauline, the lower petioles up to 15 cm. long, gradated to 2 cm. long above; leaflets 8-10, linear-elliptic-oblanceolate, pubescent on both sides, the largest 3-6 cm. long and 4-8 mm. wide; peduncles 2—3 cm. long; racemes 5-10 cm. long, rather dense, the flowers scattered or ver- ticillate, the verticils 8-10 mm. distant; bracts subpersistent to caducous; pedicels 3—4 mm. long; flowers 11—14 mm. long excluding the spur; upper lip of the calyx 6—7 mm. long, including the spur of 1.6-3 mm. in length, 70 MADRONO [Vol. 14 the lower lip 5.5—6.5 mm. long; banner with abundant pubescence on the back or sparsely ciliate in the dorsal grooves, rarely glabrous, also pubes- cent in the ventral sulcus; wings pubescent laterally near the tip, keel ciliate toward the acumen; ovules 5—6; pods 25-35 mm. long, 8-10 mm. wide, villous. The subsp. calcaratus occupies the mountains of the northern Great Basin, commonly in the Artemisia and Juniperus belt, but some interme- diates are known to extend up into the spruce zone. The blue and lavender forms grade into subspecies arbustus var. montanus. The flowers of subsp. calcaratus are more slender and the calyx spurs longer than those of subsp. arbustus var. montanus. The upper lip of the calyx may be exposed or partially hidden by the banner. In its typical yellow-white form, it is quite distinct. There also appears to be gene flow between subsp. calcara- tus and subsp. sdvicola in the Sierra Nevada of California. The effects of introgression are also apparent between subsp. calcaratus and subsp. pseudo parviflorus in Idaho. Distribution (partial citation). CaLirornia. Alpine County: Leviathan Creek, J.C. Johnson 121 (WS). Inyo County: (w)! Andrews Camp, Bishop region, Peirson 529 (RSA). Lassen County: (b) 14 miles west of Madeline, Balls 14788 (RSA). Mo- doc County: (b) Austin & Bruce 2146 (POM); (lav) Canby Bridge, Balls 14742 (RSA); (b) Cedar Pass, Heller 16205 (RSA); (w) below Clear Lake, Balls 14798 (RSA). Mono County: (w) Convict Lake, Woglum 1875 (RSA), May 30, 1937, Train (UMO); (lav-w) Sonora Pass, A. L. Grant 348, 159, 313 (POM, MO); (w) Long Valley, Fendge 1496 (POM); Virginia Lakes Basin, Peirson 11215 (RSA); Hot Creek region, Peirson 12439 (RSA). Plumas County: (lav) five miles north Chilcoot, Munz 11822 (RSA). Tuolumne County: (b-lav) Sonora Pass, Wiggins 8127 (RSA). IpaHo. Ada County: (w) Boise, Clark 13 (POM, WS); Owyhee, M. E. Jones 25464 (LA). Blaine County: (lav-w) Galena Summit, Macbride 3719 (POM, WS, MO). BorsE County: (yel-w) Squaw Creek (Sweet), Macbride 844 (POM, WS, MO). Custer County: (an intermediate with L. leucophyllus) 5 miles west of Basin- gers, Little Lost River Valley, Hitchcock 15729 (RSA); (law-w) Malkay, Nelson & Macbride 1530 (POM, WS, MO). Elmore County: 15 miles north of Mountain Home, Hitchcock & Muhlick 8670 (WS, MO); 3 miles east of Featherville, Hitch- cock 8780 (WS, MO). Twin Falls County: southeast of Hollister, Piemeisel 44, 1032 (RSA). Washington County: Weiser, M@. E. Jones, July 7, 1899. (POM, MO). Nevapa. Douglas County: (lav-w) Glenbrook, Smith 3803 (POM); Kingsbury Grade to Lake Tahoe, Train 3161 (RSA); (lav-w) Spooner, Smith 3808 (POM), (lav-b) June 23, 1902, Baker (POM). Elko County: 6 miles east of Wells, Train 3639 (RSA) ; near Deeth, Heller 9098 (UC); Ruby Mountains near Deeth, Heller 10550, 10551 (POM, MIN, Nevada Agri. Exp. Sta.). Esmeralda County: Emigrant Pass, Victory Highway, Eastwood & Howell 231 (RSA). Ormsby County: (lav-w) Kings Canyon, Baker 923 (POM, MO); (w) Snow Valley Mountains, Smith 3837 (POM). Washoe County: Alum Creek, Heller 9744 (WS) ; 2.5 miles northeast of Mount Rose Pass, Martin 5539 (RSA, MO); (w) Franktown, Jones 3812 (POM), 3815 (MO) ; (w) Kennedy Pass, Mount Rose, Heller 10342 (POM, WS, MO). Orecon. Baker County: north of Robbinette, Cronquist 6524 (WS). Harney County: Andrews, Applegate 5635 (WS). Malheur County: 10 miles east of Ironside, Peck 26065 (WS); Jamieson, Peck 26059 (WS). Sherman County: DeMoss, Hill 12 1 The letters in parentheses at the begining of each citation refer to the color of the flowers of the specimen: (w)—white. (b)=pblue. (lav) =lavender. (yel) =yellow. 1957] DUNN: LUPINUS ARBUSTUS 71 (WS), 43175 (WS). Cassia County: 23 miles east of Rogerson, Christ 18496 (NY). part, Cusick 1896 (POM, WS, MO) ; (lav) Lower Powder River, Cusick 2514 (POM, WS, MO). Wheeler County: 15 miles northeast of Mitchell, Cronquist 6977 (RSA) ; John Day River, 1 mile south of Service Creek, between Fossil & Mitchell, Cron- quist 6304 (RSA, WS). UrauH. Tooele County: (w) Mount Ibapah, M. E. Jones 11662 (POM, MO), Jones, June 23, 1891 (POM). 2{. LUPINUS ARBUSTUS subsp. PSEUDOPARVIFLORUS (Rydb.) Dunn, Leafl. West. Bot. 7:255. 1955. Lupinus pseudoparviflorus Rydb. Mem. N. Y. Bot. Gard. 1:232. 1900. Type. Bridger Mountains, Montana, Ryd- berg & Bessey 4441 (NY; isotypes MIN, WS). Lupinus laxiflorus var. pseudoparviflorus (Rydb.) C. P. Sm. & St. John in St. John, FI. S. E. Wash. 227. 1937. Lupinus laxispicatus Rydb. Bull. Torrey Club 34:42. 1907. Type. High mountains of Kootenai County, Idaho, J. H. Sandberg, July, 1887 (NY). Lupinus laxiflorus var. durabilis C. P. Smith, Journ. Bot. 13:529. 1926. Type. Bonner County, Idaho, Priest River Range, J. B. Lieberg 2731 (isotype, US). Lupinus laxispicatus var. whithamu C. P. Smith in St. John, Fl. S. E. Wash. 227. 1937. Type. Junction of Di- vide and King Creek trails, Kaniksu National Forest, Washington, C. P. Smith, St. John & Whitham 4170 (DS-Sm; isotype and paratypes at WS). Lupinus sulphureus subsp. whithamii (C. P. Smith) Phillips, Res. St. State Col. Wash. 23:193. 1955. Lupinus laxiflorus var. elmerianus C.P. Smith, Sp. Lup. 106. 1939. Type. Paradise, Wallowa County, Oregon, E. /. A pple- gate 6483 (DS). Lupinus mucronulatus var. umatillensis C. P. Smith, Sp. Lup. 108. 1939. Type. Table Rock, Umatilla County, Oregon, July, 1930, Eldon W. Lyle (DS). Lupinus fieldianus C. P. Smith, Sp. Lup. 567. 1946 (an intermediate with subsp. neolaxiflorus Dunn). Type. Thorn Creek, Idaho National Forest, Idaho, R. C. Fields 224 (USFS 23418). Lupinus lacus-payetti C. P. Smith, Sp. Lup. 574. 1946. Type. Payette Lake, Idaho, M. E. Jones 6251 (POM, MO, US). Plants 3—6 dm. tall, with several simple stems from a woody caudex, these branching later from the upper nodes, the stems 1.5—2.5 mm. in di- ameter, finely, thinly sericeous, with basal leaves generally present at anthesis; petioles of the basal leaves 8.5—-15 cm. long, those of the upper cauline leaves 2.5—3 cm. long; leaflets 7-11, the largest 4-6 (—8) cm. long and 6-10 mm. wide, the tips rounded, or obtuse to acute in intermediate forms; leaflets commonly glabrous on the upper surface, the intermediate forms sparsely pubescent; peduncles 3-6 cm. long; racemes 3-12 cm. long, lax or dense; bracts caducous or tardily deciduous; flowers 9.5—13 mm. long; pedicels 3-5 mm. long, sometimes up to 10 mm. long; upper lip of the calyx gibbous or with a spur 0.6—1.4 mm. long, the tip usually covered by the sides of the banner; banner finely pubescent in the central area or glabrous in 10-30 per cent of the specimens; the wings glabrous in about 30 per cent of the specimens, the others pubescent near the tip; the keel ciliate above near the acumen, sometimes glabrous; pods 7.5—9 mm. wide, 25-35 mm. long, villous, with 3—6 ovules. 72 MADRONO [Vol. 14 Lupinus arbustus subsp. pseudoparviflorus is found in relatively moist habitats but with some drainage, often in considerable shade, from the Douglas fir or lodge-pole pine zones on up to the aspen, spruce and white- pine zones. The center of distribution is in the mountains of western Mon- tana and northern Idaho and southward into Colorado. The occasional plants in the mountains of southern Idaho and the one collection from northeastern Nevada may be thought of as relictual populations. The material treated here as subsp. pseudoparviflorus is somewhat he- terogeneous, and breeding studies may reveal that some of the taxa cited aS synonyms may require recognition. Smith’s var. elmerianus is inter- mediate to subsp. arbustus var. montanus, having the acute leaflet tips of that variety and a vegetative habit which approaches it, but having gla- brate upper leaf surfaces. The two specimens of L. scheuberae Rydb. which were seen resemble subsp. pseudoparviflorus and have been in- cluded in the synonymy of pseudoparviflorus by authors, but the flower size of 14 mm. and the vegetative characters suggest to me that L. scheu- berae may have resulted from hybridization between L. burkei S. Wats. and subsp. pseudoparviflorus. I have not included L. scheuberae here as a synonym of the latter taxon since the morphological hiatus suggests it may warrant treatment as a named hybrid. There is also a race of subsp. pseudoparviflorus present in the Bitterroot Valley, Missoula, Montana, which is one of the more distinctive segregates. This race has the narrow- est leaflets, racemes up to 25 cm. long, pedicels to 10 mm. long, and be- comes 7—9 dm. tall. It has not been named, but it is far more distinctive than most of the variants that have names. It seems best at present merely to call attention to the fact that these plants form what appears to be a morphological extreme within the range of variation attributed to pseudo- parviflorus. The var. whithamiu C. P. Smith may warrant recognition, but the “gla- brous wings” which characterize this variety occurred in a high percen- tage of the specimens in subsp. neolaxiflorus, as well as in a fair percen- tage of the rest of the population of subsp. pseudoparviflorus. The highest percentage of individuals that were glabrous in one or more of the flower parts occurred in the population in northern Idaho and Washington. The glabrous nature was observed again in the specimens from Colorado, but there it appears to be due to introgression from L. rubricaulis of the L. argenteus complex. Distribution (partial citation). CoLorapo. Larimer County: near North Park, Aug., 1894, Osterhaut (PHIL). County ?: Ursten’s Pass, July, 1873, Coulter (PHIL). IpaAHO. Bonner County: Priest River, Experimental Forest, Daubenmire 43122 (WS), 42175 (WS). Cassia County: 23 miles east of Rogerson, Christ 18496 (NY). Bear Lake County: 13 miles west of Bloomington, Christ 18678 (NY). Clearwater County: above Orofino, Constance, Dimond, Rollins & Worley 1082 (WS). Idaho County: Indian Post Office, Q. Jones 293 (RSA); 5 miles south of Harpster, cliffs above Clearwater River, Hitchcock & Muhlick 8452 (WS). Latah County: 8 miles south of Troy, Daubenmire 46131 (WS); Paradise Ridge, Daubenmire 37403 (ap- proaches L. scheuberae) (WS). Lemhi County: 6 miles north of Gibbonville, Salmon 1957] DUNN: LUPINUS ARBUSTUS 13 River, Christ & Ward 14695 (NY); Panther Creek 8 miles north of Cabin Creek, Hitchcock 14270 (POM, RSA). Lewis County: Mission Creek, St. John, Cary, Put- nam & Warren 3247 (WS); north edge of Winchester, Dawbenmire 46237 (WS). Nez Perce County: near Lewiston, Heller 3224 (WS, MO); bluffs of Clearwater River north of Spalding, Daubenmire 37479 (WS). Shoshone County: Siwash Peak, St. Joe National Forest, Moore 433 (WS); sides of Quarles Peak, Wilson 193, 211, 205, the latter approaching L. scheuberae (WS). Montana. Flathead County: southeast of McDonald Lake, Hitchcock 18280 (RSA, WS) ; Columbia Falls, Dunn 9723 (RSA, LA, MIN, UMO), June, 1894, Wil- liams (UMO). Gallatin County: Bridger Mountains, 1 mile south of Brackett Creek, Hitchcock & Muhlick 12460 (WS, MO); Middle Creek Canyon, Bozman, Blankin- ship, June 26, 1900 (WS). Lake County: 10 miles northeast Polson, Flathead Lake shore, Hitchcock 15344 (RSA, WS). Lewis & Clark County: 3 miles east of Danaher Ranger Station, Hitchcock 18718 (RSA, WS). Missoula County: Bitterroot Valley, M. E. Jones 11655 (POM); Rattlesnake Drainage, Hitchcock 14560 (RSA, WS). Park County: 15 miles south of Wilsall, Hitchcock & Muhlick 12438 (WS). Powell County: 8 miles northeast of Helmville, Hitchcock 17843 (RSA, WS). Ravalli Coun- ty: Palisade Pl. Ranger Station, Bitterroot Mountains, Hitchcock 15344 (RSA). Sanders County: 3 miles west of Dixon, Hitchcock 2873 (WS). County ?: Nanicke, Cooke 17324 (WS). OrEGON. Wallowa County: Paradise, Applegate 6483; east of Sacajawea Camp over the Snake River Canyon, Kruckeberg 2456 (RSA, WS). WasHINGTON. Chelan County: Mission Canyon, 10 miles south of Cashmere, Hitchcock 17304 (WS); 15-20 miles up Wenatchee River, near Cascade Mountains, 1889, Vasey 38752 (WS). Okanogan County: head of Cedar Creek, south of Con- conully, Fiker 831, 832, 883 (WS). Pend Oreille County: divide & Kings trail, junc- tion, Kaniksu National Forest, Smith, St.John & Whitham 4171, 4177 (WS) ; Kings Lake Road, below south Skookum Drive, Kaniksu National Forest, Smith, St. John & Whitham 4169, 4180 (WS). Spokane County: Newman Lake, Smith & St.John 4135, 4136 (WS). University of Missouri, Columbia, Missouri. LITERATURE CITED AGARDH, JACOB G. 1835. Synopsis Generis Lupini. Lundae, Carlos Fr. Berling. ANDERSON, Epcar. 1949. Introgressive Hybridization. Wiley & Sons Inc. New York. Dovuctas, Davin. 1914. Journal kept by David Douglas during his travels in North America, 1823-1827. Royal Hort. Soc. London. Dunn, Davin B. 1954. A method of mounting pressed flowers for study and preser- vation. Rhodora 56:258-260. . 1955. A new subspecies and new combinations in Lupinus. Leafl. West. Bot. 7:254-5. . 1956a. Leguminosae of Nevada, Part 2—Lupinus. Contr. toward a Flora of Nevada, No. 39. Bureau of Plant Industry, U.S. Dept. Agri. Washington, D.C. —. 1956b. The lupines collected by David Douglas from 1825-1827: their type localities and nomenclature. Leafl. West. Bot. 8:47—54. GOLDSCHMIDT, RicHArD B. 1952. Evolution, as viewed by one geneticist. Amer. Sci. 40:84—-98. INTERNATIONAL CODE OF BOTANICAL NOMENCLATURE. 1952. Utrecht, Netherlands. LAWRENCE, DONALD B. 1950. Paper presented at Columbus, Ohio, A. I. B. S. meetings (unpublished) . PHILLrPs, LYLE. 1955. A revision of the perennial species of Lupinus of North Amer- ica. Res. St. State Coll. of Wash. 23:161—201. PursH, F. 1814. Flora Americae Septentrionalis. White, Cochrane & Co., London. 74 MADRONO [Vol. 14 SOME ADDITIONS TO THE CALIFORNIA MOSS FLORA Howarp Crum! Considerable work has been done on the moss flora of California during recent years, and a firm basis has been laid for further floristic studies by the publication of a checklist of California mosses by Koch (1950). The size of the state and the extreme diversity of available habitats make it probable, however, that many more species will be discovered in Califor- nia, not only by careful collecting but also by critical study of many difficult genera, particularly of the Pottiaceae and the Brachytheciaceae. Recently, from 1954 to 1956, Wilfred B. Schofield of Wolfville, Nova Scotia, made a sizable and very interesting collection of mosses in central California, mostly in the San Francisco Bay region. In addition to many rarities previously recorded from the area, he found one very distinctive new species and several important range extensions which are reported below. A few additional specimens worthy of recording here were sent me for determination or confirmation by Mrs. Fay A. MacFadden of Los Angeles. All the specimens are deposited in the Herbarium of the National Museum of Canada (CAN). Hymenostomum (Kleioweisia) inoperculatum sp. nov. (Figs. 1-6). Planta tenella, usque ad 1 mm., sordido-viridis, paucifoliatus, folia usque ad 2 mm. longa, madida patula, sicca crispula, inferiora minuta et ovata, caetera sensim major, oblongo-lanceolata, acuta et apiculata, concava, marginibus integerrimis, superne late involutis; costa basi 48—56y, bre- viter excurrens; cellulae basilares in dimidio inferiore hyalinae, oblongae, parietibus tenuibus, cellulae laminae superioris hexagonae, opacae, den- sissime papillosae, dioicum (?); flos masculae non vidi, seta 3 mm. longa, pallida, erecta; capsula 1.25—-1.75 mm. longa, exserta, erecta, oblongo- cylindrica, oblique et longe rostrata, clausa, sine operculo, sporae 16—19yp., minute et dense papillosae. Type. On soil in garden in front of Sequoia Hall, Stanford University, Santa Clara County, California, February, 1955, W. B. Schofield (s.n.) ; growing with Pottia arizonica var. mucronulata Wareham (CAN). This species is distinct from all other American species of Hymenos- tomum known to me in having clearly exserted, cleistocarpous capsules. It is related, perhaps most closely, to H. exsertum (Broth.) Broth. of China and Japan, but judging from the single Japanese specimen which I have seen, named by Dr. Akira Noguchi and kindly communicated by Dr. Harumi Ochi, the Californian species differs markedly in having much shorter and relatively broader leaves which are less strongly crisped when dry and capsules which are narrower and cylindric to subelliptic, rather than subglobose, and end in longer, more slender beaks. A further differ- ence is that the capsules are clearly exserted, and the setae much exceed the uppermost leaves in length, whereas in H. exsertum, although the 1 Published by permission of the Director of the National Museum of Canada, Ottawa. 1957] CRUM: CALIFORNIA MOSSES aS ALLA r Wall Fics. 1-6. Hymenostomum inoperculatum. 1, leaves, * 30; 2, upper cells of leat and apiculus, x 275; 3, two sections taken in upper portion of leaf, « 275; 4, two capsules, * 30; 5, areolation of capsule wall showing the irregularity of cells at the zone of dehiscence in most other species of mosses, K 275; 6, spores, & 520. setae are fairly long and the capsules appear to be exserted, they are actu- ally surpassed in length by the strongly contorted and spreading upper leaves. Hymenostomum ino perculatum conforms very well to the descrip- tion of H. semidiaphanum Thér. of Mexico, but the type of that species was examined and found to be inaccurately characterized. The operculum is clearly differentiated and not “‘haud secedens.” (Thériot was probably 76 MADRONO [Vol. 14 misled by the juvenile development of the capsules in the type collection of H. semidiaphanum.) The position of H. semidiaphanum in the sub- genus Kleioweisia seems highly unlikely. POTTIA ARIZONICA var. MUCRONULATA Wareham. On soil in garden in front of Sequoia Hall, Stanford University, Santa Clara County, Febru- ary, 1955, W. B. Schofield (s.n.). 1 have examined the only previous Cali- fornia collection, from Los Angeles County (Koch, 1950), as well as the type from Arizona. STEGONIA LATIFOLIA var. PILIFERA (Brid.) Broth. This interesting moss, previously known only from the Yukon and from the Canadian Rocky Mountains in North America, was collected among rocks at 13,950 feet altitude on Mount Barnard, Tulare County, by Peter H. Raven on July 25, 1955. The specimen was sent me by Mrs. MacFadden. TORTULA STANFORDENSIS Steere. On moist, hard-packed soil of open area, Purissima Creek, San Mateo County, November 8, 1955, W. B. Scho field 6322. This species, recently described from Santa Clara County (Steere, 1951), has also been reported from Alameda County (Koch & Ikenberry, 1954). GRIMMIA MARINIANA Sayre. Dark, hyaline-tipped cushions among boulder crevices, Mount Tamalpais, Marin County, March 26, 1955, W. B. Schofield 5798; steep, exposed rocks, east side of East Peak, Mount Tamalpais, Marin County, April 8, 1956, John T. Howell H184 (comm. Mrs. MacFadden). Both of these specimens come from the general vicin- ity of the type locality and conform very well to Sayre’s recent description (1955), except that the alar cells of the leaves have thickened cross-walls, a character which Sayre perhaps overlooked. EPHEMERUM MINUTISSIMUM Lindb. Locally abundant on small, clayey patches in an open field, Big Basin, Santa Cruz County, March 24, 1956, W. B. Schofield 6450; local, on bare patches near margin of field, near Ventura Hall, Stanford University, Santa Clara County, February 23, 1955, W. B. Schofield 5723. Duplicates of these specimens were sent to Dr. Virginia S. Bryan, who confirmed the determinations tentatively, pending further understanding of the puzzling variation exhibited by E. serratum and E. minutissimum in the West. Koch (1950) listed only E. serratum (Hedw.) Hampe from California, where it had been collected only once, in San Francisco. Grout (1928-40) mentioned only Massachu- setts and Saskatchewan in the distribution of E. minutissimum. MIELICHHOFERIA MIELICHHOFERIANA (Funck) Limpr. (figs. 7-13). On a tree trunk, Little Butano Creek, Santa Clara County, November 6, 1954, W. B. Schofield 5748. Extremely rare and localized in its distribu- tion, this species has been collected in only a few other widely separated places in North America (in the Lake Superior region of Michigan and Ontario and also in Maine). In the Old World it is known from Scandi- navia, the Alps and elsewhere in Central Europe, the Pyrenees, and the Caucasus. The plants which Mr. Schofield collected in the Coast Ranges 1957] CRUM: CALIFORNIA MOSSES 77 Fics. 7-13. Mielichhoferia mielichhoferiana. 7, leaves, X 23; 8, basal cells of leaf, < 205; 9, upper cells of leaf, * 205; 10, habit of fruiting plant, « 8; 11, portion of peristome, annulus, and upper exothecial region, X 205; 12, portions of peristomes, showing variation in shape of teeth, x 65; 13, cells from neck of capsule showing stoma, X 205. of central California seem to me typical in every structural detail and only slightly different in habit of growth, as they are somewhat more elon- gate, less densely compacted in tufts and more loosely and freely branched than any other material that I have examined, although they greatly re- semble in growth form the illustrations given in the “Bryologia Europaea”’ (plate 328, as M. nitida Hornsch.) and reproduced in Grout’s ‘““Moss Flora of North America.” These differences are indeed slight and almost 78 MADRONO [Vol. 14 surely not of a genetic nature, and they are doubtless associated with the unusual habitat in which the plants were found, namely, on the bark of a tree. Insofar as I have been able to ascertain, the species has never before been reported from any substrate other than rock and, more specifically, from rocks containing alum or iron and copper ions. Persson (1948) has recently reviewed the evidence for believing that the genus Mzelichho feria shows a predilection for copper-containing rocks. More typically developed, though sterile, plants of a Muielichhoferia, tentatively named M. mielichhoferiana by A. LeRoy Andrews, were also collected in California by John Thomas Howell “‘on shaded, moist rock, Merced River Canyon, just above Briceburg, Mariposa County.” They were recently sent to me by Mrs. MacFadden, and I see no reason to doubt the determination. I have prepared the accompanying camera lucida drawings in order to give greater publicity to M. mielichhoferiana, as this interesting and rare- ly collected species has not been illustrated from American material be- fore. It should be noted that the shape of the capsule varies with age. When old, empty and somewhat wrinkled, it is oblong-cylindric, but when operculate, it is more ovoid with a more distinct neck. ZYGODON VIRIDISSIMUS ( Dicks.) Brown. In crevices of bark of Litho- carpus, forming small patches, Big Basin, Santa Cruz County, November 27, 1955, W. B. Schofield 6371. Subsequent to the publication of his checklist, Koch (1950a) reported this species for the first time from Cali- fornia, from a collection made in Humboldt County. BRACHYTHECIUM WASHINGTONIANUM Eaton ex Grout. On shaded soil bank, Purissima Creek, San Mateo County, November 8, 1956, W. B. Scho field 6337; on wet bank, same locality and date, 6346; rock outcrop- ping of brook bank, Big Basin, Santa Cruz County, November 27, 1955, W. B. Schofield 6357. Koch (1950) listed three collections of this species from Trinity, Siskiyou, and Humboldt counties. These collections extend the range southward considerably. I have studied the type, which comes from Washington. The alar cells of the stem leaves are large and lax and much more clearly differentiated than Grout (1928-40) indicated and seem to me indicative of a clearer relationship to B. rivulare B.S.G. than Grout appears to have realized. National Museum of Canada Ottawa, Canada LITERATURE CITED Grout, A. J. 1928-40. Moss flora of North America north of Mexico. 3 vol. Newfane, Vt. Kocu, L. F. 1950. Mosses of California: an annotated list of species. Leafl. West. Bot. 6: 1-40. . 1950a. Mosses of California. II. Zygodon viridissimus. The Bryologist §32213-2115. & G. J. IKENBERRY. 1954. Preliminary studies of California mosses. II. The Bryologist 57:291-300. 1957] REVIEW 79 Persson, H. 1948. On the discovery of Merceya ligulata in the Azores with a discus- sion of the so-called “copper mosses.” Rev. Bryol. et Lichénol. 17(1/4) :75—-78. SAYRE, GENEVA. 1955. Grimmia mariniana, a new species from California. The Bry- ologist 58:323-325. Fig. 1. STEERE, W. C. 1951. Tortula stanfordensis, a new species from California. The Bry- ologist 54:119-123. Figs. 1-9. REVIEW The Genus Nicotiana. By T. H. GoopsPEep. xxi + 536 pp., illustrated. Chronica Botanica Vol. 16, No. 1/6. 1954. Waltham, Mass.: The Chronica Botanica Co.; San Francisco, Calif.: J. W. Stacey, Inc. Buckram. $12.50. According to Professor Goodspeed: “The primary objective of the Nicotiana investigations . . . has been the accumulation of evidence bearing upon the origin, evolution and relationships of the modern species of the genus (p. 1).” Merely as a general presentation of this accumulated information this book is impressive. As Volume 16 of Chronica Botanica, it contains some 536 pages, 50 tables, and 118 plates as well as text-figures. Studies in distribution, morphology, and cytogenetics (com- prising the first four parts of the book), many of which began as long ago as 1904, are discussed in detail. The important contributions of the Nzcotzana work to gen- eral genetics and cytology are not touched upon, however, for obvious reasons. It is very useful, of course, to have all of this material, which has appeared in many different journals over the years, brought together and integrated. One cannot fail to be impressed with the Olympian outlook of the author as he discusses a genus which has contributed so much to an understanding of basic biological problems. One must always keep in mind, however, that, in a summary work of this sort, there is a great deal more interpretation, or even opinion, than in the individual contri- butions. Part I is concerned with the evidence from distribution. Since there is no paleo- botanical evidence to use in determining past distribution, Goodspeed discusses in detail the present distribution upon the major land masses. Much of this is repetitive from chapter to chapter and speculative, and suffers from the defects of reasoning based upon patterns of distribution. It may be true that in paleontology one must assume that the present is the key to the past. But when one is dealing with patterns of geographical occurrence he never has all of the requisite facts of present-day ecol- ogy and genetics to make an adequate judgment of the factors involved in the past. Nevertheless, Goodspeed deduces that Nicotiana originated in South America in the early Tertiary, thence spreading to North America and across Antarctica to Aus- tralia, New Zealand, and the islands of the South Pacific. The bases for these deduc- tions appear to be primarily the times at which the various land masses involved became joined or separated and the presumed length of time for a particular “quan- tity” of evolution to have taken place. The limits of error of these two criteria have scarcely been securely set. But then this is more or less standard operating procedure in taxonomic problems and has been for some time. All but four of the sixty recognized species of Nicotiana have been grown in the greenhouse or field and have been available for experimental studies. Morphological and anatomical investigations have been carried out on all. Goodspeed limits his dis- cussion to general morphology, anatomy, and trichomes (Part II), cytology of species (Part III), and cytology of Fy interspecific hybrids (Part IV). Nicotiana is not char- acterized by any one specialized feature and there is marked variation between species in habit, inflorescence, and flower, all of which is well illustrated. As one would expect, the majority of anatomical distinctions are quantitative and thus of less use at the level of species. Goodspeed attaches especial importance to the trichomes which are believed to follow patterns paralleling those developed from studies of general morphology, geographical distribution, and cytogenetics. An especially interesting part of the book for me is the section on the cytology of 80 MADRONO [Vol. 14 species and the cytology of Fy, interspecific hybrids. The details of chromosome mor- phology and behavior for fifty-six species and their hybrids are described in five chapters. Apparently there is no species known to be an obligate apomict, but some may be facultatively apomictic upon occasion. Chromosome morphology is quite homogeneous. Aneuploidy appears to be rare, but amphiploidy is of frequent occur- rence and is thought to be a basic factor in the earliest evolution of the genus. Karyo- types and complete meiotic sequences in representative species and hybrids are illus- trated. In the interests of establishing chromosome homology, which is expressed in the amount and type of pairing in F; hybrids, Goodspeed makes an analysis of 215 cases. He finds that they may be grouped into five categories ranging from complete pairing to almost no pairing. It is obvious that the analysis of chromosome nature and behavior forms a major source of criteria for the segregation and alignment of groups of species and sections. Part V, entitled “Phylesis,” consists of two chapters. The first paints in very broad and general terms the author’s views concerning the origin, relationships, and evolu- tion in the genus. The second discusses the possible “future” of the genus and is based upon what is deduced about the present and the past. Here, as elsewhere in the book, little information is given about the genetics of species populations. One does not gain any concept of population structure or, indeed, of the genetic systems at work at the level of species and below. This is a very noticeable and, I would think, serious gap in the presentation and in the data. Goodspeed envisions the present-day species to have arisen from a complex comprising pre-Nicotiana, pre- Cestrum, and pre-Petunia elements. Evolution of these hypothetical forms presumably gave rise to the modern genus with its cestroid and petunioid complexes. Nicotiana can be considered the residue of a polyploid complex based on the chromosome num- ber n=6. Forms with this number are no longer extant, however, and the existing species may be arranged on two levels of polyploidy with about one-third of these at the higher level. These conclusions are diagrammed in an interesting representation of the levels of advancement and the degree of relationship of the various taxa. Goodspeed believes that in the future: “... the genus may be expected to expand both genetically and geographically, with greatest increase in number of species on the higher polyploid level.” The concluding section of the work, written with the collaboration of Helen- Mar Wheeler and Paul C. Hutchison, is the section on formal taxonomy and nomen- clature. Its 170 pages contain the usual historical resumé, generic and specific descrip- tions, keys to taxa, and lists of cited specimens. Here the relatively minor attention devoted to lower categories is emphasized by the use of variety as the only category below the rank of species and, while these are keyed out, they are not discussed in detail. One cannot be sure whether the authors really mean these forms to be con- sidered taxa below the level of subspecies (as the International Rules of Botanical Nomenclature would indicate) or if they consider the term variety to be equivalent to subspecies or to be the only infraspecific category. The reader will almost certainly be drawn to make a comparison between ‘The genus Nicotiana” and other general works on genera of plants and animals which have appeared in the last few years. I feel that it is not only difficult, but somewhat unfair, to make such comparisons. Investigators have naturally emphasized those features of the groups upon which they have worked which are the most amenable to study and which thus make a significant quantitative as well as qualitative contri- bution. For Nicotiana it would seem that the greatest contributions lie in the fields of morphology and cytogenetics. For this reason Goodspeed’s book will be an indis- pensable reference work for the biologist interested in evolutionary problems.— Ricuarp W. Horm, Department of Biological Sciences, Stanford University. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication should not exceed an estimated 20 pages when printed unless the author agree to bear the cost of the ad- ditional pages at the rate of $15 per page. Illustrative materials (includ- ing “typographically difficult” matter) in excess of 30 per cent for papers up to 10 pages and 20 per cent for longer papers are chargeable to the author. Subject to the approval of the Editorial Board, manuscripts may be published ahead of schedule, as additional pages to an issue, provided the author assume the complete cost of publication. Shorter items, such as range extensions and other biological notes, will be published in condensed form with a suitable title under the general heading, ““Notes and News.” Institutional abbreviations in specimen citations should follow Lanjouw and Stafleu’s list (Index Herbariorum. Part 1. The Herbaria of the World. Utrecht. Second Edition, 1954). Articles may be submitted to any member of the Editorial Board. MADRONO A WEST AMERICAN JOURNAL OF BOTANY A quarterly journal devoted to the publication of botanical re- search, observation, and history. Back volumes may be obtained from the Secretary at the following rates: Volume I,1916-1929 . . . . . $5.00 Volume II, 1930-1934. . . . . 5.00 Volume ITI, 1935-1936 . . . . 5.00 Volume IV, 1937-1938 . . . . 5.00 Volume V, 1939-1940. . . . . 5.00 Volume VI, 1941-1942 . . . . 5.00 Volume VII, 1943-1944 . . . . 5.00 Volume VIII, 1945-1946 . . .. . 5.00 Volume IX, 1947-1948 . . . . 6.00 Volume X, 1949-1950. . . . . 7.00 Volume XI,1951-1952 . . .) 8.00 Volume XII, 1953-1954 . . . . 8.00 Volume XIII, 1955-1956 . . . 8.00 Single numbers, Volumes: and.2 (32. 4. 96 ae .50 Succeeding numbers, (except Volume XI, No.2) . . .- ~~ 1.00 Volume XI, No.2? 2 a 1.50 The subscription price of MaApRONo is $4.00 per year. If your institution does not now subscribe to MADRONO, we would be grateful if you would make the necessary request. Since there is no extra charge for institutional subscriptions, an individual may hold membership in the California Botanical Society on the basis of his institution’s subscription. Address all orders to: G. THoMAsS Rossins, Corresponding Secretary Department of Botany University of California, Berkeley 4, California % R ot a4 % \ MADRONO VOLUME 14, NUMBER 3 JULY, 1957 Contents PAGE THE CONCEPT OF THE FLOWER AND THE THEORY oF Homotocy, Herbert L. Mason 81 MitToTic CHROMOSOME STUDIES IN THE GENUS ASTRAGALUS, S. Conrade Head 95 INNOVATIONS IN DupLEyA, Reid Moran LOG) » NOoTES AND NEWS: CLEISTOGAMY IN MiInuLus Douc- LASII GRAY, V. F. Hesse; NOTES ON CALIFORNIA GrassEs, Beecher Crampton 108 CALIFORNIA BOTANICAL SOCIETY, REPORT OF THE TREASURER FOR 1956 110 DOCUMENTED CHROMOSOME NUMBERS Dat A WEST AMERICAN JOURNAL OF BOTANY | PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO A WEST AMERICAN JOURNAL OF BOTANY Entered as second-class matter at the post office at Berkeley, California, January 29, 1954, under the Act of Congress of March 3, 1879. Established 1916. Subscription price $4.00 per year. Published quarterly and issued from the office of Madrono, Herbarium, Life Sciences Building, University of California, Berkeley 4, California. BOARD OF EDITORS HERBERT L. Mason, University of California, Berkeley, Chairman EpGaAr ANDERSON, Missouri Botanical Garden, St. Louis. LyMAN BENSON, Pomona College, Claremont, California. HERBERT F. CopeELAND, Sacramento College, Sacramento, California. Joun F. Davinson, University of Nebraska, Lincoln. Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts. MivpreD E. Maruias, University of California, Los Angeles 24. MARION OWNBEY, State College of Washington, Pullman. IrA L. Wiccins, Stanford University, Stanford, California. Secretary, Editorial Board — ANNETTA CARTER Department of Botany, University of California, Berkeley. Business Manager and Treasurer—MAtcoLm A. Noss Carnegie Institution of Washington, Stanford, California CALIFORNIA BOTANICAL SOCIETY, INC. President: Rimo Bacigalupi, Jepson Herbarium, Department of Botany, Univer- sity of California, Berkeley, California. First Vice-president: Richard W. Holm, Natural History Museum, Stanford University, Stanford, California. Second Vice- president: Lyman Benson, Department of Botany, Pomona College, Claremont, Cali- fornia. Recording Secretary: Mary L. Bowerman, Department of Botany, University of California, Berkeley, California. Corresponding Secretary: G. Thomas Robbins, Jepson Herbarium, Department of Botany, University of California, Berkeley, Cali- fornia. Treasurer: Malcolm A. Nobs, Carnegie Institution of Washington, Stanford, California. 1957 | MASON: CONCEPT OF THE FLOWER 81 THE CONCEPT OF THE FLOWER AND THE THEORY OF HOMOLOGY’? HERBERT L. MASON In seeking sound concepts around which to build our ideas of the tax- onomy of the flowering plants, we find that confidence in our concepts of the flower has steadily deteriorated since the exposition and later clari- fication of the classical theory of the flower and especially since the pub- lication by Zimmermann (1930) of the telome theory. The ideas embodied in the classical theory—namely, that the parts of the flower are metamor- phosed leaves—had their beginnings early in botanical history, going back at least to Linnaeus and several of his associates. The most significant enunciations of the theory, however, are accorded to Caspar Friedrick Wulff (Samassa, 1896) and to the poet-philosopher Wolfgang von Goethe (1790). It is important to realize that the classical theory embodied the concept of homology that was being discussed by the 18th and 19th cen- tury anatomical philosophers of Germany and France, including also such men as Cuvier (1800), Oken (1807), and St. Hilaire (1807). The concept originated in the thinking of the classical geometricians and was here being applied to biology. The language was vague because its abstractions, although clear to its authors, were burdened with confusions of logical types that led to confusion of understanding among the readers in their attempts to apply it. For instance, Goethe spoke of the parts of the flower as being “metamorphosed leaves” while Oken created bewilderment among zoologists by speaking of the “humerus of the head.” These ideas became variously known in botany as the ‘“‘classical theory of the flower”’ or, more broadly, as “‘the theory of morphology,” alluding here to the con- cept of homology that was embodied in it. It met with immediate criticism as presenting an illogical consequence, since “‘that which never was a leaf could scarcely be a metamorphosed leaf.” There followed a period of great discussion. Of the clarifications which emerged, that expressed by the British botanist, Lindley (1838, p. 59), seems especially worthy of our attention because it constitutes one of the first clear expressions of the concept of morphological equivalence. Said 1 The substance of this paper was presented to the Eighth International Botanical Congress in Paris in 1954 under the title, “The Controversy of the Flower and its Bearing on Phylogenetic Taxonomy.” Prior to this it was presented in seminars at the Davis and Berkeley campuses of the University of California and at Stanford University. Since these preliminary presentations the manuscript has been amended to indicate the formal nature of the Theory of Homology, and the title has been changed to indicate more precisely the scope of the paper. Acknowledgments are due to Jean Langenheim for many lively discussions lead- ing to the organization and clarification of the subject matter, and to Annetta Carter, Francia Chisaki, Helen Sharsmith, and Isabelle Tavares for assistance in perfection of the manuscript. MaproNno, Vol. 14, No. 3, pp. 81-112, July 30, 1957. Hes, 82 MADRONO [Vol. 14 Lindley, ‘It has been observed in a report made to the British Association at their meeting in Cambridge in 1833, when adverting to this doctrine, that when those who first seized upon the important but neglected facts out of which the modern theory of morphology has been constructed, asserted that all appendages of the axis of the plant are metamorphosed leaves, more was certainly stated than evidence would justify: for we cannot say that an organ is a metamorphosed leaf which in point of fact, never was a leaf. What was meant, and that which is supported by the most conclusive evidence, is, that every appendage of the axis, whether leaf, bract, sepal, petal, stamen or carpel, is originally constructed of the same elements, arranged upon a common plan, and varying in their man- ner of development, not on account of any original difference of structure, but on account of special and local predisposing causes; of this the leaf is taken as the type because it is the organ which is most usually the result of the development of those elements; is that to which other organs generally revert when, from any accidental disturbing cause, they do not assume the appearance to which they were originally predisposed; and moreover is that in which we have the most complete state of organ- ization.” There are several interesting ideas embodied in Lindley’s statement with which we may or may not agree. But what is important to my thesis is the concept of morphological equivalence which assumes that equiva- lent structures have had a like origin, are arranged in accordance with a common plan, and differ from one another in their further development as a consequence of special and local predisposing causes. This assumes some sort of identity in the early ontogenetic stages. Some such ideas as here expressed have governed the thinking in morphological research for a century and a half. To a large measure they have been responsible for the common practice of interpreting one structure in terms of another on the assumption that all structures are deviations from a prior “‘common plan.” Each structure is assumed to have arisen through the ontogenetic modi- fication of a pre-existing structure. Thus the concepts of the classical theory and of morphological equivalence, at least as to their explanation, are definitely ontogenetic concepts whose explanation rests in knowledge of ontogeny. The alternative to these ideas is that structures arise anew, possibly upon a foundation of the old, but in no sense to be regarded as being a modification of the old. A structure arising in this manner is said to arise sui generis. This idea also has followers, among them Gregoire (1938). I do not imply in placing this theory in apposition to the preceding that it does not also entail an ontogenetic explanation. As we look into modern concepts of the flower we find that both the classical theory and the concept of homology generally are being seriously questioned as logical bases for morphological interpretation. New ap- proaches to the problem are being investigated outside the scope of the classical theory and the concept of homology. Since a sound taxonomy of 1957] MASON: CONCEPT OF THE FLOWER 83 the flowering plants and especially a sound basis for phylogenetic inter- pretation must rest upon an understanding of the flower, I have attempted to review these concepts and evaluate this controversy in the hope of de- veloping at least an adequate working hypothesis concerning the flower that might prove of value to taxonomy. As we examine the various theories of the flower that have deviated from the classical theory, nearly all point to the fact that the classical theory fails to explain one aspect or another of the flower. Most workers pointed to various features of the carpel, especially to its relation to struc- tures resulting in an inferior ovary, as not being interpretable in terms of the homologies of appendages. With these difficulties as a beginning, other difficulties became apparent, and there ensued a re-evaluation of all points of difference between carpels and typical leaves and of other parts of the flower in their departure from appendages homologous with leaves. Some workers have investigated the organogenesis of the flower apex in com+ parison with that of the shoot and have pointed to differences. As a result of their researches upon these problems, Thomas (1934), Thompson (1935), Gregoire (1938), and Lam (1948) insist that an entirely new approach to the interpretation of the flower is called for. In fact Gregoire would overthrow homology as being inapplicable to any comparison of the flower and the shoot, and would insist that the flower is an organ suz generis and not in any ontogenetic sense comparable with a shoot. This point is amply discussed and the argument met by Boke (1947). I do not intend systematically to discuss here the points raised by each of these investigators; what I am concerned with is a re-evaluation of the investigative and intellectual approaches to the problem. There are at least two basic approaches to the study of the flower. The first of these, the traditional method, employed the concept of homology as a system of logic interpreting structures in terms of abstract categorical levels of morphological equivalence, without necessarily knowing the pre- cise details of ontogenetic elaboration beyond identity, origin, and relative position. This method has been useful, for, although it did no more than categorize structures in terms of the formal relations of a unitary classi- fication, it aided in determining what were comparable structures and at the same time provided a basis, on the one hand, for explaining points of likeness between two structures, and, on the other, a point of departure for ordering and explaining the differences between structures. This has been the chief use to taxonomy of the classical method of morphology. The other approach to the problem seeks more precise information. It presumes to base its interpretations upon discovering the details of onto- genetic elaboration. Obviously, if such researches are faithfully followed to their conclusion and the details of ontogenetic elaboration carefully worked out, we will have a considerable body of fact upon which to base our judgments and our interpretations, but such facts will in no way eliminate the need of a logical system nor obviate the necessity of estab- lishing intellectual concepts for interpretation. 84 MADRONO [Vol. 14 As we review the controversy of the flower, it appears to be a contro- versy between these two approaches to the interpretation of the flower. The researches of Zimmermann and his development of the telome theory call for detailed investigation of anatomical organization and ontogeny at a level of structural organization below that envisaged in the system of logic developed by Goethe for the classical theory of the flower, which operated strictly on the level of gross organography. If we are to employ the concept of homology to function at this more detailed level, we will have to redesign our system of logic so that it will be useful in terms of the structures that pertain at this level of organization. To some, the telome serves this purpose (Zimmermann, 1930). The telome is defined in terms of empirical criteria derived from the structure of the extinct Psilophytalean genus Riynia wherein the terminal segment of the axis beyond the last dichotomy is accepted as the basis of the concept of the telome. The segment of the axis upon which the telome rests is called the mesome. Thus telomes, as they give rise to new telomes in their ontogeny, become mesomes. It follows that if we can get at the facts of anatomy, we should be able to trace, through the structure of the individual, the onto- genetic history of telome elaboration, and, through the ontogeny of elabo- ration of the individual, we may be able to interpret the phylogeny of telome differentiation in the origin of organs. This is the simple logic of the case, and we read such phrases as “‘seeking the evidence of ancient dichotomies” (Lam, 1948). Obviously the details of morphological differ- entiation in the higher plants are more complicated than those evident in the fossil Riynia. These complications are systematically explained in the telome theory by a series of elementary processes accounting for the origin of the telome and its phylogenetic elaboration to produce the most complicated of higher plants. I shall not go into this in detail, but will discuss some limitations to its use in comparative morphology. It is suf- ficient here to call attention to the fact that, like the cell, the telome is thought of as a ubiquitous structure in the plant. Not all of the discontent with the classical theory emanates from fol- lowers of the telome theory. My point is that the wave of reactivated dis- content began with the detailed anatomical analysis of structure called into operation by the telome theory and rests primarily upon evidence pertaining to an anatomical level of organization not adequately account- ed for in the classical theory of Goethe based on organography. It should be clear that the telome theory—namely, that plant struc- tures are compounded of telomes in accordance with the operation of the elementary processes — like the classical theory and the principle of homology is a system of logic. As a system of logic it is designed to inter- pret structures on the basis of criteria significant to telomes and the ele- mentary processes significant to their elaboration. The principle of homol- ogy, on the other hand, is a system of logic designed to interpret structures through the logic of comparison resting upon comparable criteria signifi- cant to morphological equivalence. The logical consequences of the prin- 1957 | MASON: CONCEPT OF THE FLOWER 85 ciple of homology are not to be judged by the criteria of the telome theory. These criteria are pertinent only to the logical system for which they were developed. There is nothing inherent in the telome theory that can invali- date either the classical theory or the principle of homology. We, as the designers of these systems of logic, can, if the facts warrant, say that one system explains the situation much better than the other and we may accept or reject on that basis. On the other hand we may alter the system of logic to make it more effective, and, if the altered system proves to be more efficacious, we may replace the old with the new. Having investigated the problem of the flower both by studying its presentation in the literature and by reviewing the structural features of a large number of different kinds of flowers, I have assessed this contro- versy as resting primarily in difficulties with our system of logic, in part with our failure to establish adequate diagnostic criteria for the categories into which we would classify the structures of the flower and in part with a confusion of description with ontogeny and phylogeny, rather than with any unusual difficulties inherent in floral anatomy. The chief anatomical difficulty is largely a matter of determining ‘when is a structure a new structure?’ meaning by this, when does it depart from being an integral part of the structure that bears it? This, I think, can be answered strictly within the framework of the principles of homology, viewing each case, to be sure, in its ontogenetic setting, but primarily viewing each structure for what it is. In discussing the concept of homology I shall employ a diagrammatic design (fig. 1) which may be spoken of as an organization system (Wood- ger, 1929). This is strictly a design for displaying an idea, and as I em- ploy it, it is not to be construed as meaningful to any other purpose. Its purpose here is to display the scope and detail of the application of the concept of homology as it appears serially in the plant, how homology is to be interpreted in cases of regeneration, and through this, how structures that may otherwise appear as anomalous may be effectively explained. From serial homology we may proceed to general homology, a transitive relation which rests upon the notion that “things equal to the same thing are equal to one another.” We will, however, use the term “equivalent” in the sense we attribute to Lindley, rather than the term “equal,” which implies detailed identity. I shall employ as exemplary the homologies of the appendages of the axis of the shoot and of the flower in the flowering plants. ORGANIZATION SYSTEM OF A HIGHER PLANT Whether an organism develops from a one-celled zygote, from a single meristematic cell, or from a group of meristematic cells, there is a pattern of organization resulting from its ontogeny that reflects the plasticity of the cell in its capacity for division and differentiation. This pattern of organization has a dual aspect. It is reflected first as an increase in struc- tural complexity brought about simply by the continued increase and 86 MADRONO [Vol. 14 H \ | / Fic. 1. Diagram of organization system resulting from increase in complexity due to cell division and cell differentiation. A to BC represents the complexity arising purely from the multiplication of cells. The three cell lineages A to DG, A to EH, and A to FI represent differentiation through the activities within each of the three meristem systems and represent a com- plexity resulting from cell differentiation. Regeneration homology is indicated by A’ B’ C’ where a new system is generated at the organ level. Interpretation is achieved by correlating A’ with A, B’ with B, and C’ with C. disposition of large numbers of cells, as A to BC (fig. 1), especially where the pattern results from different planes of cell division. The pattern of organization is reflected secondly as an orderly increase in complexity resulting from the differentiation of the meristem DEF and channelling the diversity into lineages of differing tissues, tissue systems, and organs ordered in accordance with their respective meristem origins, as A to DG, A to EH, and A to FI. Let us suppose that these represent the lineages originating in the three meristem systems from promeristem through pro- toderm, procambial strands, and ground meristem, and the cell and tissue lineages of each meristem system to its structural destiny. By superimpos- ing the idea of increased structural complexity through cell division on different planes over that of the results of meristematic differentiation, as shown in the system chart (fig. 1), we obtain the impression of an organ+ ization system that may be of great utility in the interpretation of com- parable structures on any given plant, as well as a basis for determining what on any given plant is comparable; through such serial comparison we may determine what structures on different plants are comparable. We may interpret the design as constituting a hierarchy of different levels of 1957 | MASON: CONCEPT OF THE FLOWER 87 organization in which each of the lower levels participates within the limits of its meristematic origin in the structural organization of each suc- cessively higher level, and at length each is an integral part of the organ- ismic whole. Analytically this presents a design in order approachable through mathematical induction. Aside from the cell division hierarchy and the resulting segregation mentioned above, it is also possible to trace in these same patterns hier- archies of structural specialization and of associated functional specializa- tion. And since each phylogenetic step of structural advance was accom- plished as a precise step in the ontogeny of an individual through the activity of mutagenic agencies or other cytological phenomena by which an individual deviated from its predecessors to the extent of this step, the hierarchy of organization also reflects, to a limited extent, through logical implication, the basis for a phylogenetic explanation as it relates to struc- tural pattern. The reasoning here is clear, but because of the nature of the facts it may be clearer than thé evidence. The reasoning is as follows. There is no step in phylogeny that structurally is not first evident as a step in the ontogeny of an individual regardless of its cause. So far as the empirical phenomena upon which these rest are concerned, and especially as phylogeny relates to any particular phylogenetic step, ontogeny and phylogeny are one. It is possible to abstract successive steps in phylogeny out of their context of successive ontogenies. We may then speak of a unitary phylogenetic sequence. Through the many changes that have in- volved the structural differentiation of plants, however, it is no longer possible to separate these as discrete historical steps in either phylogeny or ontogeny. Because of this, the ideally sound biogenetic law of Haeckel (1876) that ontogeny recapitulates phylogeny is difficult of application except in very general terms. The common properties of equivalent struc- tures must be assumed to represent their common ancestral connections from which they deviate in their separate ontogenies and phylogenies. It would seem, therefore, that the concept of morphological equivalence rests, to this extent, on the biogenetic law. In our organization diagram, therefore, as they relate to an individual, the older phylogenetic facts are probably to be found lower in the system than are the younger. (In this predecessor-successor relation of morphological facts based upon levels of identity, we see the basis for the connected system necessary for the logic of homology.) To my thesis, the operation of the biogenetic law is most clear when a preceding structure serves as a structural foundation for a phylogenetically following structure and must be resynthesized in each ontogeny in order that the following structure may be synthesized. This may be the full scope of the meaningfulness of the biogenetic law in plants. THE CELL AND ITS ROLE IN ONTOGENY Since each individual begins as a single cell and each step in the com- plexity and differentiation of the whole results from cell division into like or different kinds of cells, it will be apparent that all known operational 88 MADRONO [Vol. 14 dynamics in the structural ontogeny of the individual are on the level of dividing cells, plus any subsequent modifications of any individual cell. Dividing cells occur at nearly all structural levels of the hierarchy, and such meristematic cells, through the subsequent development of descend- ent cell lineages, may either carry the main organization system toward its completion or may initiate a cell lineage capable of developing a new structure that may be an ontogenetic departure from the structure in which the original cell arose. Thus at the junction of the petiole and the blade of the leaf of Tolmiea menziesu and in the crenations on the leaf margin of various species of Bryophvllum, buds develop that generate new plants as complete as those which develop from seed. Organographically these initiating meristems occur at a high level in the organization system. However, in the inception of this activity we must think of the meristems as returning to a lower level in the system and thus as regenerating a new system. It is a very important fact that neither the resulting system nor the structure produced in any way distorts the interpretation of the organ on which these initiating meristems arose. In the cases here mentioned the organs bearing the new “‘system” are still and always will be leaves. How and where new anatomical lineages may develop and to what extent they develop will depend upon the “‘special and local predisposing causes of ontogeny.” These predisposing causes may result in the produc- tion of such structures as a sporangium on a sporophyll, the extension of a structure by toral growth, the production of some teratological struc- ture, the development of a tissue, or the extension of a vascular system. If such a structure is defined in terms of its inherent properties, it is pri- marily significant for what it is. Only if it is defined in terms of some posi- tional relation is it ever significant for where it is. In some cases it may be significant to the defining type of the structure that bears it, as for instance, a megasporangium on a sporophyll is significant to this sporo- phyll being also a carpel; but neither what the sporangium is, nor where it is, is in any wavy significant to the interpretation of the carpel as being also an appendage. Such ontogenetic and phylogenetic facts are useful in the explanations of likenesses and differences by which we classify, but they neither serve to describe nor define the classes of like structures. As we contemplate these matters we are first concerned with the identity of the structure, then with its connected relations in a classified system of structures. We discover in the properties of our material the order and connectedness that relates the classes of structures. We then explain it with our notions of ontogeny and phylogeny. Our first task is to discover and describe and thus determine what are homologous structures. We move then to ex- planation. It would seem important, therefore, that if we are to seek to explain such structures and their relations, we must direct our attention first to the organization level of dividing and differentiating cells regardless of the higher organization level of which these cells may have been a part. 1957 | MASON: CONCEPT OF THE FLOWER 89 The primary fact is the capacity of a meristematic cell. The anatomical environment of this meristematic cell may influence the nature of what develops from it, but we must realize that when this development is com- plete the resulting structure may be categorically independent of the structure that originally produced the cell. What is important is the nature and identity of this resulting structure. Through the vagaries of meristematic initiation, such structures may develop anywhere in the organization system. Where the predisposing causes of their synthesis become organized in the gene pattern, they are to be regarded as normally a part of the ontogeny of the plant and may be represented by any tissue, tissue system, or organ, or by modifications or by parts of these. In effect, in their ontogenetic elaborations they consti- tute organization systems of their own and can be abstractly superimposed upon the general organization system and correlated in a one to one man- ner with comparable structures as they may occur anywhere on the plant (fig. 1, A’, B’, C’). Thus parenchyma is parenchyma, anywhere on the plant, irrespective of the structure that gives rise to its initiating cell or cells. We may think of it as being related solely to the predisposing causes of parenchyma, whatever that combination of conditions and events may be. Likewise any other structure as it may occur normally or abnormally is to be regarded primarily in terms of what it is rather than the nature of its meristematic origin. What is important in each of these cases is, ‘What is the structure that results from such activity?’ and not, ‘From where did it arise?’ THE CONCEPT OF HOMOLOGY As mentioned above, the concept of homology had its origin among the ancient geometricians and was applied to biology by the anatomical phi- losophers at the beginning of the last century. In biology it found its first linguistic form in the doctrine of metamorphosis of Wulff and Goethe. Thus homology is not solely a biological notion. It is to Owen (1848) that we turn for most of the current notions of homology as applied in biology. Owen thought of homology as a serial system, although he expressed him- self in a manner that introduced some confusion between his serial sysi:em and his example, namely the serial segmentation of animals to which he applied the notion. Homologies outside of a given serial system could be drawn wherever the necessary comparisons could be made and relations established. These relations served to connect the serial systems of diverse organisms. This Owen spoke of as establishing general homology. In view of current controversy it appears that some clarification of these notions is in order. Homology as employed in biology is an adaptation of the notions of set theory, relations theory, and the theory of types, together aimed at establishing the formal relations between the structures of organisms (Woodger, 1937). The relations that we in biology speak of as “homology” differ from ordinary relations because the field of the relations is the 90 MADRONO [Vol. 14 organism, which comprises a dynamic system. The relations thus imply the operations of the dynamic system that have brought them about. So compelling is this implication that to many it has become an integral part of the concept, with the dire consequences that the formality of the rela- tions is often confused by having some anticipated notion of the implica- tion built into it. However, the actual homology rests solely upon the formal relations in the serial system, and these should be established in- dependently of any implication to be derived from them or to be antici- pated in their behalf and independently of any explanation that may be forthcoming from our understanding or our misunderstanding. Serial homology actually is the serial system of class relations in ascend- ing order into which any given structure may be classified. For example, a petal of a flower belongs to the following series of classes:—petal, co- rolla, perianth, flower part, appendage. Each class level represents an in- crease in the scope of structures that are included as homologous organs. Each class expresses the homologous relation of its included terms. Thus a leaf and a petal bear the homologous relation “appendage” to one an- other. Each class at each level comprises a logical type. One may insert classes or subclasses into this system as occasion may demand so long as they are inserted in a proper serial order determined by the scope of the relations that are expressed. The number of properties and the nature of the properties upon which the class is based is not im- portant. Its position in the series is. This is because each such class stems from the logical conjunction of subclasses, and hence it represents the logical connection of the subclasses in series with the class. We thus have in the system the basis for establishing order and connectedness among the morphological properties of organisms. Through their connectedness their morphological equivalence is implied. As one approaches the system analytically, it will be found that the basis of the implication accrues by mathematical induction, each subclass at each level adding its increment to the logical type of its predecessor to establish its own logical type. And as each new logical type is thus estab- lished, the scope of the relations expressed is thereby narrowed. The com- pelling feature of the implication is enhanced by the orderliness of these accruing foundations of relations and the realization that they are found- ed in ontogeny. Also, as they relate to the interpretations of phylogeny they give some semblance of meaning to the biogenetic law in that the properties of each predecessor class must be synthesized in ontogeny as the foundation for the synthesis of the properties of successor classes. This is solely logical implication and not logical proof. The scope of the rela- tions of the more inclusive classes, however, is significant, since it serves to give a sequence of order to the properties. Because of the complex nature of higher plants, the serial system must be thought of in terms of various structural levels of homology rather than in terms of a single organographic series culminating in the individ- ual as the class that includes all. The organographic series provides a very 1957] MASON: CONCEPT OF THE FLOWER 91 incomplete concept of serial homology. It should be clear that homologies are possible on different levels of organization of ontogeny. The cytologist employs the concept at the level of cells and cell contents. Here homolo- gies of very broad scope are possible. The basis of homology rests in the identity of the structure, and it implies origin from a pre-existing similar structure. At a higher level in the system the anatomist employs the con- cept of homology for tissues and tissue systems. Here like or equivalent cells are organized into tissues arising from a common or equivalent meri- stem as they become a part of a higher organizational tissue system. In such cases the homology often becomes restricted within the scope of the ontogenetic lineage cut off in the differentiation of the promeristem into protoderm, procambium strands, and ground meristem. A third level of homology may be spoken of as the organographic level involving identity and positional relations of organs and appendages. Here we are concerned with cells, tissues, and tissue systems constituting organs and appendages as these arise similarly from like meristems and are ar- ranged upon a common plan. As we ascend the levels of structural complexity, the scope of possible homologies is reduced and becomes more localized as to areas of the plant and involves fewer serial structures. Likewise, in general homology, the higher the organizational level the fewer are the homologies that can be broadly drawn. Thus, on the organographic level one is less able to draw broad homologies than one is on the level of tissues and cells. For in- stance, carpels are confined to the angiosperms, tracheids are present in several classes of plants, while cells are characteristic of all plants. Because of the capacity of dividing cells to generate structures morpho- logically unrelated to the organ that bears them, one must recognize this state of affairs in his concept of homology. It calls for an assessment of the situation in terms of what structures result and how they may be cor- related with the organization system as a whole. The homologies of such structures become evident through their identity and their assignment to the proper level in the serial system of the whole. For such homologies I shall use the term “‘regeneration homology.” Any deviation that is evident in a practical application of this system must seek its explanations in regeneration homology. This will be particu- larly evident when a structure attains properties assignable to a structure other than that on which it occurs. It will also be evident when a tissue occurs in the confines of a meristem system in which such tissues do not ordinarily arise. THE NEED FOR CLEAR DEFINITIONS In the practical application of homology to the problems for which the logical system is devised, it becomes imperative that such structures and abstractions employed be clearly as well as inclusively and exclusively defined. The chief problems in establishing homology are precisely trace- able to these difficulties of definition or to semantic problems in not under- 92 MADRONO [Vol. 14 standing the level of abstraction and the resulting error as to what criteria are involved. Whereas a sporangium may be significant to the concept “carpel,” it is not significant to the concept “appendage” and has nothing to do with the inclusion of carpels as appendages. There may be many other features of carpels that may be significant to the concept of carpel or significant to the carpel of some particular kind of plant, but they play absolutely no role in the abstractions of the category “appendage” which are based upon the common characters of kinds of appendages. We are not concerned with the characters of carpels whereby they may be unlike other appendages. We are only concerned whether or not they possess the characters that are diagnostic of the abstract concept ‘‘appendage.” To the extent that these characters are present, we classify the carpel as an appendage and to this extent it fulfills the classical theory as amended by Lindley. To this extent also, the assignment of the carpel to the concept “appendage” constitutes another relation step in our logical system. We must be aware, however, that logic is not self-validating, and in this case it is no more valid than is the significance of the criteria of morphological equivalence that we have accepted. The logical system is only a method of handling the facts that we accept as valid. We cannot hold the system responsible for the validity of these facts. Our logic cannot correct our mistakes, although sometimes it may assist in pointing them out. When we find carpels that display characters that confuse us as to what structural category in which to include them, it seems to me necessary to investigate whether or not we are dealing with a simple situation or with a phenomenon that may be better handled through the logic of regeneration homology. Perhaps the occurrence of sporangia on sporophylls is such a case. Some seem to think that a sporangium is normally terminal on a cauline structure. Should this be true the problem of sporangia on append- ages fits naturally into the logical system of regeneration homology. We would presume that through the predisposing causes of ontogeny a meri- stematic event takes place that generates a sporangium on an appendage, and these facts make of the appendage a sporophyll. When provision for the synthesis of the predisposing causes becomes organized in the gene system, the production of sporangia on appendages is normal. Such an event in no way destroys the homologies of the sporophyll as an append- age. It isin every way comparable to the occurrence of buds on the margin of the leaf of Bryophyllum. I deliberately employed the sporangium in this example because it was easy to discuss and seemed to fit in well as an example. There are other stem-like tissues reported in carpels that have raised questions of their reference to appendages. The explanations of these should be sought upon an anatomical level of homology and evidence sought as to what struc- tures are represented and what has been their ontogenetic history. Next the problem is to find the extent of their organization and what their re- sulting homologies may be. The fact that we seek identity in the old does not mean that new struc- 1957 | MASON: CONCEPT OF THE FLOWER 93 tures might not appear and that new structures are not possible in the gynoecium. It does mean that before we can assign them as being new structures, we must have some criteria other than our confusion in the interpretation of the old, upon which to base our decision. A new struc- ture is one which does not possess the diagnostic criteria that would per- mit it to be classified in any existing category of structures at the same organization level. Much of this confusion stems from lack of clarity as to what constitutes any particular organ or structure. For example, much of the problem of the inferior ovary stems from the fact that we do not have an adequate definition of a receptacle drawn from characters that are diagnostic of a receptacle. The result has been that we seek evidence of receptacles in criteria that are not diagnostic of such a structure, as, for example, the employment by Smith (1943) of recurrent bundles in the structures sur- rounding the ovary in Santalaceae as evidence that this structure is a receptacle. If the structure surrounding the inferior ovary and traversed by the recurrent bundles in the Santalaceae is a receptacle, it must be so on some other evidence, for recurrent bundles are not diagnostic of recep- tacles, nor are branched or unbranched bundles so diagnostic. In seeking diagnostic criteria for the receptacle, about all that I can find common to the several hundred kinds of flowers I have examined is that the recep- tacle of all of them bears flower parts and is at least operationally terminal on an axis. I have seen no evidence whatsoever in the vascular system that provides for a universally applicable set of criteria by which one might recognize a receptacle as different from any other structure re'ated to stems. We must bear in mind that the plant does not define “receptacle.” We as humans define it, and we define it for our own devices. ON THE ROLE OF THE TELOME If the telome-mesome complex is significant in the interpretations of morphology, it cannot be employed in any problem of diagnostic com- parison that employs the logic of homology above the level of meristem differentiation in the organization system. This is because the telome, like the cell, is a ubiquitous structure. It is presumed to be a characteristic of everything and therefore cannot be diagnostic of anything. The logic of homology follows strictly the pattern of diagnostic comparison. To be significant, an attribute must be diagnostic. The cell is significant to ho- mology only on a cytological level where cells are compared with cells. If the telome is significant to homology, it can be significant only on levels where telomes are significant to the diagnosis of the structure. It may be possible to make such diagnoses on the basis of kinds of telome systems if these are valid structures. CONCLUSION This reorganization of our logical system I believe will give us at least a sound working hypothesis for the flower which we can use in developing 94 MADRONO [Vol. 14 our concepts of phylogenetic taxonomy. The concept of homology, be- cause of its ontogenetic implications and because phylogeny also has ontogenetic implications, provides us with the necessary basis for handling the problems of likenesses and differences as they involve comparable structures. There is no problem of phylogenetic divergence, as it is sub- ject to investigation in taxonomy, that is not pursued from the point of view of the logic of the concept of homology as it relates to the properties of organisms. When we turn from the plant to contemplate what we have seen and to interpret its significance, we must inevitably rely either upon the intuitive judgments of what appear to be immediately self-evident facts, or we automatically resort to a logical system based upon reasoning. The one is the intellectual foundation of sight recognition, the other the intellectual foundation of identification. Both have an important place in interpretive science. Neither is self-validating and therefore either may lead to erro- neous judgments. The security of the one rests in the validity of the imme- diate self-evidence. The security of the other, as it relates to our problem, rests first in the validity of the logical system and then in the validity of the diagnostic criteria. It therefore becomes important that we develop an adequate logical svstem such as we strive for in perfecting the system of homology. It is also important that we seriously re-examine the validity of the diagnostic criteria of the structures and the abstractions that we employ as signifi- cant to our interpretations. Upon this will rest the validity of their self- evidence, so important to our intellectual manipulations. We must not assume that simply because we may have more detailed and complete facts, as important as this is, that we can avoid operating within the framework of intuitive judgment and logical systems in our interpretation. If we had all of the facts of structure and ontogeny, we would only shift in our intellectual contemplation of them from a prepon- derant leaning upon logical systems to a heavier reliance upon intuitive judgment. This is because sight recognition would play a greater role. The validity of our judgments will still rest upon individual human capacities for discrimination as to significance. We still will be plagued by the curse of him who, without understanding them, employs the faulty judgments made by himself or others. The plant is responsible for what is there. What the plant and its structures mean to us is our responsibility, and it is not solely a responsibility of discovery, important as this is, because discov- eries must be interpreted to be understood. Department of Botany, University of California, Berkeley. 1957 | HEAD: GENUS ASTRAGALUS 95 LITERATURE CITED Boxe, Norman H. 1947. Development of the adult shoot apex and floral initiation in Vinca rosea L. Am. Jour. Bot. 34: 433-439. Cuvier, Georces. 1800. Lecons d’anatomie comparée. Tome i, ii. Paris. GorTHE, J. W. 1790. Versuch die Metamorphose der Pflanzen zu erklaeren. Gotha. GREGOIRE, V. 1938. La morphogénése et l’autonomie morphologique de lappareil floral. I. Le carpelle. La Cellule 47: 287-452. HaEcKEL, Ernst H. P. A. 1876. The history of creation. Transl. by E. Ray. New York. Lam, H. J. 1948. Classification and the new morphology. Acta Biotheoretica 8: 107- 15S: LINDLEY, J. 1838. Botany. Library of Useful Knowledge. London. OKEN, Lorenz. 1807. Programm, uber die Bedeutung der Schadelknochen. Jena. OweEN, RICHARD. 1848. The vertebrate skeleton. London. Samassa, Pau. 1896. Caspar Friedrich Wulff’s Theoria Generationis. Leipzig. Smi1TH, Frank H. & Suir, E. C. 1943. Floral anatomy of the Santalaceae and some related forms. Oregon State Monographs, Studies in Botany. The College Press. Corvallis. Sr. HinarreE, GEOFFREY. 1807. Considérations sur les pieces de la téte osseuse des animaux vertebres, et particuli¢rement sur celles du crane des oiseaux. Ann. Mus. Hist. Naturelle, Paris. 10: 342-365. Tuomas, H.H. 1934. The nature and origin of the stigma. A contribution towards a new morphological interpretation of the angiosperm flower. New Phytol. 33: 173-198. Tuompson, J. M. 1935. The acarpous nature of modern flowering. Proc. Sixth Internat. Bot. Congress 2: 122-124. Wooncer, J. H. 1929. Biological principles. New York. . 1937. Axiomatic method in biology. Cambridge Univ. Press. ZIMMERMANN, WALTER. 1930. Die Phylogenie der Pflanzen, ein Uberblick iiber Tatsachen und Probleme. Jena. MITOTIC CHROMOSOME STUDIES IN THE GENUS ASTRAGALUS! S. CONRADE HEAD The genus Astragalus L., tribe Galegeae of the Leguminosae, consists of about 1,500 species occurring in northern Africa, Europe, northern and central Asia, and in the western hemisphere. Some sixty genera have been proposed as segregates from it, and several taxonomic revisions of the genus or parts of it for North America, based on morphological characters, have been presented (Jones, 1923; Rydberg, 1929; Barneby, 1945, 1947, 1949, 1956). Of these, the more conservative treatments of Jones and Barneby have been found more practical for the purposes of this study. Very little, however, is known about the cytology of this genus. Accord- ing to Senn (1938), “Only two per cent of the species of the huge genus Astragalus have been studied. These species are based on an 8 series with 1 This paper represents a portion of a thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Botany at the State College of Washington, Pullman, 1955. 96 MADRONO [Vol. 14 two exceptions in which n—14. There are 16 diploids, 1 tetraploid, 2 hexa- ploids, and 1 octaploid. Considering that the species studied come from widely separated regions scattered over Europe and Asia, this is a remark- able consistency of chromosome number.” According to Tischler (1938), the findings of ten workers for forty-four Old World species and four New World species were: 2n—16, thirty-three species; 2n—32, three species; 2n=48, four species; 2n—64, five species; 2n—=28, 36 and 96, one species each. Vilkomerson (1943) made a survey of twenty-six species from west- ern United States and found that for eleven 2n=24, for thirteen 2n=22, whereas the other two had 2n numbers of 16 and 44 respectively. James (1951) gave chromosome counts for three species, one each of 2n=22, 24 and 26. These several surveys account for approximately one hundred spe- cies of Astragalus. Certainly the consistent chromosome number stressed by Senn no longer holds. It was with the thought of adding to the chromo- somal information for this genus that the present investigation was under- taken. The author wishes to express his appreciation to Dr. Adolph Hecht, who served as advisor during the course of the research and who with Dr. Marion Ownbey kindly offered many suggestions during the preparation of the manuscript. Mr. Robert C. Barneby provided several of the col- lections reported here and checked many of the determinations. Mr. Ralph D. Amen offered many valuable suggestions concerning cytological methods. METHODS The Astragalus collections studied are listed by species in Table 1. The source of the collection, the chromosome number, and the figure number (for those collections illustrated) are given. Voucher specimens are filed in the Herbarium of the State College of Washington. The plants were grown in the greenhouse and later transplanted to an experimental garden at Pullman, Washington. Seeds had to be scarified either by filing or by use of concentrated sulfuric acid, with treatment in the acid from forty-five minutes to one hour being most satisfactory. The scarified seeds were placed on wet filter papers in Petri dishes until the primary root had reached a length of about fifteen mm.; the root was then removed and placed in Belling’s ‘“‘“metaphase” modification of Navashin’s solution. Root tips were also obtained from pot-bound plants. Some plants were transplanted from their natural habitat to the greenhouse and later to the garden. Since most persons preparing herbarium specimens rarely collect mature fruits, herbarium sheets did not prove to be a profitable source of seed. A few seeds were obtained from herbarium sheets, however, and those as old as nine years germinated without great difficulty pro- vided they were mature when collected. Paraffin sections cut at twelve microns as recommended by Senn were prepared and stained by the crystal violet-iodine method. These prepara- tions were not as satisfactory as those obtained by a method worked out 1957 | HEAD: GENUS ASTRAGALUS 97 by Amen (unpublished).* This method provides excellent permanent slides with the cells separated from each other. One has little difficulty in viewing separated cells under the microscope, and the observer is certain that the cells are uncut. Amen plans to publish his method in detail. My modification of his procedure is as follows: Fix cut root tips in Belling’s “metaphase” modification of Navashin’s solution pre- ferably for at least two days; remove, rinse several minutes in tap water, blot off ex- cess water; place on slide in one drop of Haupt’s adhesive; cut apical 2 mm. into several pieces and squash, using flat side of scalpel; air dry slide about 15 minutes; stain in 1 per cent methyl violet for 10 minutes, wipe off excess stain and nearly air dry; wash momentarily and again nearly air dry; place in solution of 8 grams of picric acid powder dissolved in 1 liter of 95 per cent alcohol for about 20 seconds; blot excess 1 to 2 seconds, place in absolute alcohol about 20 seconds; clear in 50 parts absolute alcohol, 25 parts xylene, and 25 parts clove oil for 3—7 minutes; pass through 2 changes of xylene; mount in piccolyte. Slides were examined and camera lucida drawings were made of the metaphase plates using a Zeiss microscope with an apochromatic oil- immersion lens of N. A. 1.30 and an initial magnification of 2,250 times. The figure were drawn at approximately 4,350 times and reduced to 1,450 times in reproduction. TABLE 1. CHROMOSOME NUMBERS OF ASTRAGALUS COLLECTIONS STUDIED CHROMOSOME FIGURE SPECIES NUMBER (2n) NUMBER SOURCE SECTION HoMALOBI® A, stenophyllus T. & G. 24 1 Oregon, Morrow County: 12.9 miles southwest of Heppner, Head 598. 24 2 Oregon, Baker County: 12.5 miles southeast of Baker on the Ebell Creek Road, Head 609. 24 3 Oregon, Wheeler County: 16 miles south of Condon, Head 600. SECTION INFLATI A. lentiginosus Dougl. ex Hock. var. lentiginosus 22 4 Oregon, Baker County: 1 mile east of Quartz, Head 607. A. cusickii A. Gray 22 5 Washington, Asotin County: near the Grande Ronde River bridge, Head 569. 22,44*4 36 Oregon, Baker County: 13 miles west of Richland, Head 611. 2 Amen, Ralph D., former graduate student, State College of Washington. Present address: 2426 South University, Denver, Colorado. % Sections are those listed by Jones (1923) although this arrangement is not al- ways satisfactory. + Diploid and tetraploid cells occur in the same root tip of many Leguminosae. See discussion on polysomaty. Such counts are indicated by an asterisk. 98 SPECIES MADRONO CHROMOSOME NUMBER (2n) [Vol. 14 FIGURE NUMBER SOURCE A. beckwithii T. & G. var. wezserensis M. E. Jones A. allochrous A. Gray SECTION COLLINI a2 22 A. collinus (Dougl. ex Hook.) G. Don var. collinus var. laurentii (Rydb.) Barneby SECTION HAMOSI A. andersonii A. Gray A. arthuri M. E. Jones A. congdoni S. Wats. SECTION PopO-SCLEROCARPI A. sclerocarpus A. Gray A. pachypus Greene SECTION REVENTI-ARRECTI A. arrectus A. Gray A. sheldonii (Rydb.) Barneby A. riparius Barneby 24, 48* 24 24 24 24 26, 52* a2 24 24 24 6 8; 37 10 11 13,39 14 lS) 16 17 18 Idaho, Owyhee County: 10 miles north of Silver City, on road to Murphy, Christ 19537. New Mexico, Grant County: San Lorenzo, Barneby 11172. Washington, Asotin County: 5.5 miles northeast of Anatone, Head 585. Washington, Asotin. County: 6.3 miles northeast of Anatone, Head 588. Oregon, Morrow County: 18.6 miles east of Heppner, Head 596. Nevada, Washoe County: 6 miles northwest of Univ. of Nevada Campus, Reno, Ownbey 2925. Washington, Asotin County: 3.4 miles northeast of Ana- tone, Head 587. California, Fresno County: Piedra, Barneby 11417. Washington, Benton County: 2 miles west of Enterprise (West Richland), Head 525. California, Kern County: Caliente, Barneby 11370. Washington, Whitman Coun- ty: Prairie Strip, Botany Dept. State College of Washington, Pullman, Head 584. Washington, Asotin County: 3.4 miles northeast of Ana- tone, Head 586. Washington, Whitman Coun- ty: 3.3 miles northeast of Wa- wawai, Head 562. 1957] HEAD: GENUS ASTRAGALUS 99 CHROMOSOME FIGURE SPECIES NUMBER (2n) NUMBER SOURCE 24 19 Washington, Whitman Coun- ty: 1.1 miles east of Wawa- wai, Head 563. A. conjunctus S. Wats. 24 20 Oregon, Wheeler County: 16 miles south of Condon, Head 599. A. eremeticus Sheldon var. malheurensis (Heller) Barneby 24 21 Idaho, Washington County: just north of Weiser, Ownbey 27061. SECTION ULIGINOSI A. canadensis L. var. mortoni (Nutt.) S. Wats. 16 22 Washington, Whitman Coun- ty: north slope of Kamiak Butte, Head 613. SECTION CHAETODONTES A. spaldingit A. Gray 24 23 Washington, Whitman Coun- ty: % mile east of Lacrosse, Head 582. SECTION ARGOPHYLLI A. inflexus Dougl. ex Hook. 22 24 Washington, Whitman Coun- ty: 1 mile northeast of Wa- wawai, Head 499. A. purshtt Dougl. ex Hook. var. glareosus (Dougl. ex Hook.) Barneby 22,44* 25,38 Oregon, Baker County: 1 mile east of Quartz, Head 547. 22 26 Oregon, Morrow County: 18.6 miles east of Heppner, Head 595. 22 27. Oregon, Grant County, 2.5 miles north of Mt. Vernon, Head 603. 22 28 Oregon, Grant County: 2.4 miles north of Mt. Vernon, Head 604. var. purshi 22 29 Washington, Whitman Coun- ty: top of Steptoe Butte, Head 580. A. chamaeleuce A. Gray 22 30 Colorado, Mesa County: 3 miles south of Fruita, Weber 3782. 100 MADRONO [Vol. 14 ; CHROMOSOME FIGURE SPECIES NUMTCER (277) NUMBER SOURCE A. cibarius Sheldon a2 31 Idaho, Bannock County: 12 miles south of Portneuf, Christ 19933. SECTION MALACI A. succumbens Dougl. ex Hook. 24 32 Washington, Walla Walla County: 7.4 miles east of Wallula, Head 539. SECTION MOLLISSIMI A. mollissimus Torr. var. earlei (Rydb.) Tidest. 24 33 Texas, Jeff Davis County: southeast of Fort Davis, Barneby 11129. SECTION SARCOCARPI A. gypsodes Barneby 24 35 New Mexico, Eddy County: southwest of Whites City, Barneby 11138. SECTION UNDETERMINED A. diaphanus Dougl. ex Hook. 28 34 Oregon, Wheeler County: 2 miles east of Service Creek, Hitchcock 19235. -t DISCUSSION As the table indicates, chromosome numbers of 2n=16, 22, 24, 26, and 28 were found in the plants studied. The sections /nflati, Collint, Podo-sclerocar pt, Reventi-arrecti and Argophylli showed constant chrom- osome numbers. The section Hamosz had two different chromosome num- bers represented; A. andersonu and A. arthuru both had 2n=—24, and A. congdonu, 2n—26. A like situation was reported by Vilkomerson (1943) for the section Galegiformes. Even prior to her publication, the need for a taxonomic revision considering physiological evidence was sug- gested by Trelease (1942). Vilkomerson also reported a chromosome number of 2n—22 for A. crassicarpus Nutt.; A. gypsodes is recorded above as having 2n—24. Barneby (1956) groups these two species to- gether in the same section. James (1951) found three different chromo- some numbers represented by three species in the section Didymocar pt. Thus we see it is possible for a section to have species with different chromosome numbers. Yet, as cytological information accumulates for the genus, more sections are found to have a constant chromosome num- ber. Much more study is needed in the section Hamosi and, as Trelease mentioned, in the Galegiformes. Certain species of Astragalus can be readily identified by their charac- teristically shaped chromosomes. Among these are A. succumbens with a pair of large “question mark” chromosomes and A. mollissimus var. earlet 1957 | HEAD: GENUS ASTRAGALUS 101 with its eight pairs of ““C” chromosomes. The sections Reventi-arrecti and Argophylli may also be recognized by chromosome similarities of the included species. 7 yr ~~] 8 FvSr Me FF hip Gee i ie Se “A ¢-* NS 3 4 1 2 a a > ~ INN Ve Oat OSGRK, SUS ae a ‘N y Son : ICNG= “LMA YOU sr w 9 w P ae I ie “ais INS A SY! 9 Ke) 1] 12 i Sl Sto_ 4, a J oo m9 4 cl ox ‘ZU \Sys (of J fasve CO) (\ ra ha A ( c tcl d, ¢) i : 3 16 Ss In MAL (, 2(NIC = We je x a \ aS 3 \7 18 Fe 20 Fics. 1-20. Chromosomes of Astragalus. 1-3, A. stenophyllus; 4, A. lentiginosus var. lentiginosus; 5, A. cusickii; 6, A. beckwithii var. weiserensis; 7, A. allochrous ; 8-9, A. collinus var. collinus; 10, A. collinus var. laurentii; 11, A. andersonit; 12, A. arthurii; 13, A. congdonii; 14, A. sclerocarpus; 15, A. pachypus; 16, A. arrectus; 17, A. sheldonit; 18-19, A. riparius; 20, A. conjunctus. Camera lucida drawings, & 1450. 102 MADRONO [Vol. 14 SECTION HoMALoBI (figs. 1-3, idiograms 1—3 ).—Geographical distribu- tion of A. stenophyllus appears to have little correlation with chromosome morphology in this species. Figures 2 and 3 are from plants which grew about two hundred miles apart, yet the chromosomes appear more alike than those of figures 1 and 3 which are from plants separated by only a few miles. SECTION INFLATI (figs. 4-7, 37; idiograms 4—7).—In all of the /nflati so far studied the 2n number is 22, provided A. diaphanus is not referred here. However, the section as a whole cannot be characterized or identi- fied on the basis of chromosome similarity, for the positions of the centro- meres are not as consistent as in those groups already mentioned. Both A. allochrous and A. cusicku have four pairs of chromosomes with nearly median centromeres which take a characteristic ““C” shape. A. beckwithu var. weiserensis has but one pair of these chromosomes. Astragalus allo- chrous (fig. 7) has the largest chromosomes of any found in this study. SECTION COLLINI (figs. 8-10, idiograms 8—10).—In contrast to the low correlation of chromosome morphology with geographical distribution in A. stenophyllus of section Homalobi, here there is much similarity in chromosome morphology from plants separated by even greater distances. SECTION Hamost (figs. 11-13, idiograms 11-13).—The two species with the 24 chromosomes, A. arthurit and A. andersonii, have little in common with the 26 chromosome species 4A. congdonit. The latter (fig. 13) has five pairs of “C’’-shaped chromosomes, while the former two species have only two pairs. Astragalus arthurii is unique in that one chromosome (the last in idiogram 12) shows a prominent constriction at about the middle, which might be the centromere region. Chromosomal data beyond that now available should be obtained before a revision of the Hamosi is attempted. SECTION PODO-SCLEROCARPI (figs. 14-15, idiograms 14—-15).—Vilko- merson studied nine species belonging here, including A. sclerocarpus, and found the same chromosome number (2n=22) in all. A. pachypus is the only new report for a species of this section. Unlike A. sclerocarpus (idiogram 14), A. pachypus (idiogram 15) has several pairs of ‘‘C’’- shaped chromosomes. SECTION REVENTI-ARRECTI (figs. 16-21, idiograms 16—21).—_Members of this section have very similar chromosomes. Each of the five species studied has four pairs of “C’-shaped chromosomes. Although these chromosomes vary somewhat in length they are otherwise very similar. Astragalus eremeticus var. malheurensis (idiogram 21) differs somewhat from the others by having a pair of small “dot” chromosomes not found elsewhere in this section. SECTION ULIcINosI (fig. 22, idiogram 22).—Only one species of the Uliginosi has been studied, A. canadensis L., reported by Vilkomerson, and A. canadensis var. mortoni of this study. SECTION CHAETODONTES (fig. 23, idiogram 23).—A. spaldingu is the only member of this section thus far studied. 1957 | HEAD: GENUS ASTRAGALUS 103 SECTION ARGOPHYLLI (figs. 24-31, 38; idiograms 24—31).—The Argo- phylli, as a section, show a close likeness in chromosome morphology and number. This similarity is perhaps to be expected with closely related species such as A. inflexus and A. purshu, but it would not necessarily SPC hd ECGeee| of UC Cree ee CIC CVEOKE CEH LCE CO CEECEEe s(pcarpeceece!’ 27 EC OCU Coe se ACCOM cece td 26 (09 OC 6 s(C(ldiCccece va | (CO UCCCCES Checargenrc< {UCC( reece es a Rr ry o€(4C CHUCK OCH LOL EK nD cereal GEUOES 568 (Gta are et NQCCRP SFDC ECE of O(aHert ee C APIS COCO CEE a( CECE (cece BCL CH eCcecer ah (Prd hecges AIC CONC Cece sl (CCgceeeeec 7 UUs Pre cere «Cn UGHE Cares ePCEUC HEC Cee AAKUSECCEE IptocRAMsS 1-35. Chromosomes of Astragalus. 1-3, A. stenophyllus; 4, A. lentzi- ginosus var. lentiginosus; 5, A. cusickii; 6, A. beckwithii var. weiserensis; 7, A. allo- chrous; 8-9, A. collinus var. collinus; 10, A. collinus var. laurentii; 11, A. andersonii; 12, A. arthuri; 13, A. congdonit; 14, A. sclerocarpus; 15, A. pachypus; 16, A. arrectus; 17, A. sheldonii; 18-19, A. riparius; 20, A. conjunctus; 21, A. eremeticus var. mal- heurensis; 22, A. canadensis var. mortonii; 23, A. spaldingii; 24, A. inflexus; 25-28, A. purshii var. glareosus; 29, A. purshii var. purshii; 30, A. chamaeleuce ; 31, A. ciba- vius; 32, A. succumbens; 33, A. mollissimus var. earlei; 34, A. diaphanus; 35, A. gypsodes. Camera lucida drawings, * 1450. 104 MADRONO [Vol. 14 ane eA = ag ons com | iS ee —< a I~ sat "f WS we ~\ A SUK MWs, AFL Sa\~ & Vn > I99™ LIYE, AS 24 25 aS 2 wT ow INN (5, slo> JY ee DAP — FIG INDE a ' a4 ~, SJN a TT te wD J > 3 AT~ 32 x 33 9 J~ ‘ ae ww TS a! ht, mn Veatn VIS Oye 34 BS) NPC KE &NS VS) ~ rn ~ a) Dr e 4— SER SNELL Sve Z =f Y a var ran 4 “es Cc wy &) (“ er 9S a 9 U 39 Fics. 21-39. Chromosomes of Astragalus. 21, A. eremeticus var. malheurensis ; 22, A. canadensis var. mortonit; 23, A. spaldingii; 24, A. inflexus; 25-28, A. purshii var. glareosus; 29, A. purshiit var. purshii; 30, A. chamaeleuce; 31, A. cibarius; 32, A. succumbens; 33, A. mollissimus var. earlei; 34, A. diaphanus; 35, A. gypsodes; 36, A. cusickii; 37, A. collinus var. collinus; 38, A. purshii var. glareosus; 39, A. congdoni. Camera lucida drawings, 1450. 1957 | HEAD: GENUS ASTRAGALUS 105 extend to such distant species as A. chamaeleuce from Colorado or to A. cibarius from southeastern Idaho. It would be interesting to determine if this similarity is maintained throughout this large section. Astragalus cibarius is excluded from the Argophylli by Barneby (1947), and he sug- gests a relationship with the Malaci for this species. On the basis of chromosome morphology and number, however, it seems very much like the other ArgopAvlli and very little like the only representative of the Malaci, A. succumbens, thus far studied. SEcTION Matact (fig. 32, idiogram 32).—Astragalus succumbens has one pair of large ““C”’-shaped chromosomes and a pair of “‘question mark”- shaped chromosomes. SECTION Mo ttissimI (fig. 33, idiogram 33 ).—Astragalus mollissimus var. earlez has eight pairs of ““C’’-shaped chromosomes. SECTION SARCOCARPI (fig. 35, idiogram 35).—A stragalus gvpsodes has 11 pairs of relatively long chromosomes and 1 pair of very short ones. Astragalus diaphanus (fig. 34, idiogram 34) has not been determined as to section. This species stands alone among the North American species studied in that the 2n chromosome number is 28. The chromosomes are also the smallest observed in this study. Jones (1923) listed A. diaphanus as a variety of A. lentiginosus, a member of his section /nflati. Barneby (1945) excluded A. diaphanus from his section Diplocystium (composed of the varieties of A. lentiginosus), but did not propose a new status. A. diaphanus should be excluded from the /nflati on the bases of chromosome number and fruit morphology. These reasons also support Barneby’s ex- clusion of it from the Diplocystium. PoLtysomatTic CEtts. In Astragalus, as in many other genera of the Leguminosae, both diploid and tetraploid cells may be found in the same root tip. Vilkomerson reported polysomaty in three species, but listed its occurrence as rare. Polysomatic cells were found by Tschechow (1930) in two of the species he studied. In one of these, A. candidissimus, tetra- ploid cells were found in forty per cent of the metaphase plates. Polyso- matic cells were observed in four of the taxa of the present study: A. cusicku, A. purshi var. glareosus, A. collinus var. collinus, and A. cong- dont (figures 36, 37, 38 and 39). In the last three species the occurrence of such cells are rare, but A. cusickii had about the same percentage of tetraploid cells found by Tschechow in 4. candidissimus. SUMMARY Mitotic chromosome studies were made of twenty-six species of Astra- galus represented by thirty-five collections. The 2n chromosome numbers of 16, 22, 24, 26 and 28 were found. The basic number of 14 is added to those previously reported for the North American species. Chromosome numbers for species of the sections Homalobi, Collini, Hamosi, Reventi- arrectt, Argophylli, Chaetodontes, Malaci and Mollissimi are reported for the first time. Counts for three species substantiate those previously pub- lished. Certain species and some sections of the genus can be readily rec- 106 MADRONO [Vol. 14 ognized on the basis of chromosome morphology. Astragalus diaphanus should be excluded from the /nflati on the basis of chromosome number and morphology. Department of Botany, Oregon State College Corvallis, Oregon LITERATURE CITED Barnesy, R. C. 1945. Pugillus Astragalorum IV: The section Diplocytium. Leafl. West. Bot. 4:65-147. . 1947. Pugillus Astragalorum VII: A revision of the Argophylli. Am. Mid]. Nat. 37:421-516. . 1949. Pugillus Astragalorum X: New species, varieties and combinations. Am. Midl. Nat. 41:496-502. . 1956. Pugillus Astragalorum XVIII: Miscellaneous novelties and reap- praisals. Am. Mid]. Nat. 55:477-503. JAMES, Lots E. 1951. Observations on the taxonomy of Astragalus, subgenus Hes- perastragalus. Contr. Dudley Herb. 4:63-72. Jones, M. E. 1923. Revision of North-American species of Astragalus. Salt Lake City; Utah. RypBerc, P. A. 1929. “Astragalanae” in N. Am. Flora 24:251-462. SENN, Harotp A. 1938. Chromosome number relationships in the Leguminosae. Bib- liographia Genetica 12:175—336. TISCHLER, G. 1938. Pflanzliche Chromosomen-zahlen. IV. Tabulae Biologicae 16: 162-218. TRELEASE, SAM F. 1942. Identification of selenium indicator species of Astragalus by germination tests. Science 95:656—-657. TscHECHOW, W. 1930. Karyologisch-systematische Untersuchung des Tribus Galegeae Fam. Leguminosae. Planta 9:673-—680. VILKOMERSON, Hitpa. 1943. Chromosomes of Astragalus. Bull. Torrey Club 70: 430-435. INNOVATIONS IN DUDLEYA REID MorRAN As a thesis at the University of California, I prepared a revision of Dudleya (Crassulaceae). This revision is not yet ready for publication and may not be ready for several years. Meanwhile, two floras including Dudleya are nearly completed, and there is immediate need for certain names from the thesis. Therefore, one new subspecies will be described and several new combinations proposed. Abbreviations for the names of herbaria are according to Lanjouw and Stafleu (1956). DUDLEYA ABRAMSII Rose subsp. murina (Eastwood) Moran, comb. nov. Dudleya murina Eastwood, Proc. Calif. Acad. IV. 20: 147. 1930. DUDLEYA CyMOSA (Lemaire) Britton & Rose subsp. gigantea (Rose) Moran, comb. nov. Dudleya gigantea Rose in Britton & Rose, Bull. N.Y. Bot. Gard. 3: 23. 1903. DUDLEYA CyMOSA (Lemaire) Britton & Rose subsp. marcescens Moran, subsp. nov. A subspeciebus ceteris caudicibus tenuioribus, rosulae 1957] MORAN: DUDLEYA 107 foliis minoribus aestate marcescentibus, inflorescentiis simplicioribus differt (fig. 1). Caudex 1-3 cm. long, 2—7 mm. thick, often branching; rosettes 3—6 cm. wide, of 8-12 (—15) leaves; rosette leaves green, oblanceolate, acute to subobtuse, 114-3 (—4) cm. long, 5-12 mm. wide, 1-2 mm. thick; floral stems 4—10 cm. tall, their leaves deltoid-lanceolate, 12-114 cm. long; in- florescence of 1—2 cincinni, each 1-3 cm. long and with 2—5 flowers; pedi- cels erect, 5-12 mm. long; sepals deltoid, acute, 2’2—4 mm. long; petals bright yellow often marked with red, 10-14 mm. long, 214-34 mm. wide, connate ca. 114 mm. Fic. 1. Dudleya cymosa (Lemaire) Britton & Rose subsp. marcescens Moran, subsp. nov. (type collection) x 0.6. Type: Shaded rocky slope by the creek, Little Sycamore Canyon, Sierra Santa Monica, Ventura County, California (near 34° 05’N, 118° 57’W), at about 330 meters elevation, May 28, 1948, Moran 3078 (UC 917950). Specimens examined: known only from the type locality, Little Sycamore Can- yon, Moran 1890 (CU), 2072 (UC), 3078 (type: UC; isotypes DS, POM). Illustration: Des. Pl. Life 8: 70. 1936. The plant shown in this photograph is very lax, apparently as a result of cultivation. Uhl and Moran (1953, p. 495, under Dudleya sp. affin. D. ovatifolia) reported a gametic number of 17 chromosomes in each of two collections of D. cymosa marcescens. Thus, like the other subspecies of D. cymosa, this plant is diploid with relation to the basic number for the genus. The subsp. marcescens appears to be the most distinctive of the sub- species recognized here for D. cymosa. It is quite different from the subsp. cymosa, of central California, but in some respects these two are con- nected by the subsp. ovatifolia, which also occurs locally in the Sierra 108 MADRONO [Vol. 14 Santa Monica. The subsp. marcescens appears to be quite distinct from the subsp. ovatifolza, differing in its more slender caudex, in its narrower rosette leaves, and in the withering of its rosette leaves in summer. The only other member of the subgenus Dudleya known to be com- pletely leafless in summer is D. parva Rose & Davidson. That also is a small diploid plant very local in Ventura County: it occurs about 8 miles north of Little Sycamore Canyon. Dudleya parva is quicker to lose its leaves in summer and slower to produce new ones after the first rains. It differs from D. cymosa marcescens further in its narrower rosette leaves, its much shorter pedicels, and its less sharply acute petals, which are pale yellow rather than bright yellow. For description and photographs of D. parva, see Moran, 1948. DUDLEYA cyMosaA (Lemaire) Britton & Rose subsp. minor (Rose) Moran, comb. nov. Dudleya minor Rose in Britton & Rose, Bull. N. Y. Bot. Gard. 3: 19. 1903. DUDLEYA CyMOSA (Lemaire) Britton & Rose subsp. ovatifolia (Brit- ton) Moran, comb. nov. Dudlevya ovatifolia Britton in Britton & Rose, Bull. N. Y. Bot. Gard. 3: 20. 1903. DUDLEYA CYMOSA (Lemaire) Britton & Rose subsp. setchellii (Jepson) Moran, comb. nov. Cotyledon laxa (Lindley) Brewer & Watson var. setchellu Jepson, Fl. West. Mid. Calif. 267. 1901. Dudleya hassei (Rose) Moran, comb. nov. Stvlophyllum Hassei Rose in Britton & Rose, Bull. N. Y. Bot. Gard. 3: 35. 1903. DUDLEYA SAXoSA (M. E. Jones) Britton & Rose subsp. aloides (Rose) Moran, comb. nov. Dudleya aloides Rose in Britton & Rose, Bull. N. Y. Bot. Gard. 3: 15. 1903. DUDLEYA SAxosA (M. E. Jones) Britton & Rose subsp. cole (Rose) Moran, comb. nov. Dudleya Collomae Rose in Morton, Des. PI. Life 6: 68. 1934. San Diego Natural History Museum, San Diego, California. LITERATURE CITED Lanjouw, J., and F. A. Starteu. 1956. Index herbariorum. Part I, the herbaria of the world. Ed. 3. Utrecht. Moran, Rew. 1948. Dudleya parva, Rose & Davidson. Des. Pl. Life 20: 137-140. Unt, Cuarves, and Rem Moran. 1953. The cytotaxonomy of Dudleya and Has- seanthus. Am. Jour. Bot. 40: 492-502. NOTES AND NEWS CLEISTOGAMY IN Mimutus Dovuctasi Gray. In 1938, J. T. Howell described a small cleistogamous-flowered Mimulus as M. cleistogamus (Leafl. West. Bot. 2: 79), but he later (op. cit. 3: 127-128. 1942) recognized it as merely a “growth phase” of M. Douglasii. My own observations indicate the frequent presence of cleistogamous flowers in M. Douglasii. Their presence seems to be related to absence of sufficient water in the soil. Normally these plants grow in thin soil over sandstone, and, in years when late winter and early spring rains come often enough, they produce large open flowers. In some situations, for example, where the soil is extremely thin on 1957 | NOTES AND NEWS 109 southern exposures, this seldom happens, and year after year, only cleistogamous flowers are produced. A colony was once seen where apparently the water supply was cut down at a critical moment, following which some of the plants produced small open flowers some of which were abnormal for M. Douglasi (V. F. Hesse 918, Jepson Herbarium, University of California, Berkeley). One of these small flowers was observed to re- semble in shape the flowers of M. Congdonzi Robinson. Although M. Congdonii grows in the same area, it apparently requires a somewhat deeper soil, and has not been observed to produce cleistogamic flowers. These observations were made in the Boulder Creek area of Santa Cruz County. —V. F. Hesse, Boulder Creek, California. NoTEs ON CALIFORNIA GRASSES. Three summer weedy annual grasses are extending their range northward in California. They occur on wet soils, periodicaily flooded by irrigation during the summer months. Specimens cited are at the Agronomy Her- barium, University of California, Davis. 1. ERIOCHLOA CONTRACTA Hitchc. (prairie cupgrass). Generally of sparse occur- rence in the state. Introduced from the Great Plains originally into southern Cali- fornia. William H. Allison collected the grass in Merced County in 1939 and the author discovered it in northern Solano and southern Yolo counties (Crampton 3147, 3148) in the summer of 1955. Apparently well-established here and competing suc- cessfully with Echinochloa crus-galli and Echinochloa colonum. 2. CHLORIS VIRGATA Swartz (feather fingergrass). Occurs mostly in southern Cali- fornia and the San Joaquin Valley. Collection of the grass from near Davis, Yolo County (Crampton 3140), and Pentz, Butte County (Morse, Farm Advisor, Butte County), confirms its northward extension. The plant has been reported seen near Auburn, Placer County, but no voucher is available. 3. LEPTOCHLOA FASCICULARIS (Lam.) Gray (sprangletop). This is now a very common weedy grass preferring wet habitats ranging from loose, sandy soils of stream and river shores to heavy adobe or alkaline soils of valley plains and river bottom- lands. In its early development, the grass probably behaves as an aquatic, since ma- turing plants are often partially immersed in water, particularly in and around rice fields. Distribution of the grass in California may be designated as follows: Infrequent: in the Coast Ranges from San Francisco Bay south to Lower Cali- fornia; alkaline soils east of the Sierran Crest from Lassen County south to Inyo County. Abundant: Great Valley, Butte County south to Kern County. AGROSTIS TANDILENSIS (O. Kuntze) Parodi (A. kennedyana Beetle). This rare annual grass, previously known in California only from San Diego County, was dis- covered in Solano County (Crampton 3275, 3289, 3296, 3300) during April, 1956. The general area of occurrence of this species begins about 7.5 to 8 miles south of Dixon, on the road to Rio Vista, centers around Dozier Station, and extends south- ward for several miles. The area habitat is one of a valley plain with low and small to large hummocks interspersed with hog wallows or vernal pools that are largely alkaline and support tufts of the rhizomatous Distichls along with Eryngium, Des- champsia danthonioides, Baeria, and Pogogyne. The hummocks support largely the Mediterranean annual grasses Bromus mollis, Bromus rigidus, Festuca bromoides and close allies, Avena barbata, Hordeum hystrix, and Lolium multiflorum. In some local- ized areas, the vestiges of the old Pacific Bunchgrass region is seen in Stipa pulchra and the less common Melica californica. Some native Trifoliums are in abundance, particularly the striking Trifolium barbigerum var. lilacinum (Greene) Jepson. Agrostis tandilensis is relatively inconspicuous in the beds and along the margins of these vernal pools, and often is masked by Eryngium and Deschampsia dantho- nioides. Sometimes, though, this grass is noticeable on the somewhat barren portions of the pools. The species generally is not abundant, and is certainly not a conspicuous element of its habitat. One or two pools were found supporting many plants of the 110 MADRONO [Vol. 14 species, while most pools had none, or, if a few, the plants went unnoticed among other vegetation. In one pool the grass was associated with a related annual, Agrostis microphylla var. intermedia Beetle, remarkably distinctive from the pale green and shining panicles of A. tandilensis by its reddish panicles, but also easily overlooked among the ubiquitous Deschampsia danthonioides —BEECHER CRAMPTON, Agronomy Herbarium, University of California, Davis. CALIFORNIA BOTANICAL SOCIETY PUBLISHERS OF MADRONO REPORT OF THE TREASURER FOR 1956 RECEIPTS: Balance on hand in commercial account, January 15, 1956.....$ 464.15 From memberships and subscriptions...................000.....22.0---- From sales of back numbers of Madrono............................. Receipts trom: annual dinners 2.6 ee Received as authors’ share of publication costs..................... Contributions toxendowment funds 2n2.2 3.2 Contributions to °memornal inde see i eo SPOLAILTECEIPES Hate ee ee DISBURSEMENTS: Credited to endowment fund from sales of back numbers Ol MGA OO: ct St Se ek eee one Credited to endowment fund from contributions................. Creditedsto memorial fund... 28s st ee ee ee Corresponding Secretary’s €xpenses —..:5.22.2..2..elnc ee Cost. of-annual: dinner. = s... sete te Se ee Cancellation: vetunds: 22.25) te eee ee ee Cost of printing, binding, and mailing Madrono, Volume 13, Numbers 5346, 4 angus... ee Total disbursements: 2-2 eee eae 2,369.35 he $3,549.10 ee 2.50 1 ie $2,708.80 BALANCE ON HAND IN COMMERCIAL ACCOUNT, American Trust Co., PalovAlto; january 15.6195 (ae. eeesreaee ere ee ENDOWMENT AND MEMORIAL FUND: Palo Alto Mutual Savings and Loan Association, balance’ on hand! January 15, 195/. 2 ACCrued: INGE eSt 5 ae ce eee cous ce ee eee nee From sales of back numbers of Madrono........................... Contributions to: memorial fund... 23 Contributions to endowment fund............0.00.....00..0000-0ee American Trust Company, savings account, balance: Janwary 15, 1957/-1..3 se ee ee Aconuled! Interest)... .fns 55551 ec auest Total endowment: .....5 eee eee Accounts audited and found correct: RicHarp W. Horm, Auditor June 6, 1957 Gets $ 840.30 $4,586.47 ee 7.48 380.06 ee $4,966.53 Matcorim A. Noss, Treasurer for 1956 1957] DOCUMENTED CHROMOSOME NUMBERS DOCUMENTED CHROMOSOME NUMBERS OF PLANTS SPECIES (See MaproNo 9: 257-258. 1948.) COMPOSITAE A phanostephus arizonicus A. Gray *Engelmannia pinnatifida Nutt. Gaillardia pinnatifida Torr. *Haplo pap pus spinulosus ssp. typicus H. M. Hall * Helianthus formosus E. E. Watson * HH ymenoxys argentea (Gray) K. F. Parker * Melam podium leucanthum Torr. & Gray *Psilostro phe tagetina (Nutt.) Greene CRUCIFERAE Streptanthus am plexicaulis (Wats.) Jeps. *Streptanthus Howelli Wats. NUMBER hes 4 n= 9 n= 17 n=4 n= 51 n= 115 n=10 i= s16 1) eee 44 n= 14 * Prepared slide available. 1 Symbols for institutions are those listed by bariorum, Part I. Second edition, 1954, Utrecht. COUNTED BY COLLECTION LOCALITY R.C. Jackson, |Jackson 2059, | Bernalillo UNM1 UNM County, New Mexico R.C. Jackson, |Jackson 2042, |Torrance UNM UNM County, New Mexico R.C. Jackson, |Jackson 2033, | Bernalillo UNM UNM County, New Mexico R.C. Jackson, |Jackson 2032, |Torrance UNM UNM County, New Mexico R.C. Jackson, |Jackson 742, Marion UNM IND County, Missouri R.C. Jackson, |Jackson 2038, | Bernalillo UNM UNM County, New Mexico R.C. Jackson, |Jackson 2082, | Bernalillo UNM UNM County, New Mexico R.C. Jackson, |Jackson 2049, |Torrance UNM UNM County, New Mexico A.R.Krucke- |Kruckeberg 1553,|San Gabriel berg, WTU WTU Mountains, Los Angeles County, California A.R.Krucke- | Kruckeberg 1881,| Siskiyou berg, WTU WTU Mountains, Josephine County, Oregon (continued on p. 112) Lanjouw and Stafleu, Index Her- 112 MADRONO [ Vol. 14 SPECIES NUMBER | COUNTED BY COLLECTION LOCALITY LILIACEAE *Fyitillaria camtschatcensis n= 12.|R.‘Ornduit.& Kruckeberg & Stillaguamish (L.) Ker-Gawl. A.R. Krucke- Ornduff 4013, | River, Snoho- berg, WTU WTU mish County, PORTULACACEAE Washington + Lewisia Tweedyi n= 46 |A.R.Krucke-_ | Kruckeberg 3320,| Wenatchee (Gray) Robins. berg, WTU WTU Mountains, Chelan County, RANUNCULACEAE Washington *Delphinium bicolor n=8 |R. Ornduff, UC | Hitchcock & About 20 miles Nutt. forma Muhlick east of Prineville Helleri (Rydb.) 20781, WTU on road to John Ewan Day, Crook *Nuttallianum CONN eee oe Pritz. ex n=8 |R. Ornduff, UC | Hitchcock & About 2 miles Walpers Muhlick east of Plains, 20862, TU |Sanders County, _ Montana *cyanoreios Piper n= 8 R. Ornduff, UC | Hitchcock & About 18 miles var. cyanoreios Muhlick south of Idaho 20820, WTU_ | City, Boise County, Idaho 20835, WTU |About 21 miles west of Lowman, Boise County, Idaho 20837, WTU_ |Two miles north of Cascade, Valley County, é ge Idaho *“cyanoreios Piper forma multiplex Nene! R. Ornduff, UC | Hitchcock & About 10 miles Ewan Muhlick northwest of 20880, WTU_ | Ellensburg, Kit- titas County, Washington INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication should not exceed an estimated 20 pages when printed unless the author agree to bear the cost of the ad- ditional pages at the rate of $15 per page. Illustrative materials (includ- ing “typographically difficult” matter) in excess of 30 per cent for papers up to 10 pages and 20 per cent for longer papers are chargeable to the author. Subject to the approval of the Editorial Board, manuscripts may be published ahead of schedule, as additional pages to an issue, provided the author assume the complete cost of publication. Shorter items, such as range extensions and other biological notes, will be published in condensed form with a suitable title under the genera! heading, ‘‘Notes and News.” Institutional abbreviations in specimen citations should follow Lanjouw and Stafleu’s list (Index Herbariorum. Part 1. The Herbaria of the World. Utrecht. Second Edition, 1954). Articles may be submitted to any member of the Editorial Board. MADRONO A WEST AMERICAN JOURNAL OF BOTANY A quarterly journal devoted to the publication of botanical re- search, observation, and history. Back volumes may be obtained from the Secretary at the following rates: Volume I, 1916-1929 .. 9... gs 85200 Volume II, 1930-1934. . . . . 5.00 Volume ITI, 1935-1936 . . . . 5.00 Volume IV, 1937-1938 . . . . 5.00 Volume V, 1939-1940. . . .. . 5.00 Volume VI, 1941-1942 . . . . 5.00 Volume VII, 1943-1944 . . .. . 5.00 Volume VIII, 1945-1946 . . .. 5.00 Volume IX, 1947-1948 . . . . 6.00 Volume X, 1949-1950. . . .. . 7.00 Volume XI, 1951-1952. . . Ss. 8.00 Volume XIT, 1953-1954 . . .. 8.00 Volume XIII, 1955-1956 . . . . 8.00 Single numbers, Volumes’ leandy2> -. A jar ee .50 Succeeding numbers, (except Volume XI, No.2). . . 1.00 Volumezxt. Nov27 2s) We) tae 1.50 The subscription price of MapDRONo is $4.00 per year. If your institution does not now subscribe to MApRONO, we would be grateful if you would make the necessary request. Since there is no extra charge for institutional subscriptions, an individual may hold membership in the California Botanical Society on the basis of his institution’s subscription. Address all orders to: G. THomas Rossins, Corresponding Secretary Department of Botany University of California, Berkeley 4, California MADRONO VOLUME 14, NUMBER 4 OCTOBER, 1957 Contents PAGE SoME AMERICAN SPECIES OF MARSILEA WITH SPECIAL REFERENCE TO THEIR EPIDERMAL AND SORAL CHAR- ACTERS, K. M. Gupta 113 CHROMOSOME COUNTS IN THE SECTION SIMIOLUS OF THE GENUS MIMULUS (SCROPHULARIACEAE). IT. Barid B. Mukherjee, Delbert Wiens, and Robert K. Vickery, Jr. 128 ON THE SPECIFIC DISTINCTNESS OF RUDBECKIA LACI- NIATA AND R. AMPLA, George Neville Jones 131 PINES FROM Nuevo LEON, Mexico, Bruce Zobel and Franklin Cech 133 A WEST AMERICAN JOURNAL OF BOTANY PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY MADRONO A WEST AMERICAN JOURNAL OF BOTANY Entered as second-class matter at the post office at Berkeley, California, January 29, 1954, under the Act of Congress of March 3, 1879. Established 1916. Subscription price $4.00 per year. Published quarterly and issued from the office of Madrofio, Herbarium, Life Sciences Building, University of California, Berkeley 4, California. BOARD OF EDITORS HERBERT L. MAsoNn, University of California, Berkeley, Chairman EpGAR ANDERSON, Missouri Botanical Garden, St. Louis. LymAN BENSON, Pomona College, Claremont, California. HERBERT F, CoPELAND, Sacramento College, Sacramento, California. Joun F. Davipson, University of Nebraska, Lincoln. IvaAN M. JounsTon, Arnold Arboretum, Jamaica Plain, Massachusetts. Mitprep E. Matuias, University of California, Los Angeles 24. Marion OwnBEY, State College of Washington, Pullman. Ira L. Wiccrns, Stanford University, Stanford, California. Secretary, Editorial Board — ANNETTA CARTER Department of Botany, University of California, Berkeley. Business Manager and Treasurer—Matcoitm A. Noss Carnegie Institution of Washington, Stanford, California CALIFORNIA BOTANICAL SOCIETY, INC. President: Rimo Bacigalupi, Jepson Herbarium, Department of Botany, Univer- sity of California, Berkeley, California. First Vice-president: Richard W. Holm, Natural History Museum, Stanford University, Stanford, California. Second Vice- president: Lyman Benson, Department of Botany, Pomona College, Claremont, Cali- fornia. Recording Secretary: Mary L. Bowerman, Department of Botany, University of California, Berkeley, California. Corresponding Secretary: G. Thomas Robbins, Jepson Herbarium, Department of Botany, University of California, Berkeley, Cali- fornia. Treasurer: Malcolm A. Nobs, Carnegie Institution of Washington, Stanford, California. 1957] GUPTA: MARSILEA 113 SOME AMERICAN SPECIES OF MARSILEA WITH SPECIAL REFERENCE TO THEIR EPIDERMAL AND SORAL CHARACTERS K. M. Gupta, D.Sc. In connection with the monographic study of the genus Marsilea m India that has been undertaken in my laboratory at Jaswant College, Jodhpur, a loan of about one hundred and fifty sheets of Marsilea com- prising eleven American species was received from the University of Cali- fornia Herbarium, Berkeley, for comparison with the Indian species. Of the species received, only two, M. minuta and M. quadrifolia, are repre- sented in the Indian flora. Study of this material led to the paper here presented. Marsilea is a genus of world-wide distribution. Just as M. minuta is of very wide occurrence in India from the Panjab in the north to Travan- core in the south, W/. vestita among the American species has a wide range of distribution throughout the western and Pacific United States and, like the Indian species, is very variable. It has been observed that M. minuta possesses varieties which are anatomically distinct, particularly in the internal organization of the sporocarps, but since this subject is now being studied by my research students from both the morphological and cyto- logical point of view, I do not wish to anticipate their results herein. Miss Margaret Stason (1926), too, had hinted toward such a possibility while discussing the possible connection between MV. vestita and M. oligospora. I should like to state that both these latter species seem distinct from one another. Because the American material was limited in quantity, my study of these species is not as extensive as I would have liked. In some cases it was possible to remove only one or two sporocarps from the herbarium sheets for the study of their internal structures. However, the external characters were carefully studied and in spite of the fact that the genus is notorious for its morphological plasticity, one cannot always ignore certain differences like the shape of the leaflets (figs. 1-9) that might clearly be discernible among the allied species. I have, therefore, sum- marized these observations on the vegetative characters in Table 1. On account of the limitations imposed for want of preserved material, I chose to remove only a few leaflets and some sporocarps from each of the collections in order to examine their epidermal and soral characters. It is a well known fact that with the publication of many important works on the epidermal and cuticular studies of vascular plants (Linsbauer, 1898; Porsch, 1905; Thomas and Bancroft, 1913; Rehfous, 1917; Ban- dulska, 1923, 1926; Prat, 1932; Florin, 1931 and afterwards; Allsopp, 1952, 1953a, 1953b, 1953c, 1954, 1955), the taxonomic importance of such studies has gained world-wide significance. Linsbauer was probably Maprono, Vol. 14, No. 4, pp. 113-144. October 30, 1957. 114 MADRONO (Vol. 14 the first to study the epidermal characters in Lycopodium. Porsch had suggested the phylogenetic importance of stomatal characters, Rehfous thought that the stomatal apparatus exhibits a constant feature within a group, and according to Miller (1938, p. 322) “the structure of the stomatal apparatus is markedly different in different groups of plants.” Prat, after his comprehensive work on the epidermal structure of the Gramineae, stated that epidermal structures and their distribution of parts are specific characters. He says (1932, p. 185), ““Correctement in- terpretés, les caractéres épidermiques peuvent étre de méme valeur que les caractéres qui ont servi de base a la définition des groupes systéma- tiques.”’ Satake (1934) has shown the systematic importance of the epi- dermal elements in the leaves of Japanese selaginellas and has founded a new classification on the basis of the epidermal structure of the leaves. Chowdhury (1937) studied the epidermis of eleven Indian species of Lycopodium and found that characters of the epidermis in most of the species were of diagnostic value. Some study has been made in my labora- tory on the structure of the epidermis of some Indian species of Selagi- nella. As far as I know no detailed investigation on the epidermis of Mar- silea, a plant with pronounced amphibious habit, has been undertaken in the past. The present examination, though brief, is clearly indicative of its usefulness. The soral characters in the sporocarps of the genus have been studied in the past, but their systematic value had not yet been fully recognized. SPECIMENS EXAMINED. The University of California material studied was collected by various workers over a period extending from 1849 to 1954. Except for M. vestita, which has a very wide distribution in the United States, all of the specimens exam- ined are cited below: MarsILEA VESTITA Hook. & Grev. UNITED STATES. Western Texas to El Paso, New Mexico, Wright 811, May-October 1849. Utan. Rich County: Bear Lake near Laketown, Porter 6490, 30 June 1954. CALIFoRNIA. San Diego County: Purpus, May- October 1898. MarsILEA MOLLIS Rob. & Fern. MEXICO. Curnuanua. St. Diego, Hartman 604, 20 April 1891. Duranco. City of Durango and vicinity, Palmer, April-November 1896. MarsILEA MEXICANA A. Br. MEXICO. CurtHuanua. Near Cusihuiriachic, Pringle 2007, 23 September 1888. Sonora. San Pedro, Hartman 893, 1894. MarsILEA MACROPODA Engelm. MEXICO. Tamauttipas. Near Matamoras, Pringle 1975, 8 August 1888. UNITED STATES. Texas. (New Braunfels ?), Lindheimer 573, 1846. Ponds on the Seco, Reverchon 1630, June, no year. Dimmit County: Turkey Creek south of Crystal City, Muenscher & Winne 16506, 29 June 1945. Jackson County: Lavaca River, Tharp, 29 August 1941. MArSILEA FOURNIERI C. Chr. MEXICO. Nuevo Leon. Rio de San Juan outside of China, Barkley 14344, 27 February 1944. Sonora. Rancho San Carlos, 40 miles west of Hermosillo on road to Kino Bay, Wiggins & Rollins 181, 30 August 1941. Baya CALIFORNIA. Seventeen miles south of Pozo Aleman, Wiggins 7848, 3 March 1935; 20 miles south of Calmalli, Wiggins 5422, 18 April 1931. MarsILEA UNCINATA A. Br. UNITED STATES. Haven’s Ranch, Lemmon, July 1882. Texas. Grown at Berlin Botanic Garden from fruit collected in Texas in 1872 by E. Hall. Comanche Spring, New Braunfels, Lindheimer 1283, June 1851. Caldwell County: dry sink, prairie, Barkley 13130, 7 July 1943. 1O5%/1 GUPTA: MARSILEA 115 o gy Fics. 1-9. Photographs of the quadrifoliate leaflets of Marsilea species, rendered transparent so as to show exact outlines and venation. 1, M. cf. macropoda, X 0.8; 2, M. minuta, <* 1.3; 3, M. mucronata, X 1.3; 4, M. tenuifolia, * 1.3; 5, M. mollis, x 2; 6, M. vestita, K 1.3; 7, M. uncinata (lacking two leaflets), « 2; 8, M. oligo- spora, X 1.3; 9, M. fournieri, K 1.8. MarSILEA TENUIFOLIA Engelm. UNITED STATES. Texas. Inks Lake, Tharp, 11 August 1941. (New Braunfels?), Lindheimer 745 in 1847. Llano County: creek near Kingsland, Whitehouse 18480, 4 May 1947; creek bed, Tharp, 15 August 1940. MarsILEA OLIGOSPORA Goodd. UNITED STATES. Wyoming. Uinta County: Jack- son Lake, Nelson 6560, 12 August 1899 (isotype). Elmore County: King Hill, Nelson & Macbride 1158, 17 July 1911. Sublette County: near New Fork Lake, Payson & Payson 4437, 24 July 1925; Kendall, Payson & Payson 2920, 5 August 1922. MarsILEA MucRONATA A. Br. UNITED STATES. NortH Dakora. Valley City, 116 MADRONO [Vol. 14 Stevens 1223, 1 August 1950. Benson County: Butte, Lunell, 5 September 1905. NEBRASKA. Exeter, Wzbbe, September 1888. Wyominc. String Lake, Prettyman, 20 July 1953. MarSILEA QUADRIFOLIA L. UNITED STATES. Connecticut. New Haven, Setchell, 3 October 1883. Cromwell, Hubbard 6817, no date. Litchfield County: Bantam Lake, Bridgman, 1888, Thompson, August 1891. MaAssacuusetts. Falmouth, Brooks, 30 August 1910. Salem, Harper & Harper, 19 July 1895. Glacialis Pond, Cambridge, Pease 5672, 9 October 1904. Norfolk County: Wellesley, Brown, 14 August 1940. Kentucky. Fayette County: two miles east of Lexington, McFarland 46, 25 Sep- tember 1940. MarsILEA MINUTA L. MEXICO. Coanutra. Cerro de Cypriano, Purpus 4525, July 1910. StnaLoa. Culiacan, Brandegee, 11 November 1904. BAJA CaALIFoRNIA. San Jose del Cabo, Brandegee, September 1893, 16 October 1899, 25 October 1902. Jarisco. Near Guadalajara, Pringle 2434, 6 December 1889. MarsILEA cf. MACROPODA Engelm. MEXICO. CoAnuirLa. Rancho Agua Bueno, 43 miles north of Monclova, Gould 6405, 20 June 1952. In studying the above-cited specimens, the vegetative and reproductive structures were prepared in the following manner. Hatrs. The hairs from different regions (leaves, nodes, and the sporo- carps) were removed dry and mounted directly in liquid paraffin on a clean slide in the manner usually employed for the examination of fibers of wool or cotton, so as to bring out their medullated or non-medullated nature quite distinctly. EpmpermMIs. For studying the epidermal structures, the leaves were treated with dilute KOH solution, washed in water, carefully teased, and stained with a single stain, safranin. After usual dehydration they were mounted in Canada balsam for examination. SPOROCARPS. The sporocarps were examined dry for their external feat- ures, but in order to ascertain their soral number, they were soaked in warm water after slightly scratching their resistant walls with a knife. This facilitated the so-called germination. The coming out of the muci- laginous mass is a mechanical process and cannot be strictly described as germination, as people often believe, because the real germination would mean the future activity of the mega- and microspores leading to the pro- duction of the gametophytes. The sporocarps were dissected further to study the nature of the sorus, the structure of the mega- and microspo- rangia, and also to find out whether the sporocarps were normal or pos- sessed abnormalities which are often of a fluctuating type. RANGE OF VARIATION The habitats of Marsilea vary from aquatic to subterrestrial or terres- trial; almost all species start as aquatic plants during which time their vegetative growth is pronounced. Some, like M. minuta, also produce sporocarps in the aquatic habitat, while others, like 7. aegyptiaca, never do so. The majority of the species possess the tendency to fruit only under a subterrestrial habitat, and in fact most of the specimens from the Uni- versity of California Herbarium seem to have been collected under sub- terresirial conditions. They grow in plains (100 feet above sea level) as 117 MARSILEA GUE 1957] wpodoyoeuw Kaen TO°O X z20°O 8 9 '9°T 9 ‘o°2 | 8 te°T | Il°T: ¥O°T pedeys-ue4 9°ZT | O°FT| 9°OT| Tel4zsessezqns "390 "Ww Kaen I0°O X z0°0 LI 8T°0 : 62°0 a}eaung ve 0°9 Geir Telupsessezqns eynuiw "Ww Telsapsausey pue Tel4u}sausay #4618445 T0°O X 20°0 @}eA0GQ 9°2 -qns ‘ol1zenby | eiTostdpenb ° + W Kaen a}zeaung 16°z 2°T Telupysesuay epeuoyonu “pW Tel4psessey Kaen @}eA0gG e°t z*2 | v0 pue o1}enby esodsoBito *w +— yYHiesys TO°O X €0°0 9 8yeTOedUeTGO 6L°E €°9 S*2 o1zenby BiTosinuezy “Ww Kaen TO°O X 20°O €T a}eaung Gv'd O°e I'l Telspsausey ePeUIOUN “W Aaem’ mOsseU Tel4zpsesssa} pue ‘6u04 z0°O X 20°0 LT 8} eA0gGO 69°? vie O°T Teldszysassazyqns f4alusno}s °W yzoous Tel4pseusazyqns pue §uo7q TO°O X €0°O 6 8} eA0gO S‘*v €°s v'°T pue o1zenby epodosoew “Ww Aaem ssayq T0°0 X 20°0 8 ce °O =: €€°O ezpeauny 9°sS €°9 0°S Telszpsesuazqns eueoixaw “WwW Kaem ATubtyH c0°O X €0°0 OT 80°0 : 8T°0 8}yeaun) ¥8°0 TT s°0 peuol}yusw ON SITTOW "W —- 7461 esys Tel4zpsasue} }eyMawos pue [Tel4}sausay ‘6u0q T0°O X z0°O zz L tv't yty’t|] L tL*t | 62°0 :€€°0 @}eA0KO ve s*s | ¢°T -qns ‘o1}zenby BLIZSEA “W S[Te0 sdueoosods SOACIT sapou Tewsapida e}eWwoYS Aouanbauy STI99 $0 saqunu pue o1}zed S}ATLeIST “aAy *xew [cul $O TIM $0 AZIS Tezpewoys ‘ww ul Yyzyb6uaT abesaay }eT}eAq P/V $0 adeys @aTo1z,ed yo yzybue7q }Iqey sajoeds $o oweN Siwuapidy S4ley | [esuauag VITISYYW 40 S31903dS NVOIYSWY HLYON JHL JO SYSLOVYVHO BAILVL39SA “°T 378V1L 118 MADRONO [Vol. 14 18 19 20 | Fics. 10-21. Photomicrographs of the epidermis of American Marsileas, showing nature of the epidermal cells and distribution of stomata. 10, M. vestita; 11, M. oligospora; 12, M. macropoda; 13, M. cf. macropoda; 14, M. mollis; 15, M. mexi- cana; 16, M. fournieri; 17, M. uncinata; 18, M. tenuifolia; 19, M. minuta; 20, M. mucronata; 21, M. quadrifolia. All about xX 85. well as on mountains, 6000 feet or above. Not only this, but some species like M. vestita are found in the plains (160 feet above sea level) as well as in the higher altitudes (5800 feet), whereas a species like M/. mollis is confined to higher levels (6000 feet) only. The rhizome is creeping and branches dichotomously. The internodes are short or long, sometimes becoming very much reduced. The adventi- 1957] GUPTA: MARSILEA 119 tious roots, few or more in number in different species (one to eight), usually arise ventrally from the nodes of the rhizome. The size of the petiole varies considerably (from one to nine inches in length in different species), and so also the shape and size of the leaflets. The oblanceolate leaflets of M. tenuifolia (fig. 4) are quite peculiar to the species (Table 1). The detailed examination of the structure of the hairs revealed the dif- ferences that exist among them in the different organs in the same species or on the same organs of the different species. The main point of interest is centered around the shape and size of either the basal or the terminal cells of the hairs of these organs (Table 1). The study of epidermal structures has revealed that the species seem to differ from one another at least in two important respects, namely na- ture of the epidermal cells and the frequency of the stomatal distribution. The walls of the epidermal cells are either smooth, slightly wavy, or highly so, and the frequency of the stomata per unit area ranges from five to twenty (Table 1). The sporocarps are stalked. Their attachment to the petiole is either basal or adnate, and the number of the sporocarps in a group may be one or more, connate or free (figs. 27-38). The shape differs from species to species, for instance it is subspherical in M. minuta (fig. 37), oval in M. mollis (fig. 28) and clearly bean-shaped in M. vestita (fig. 27). All the American species possess two horns, these varying in their promi- nence and bluntness. Similarly there is a difference in their soral number, a fact that seems to be of some systematic importance in the genus Marsilea. DISCUSSION OF THE SPECIES MARSILEA VESTITA and M. oLicospora. Marsilea vestita is very well represented in the collection by 84 sheets which date from 1894 to 1954. While the general characters are given in tables 1 and 2, I should like to mention a few interesting observations on this species. The anatomy of the epidermis shows that the walls of the epidermal cells are long and more or less straight (fig. 10). The stomata are uniformly distributed throughout, their orientation being in the direction of the length of the epidermal cells. The frequency of the stomatal distribution is almost the highest among the species here examined, namely twenty-two per unit area. There is a single stalked sporocarp in M. vestita which is attached at the base of the petiole. It is bean-shaped and possesses both the horns, the upper pointed and the lower one blunt (fig. 27). An abnormal speci- men with a sporocarp adnate to the petiole was found on sheet UC 205105 (fig. 24). Most of the specimens possessed normal sporocarps with nor- mal type of both mega- and microspores (figs. 25a, 25b), but a critical search of the specimens revealed the presence of abnormal sporocarps (fig. 26a). The latter had no megaspores but only microspores. Sometimes these microspores were also of two types (fig. 26b). As pointed out in the 120 MADRONO [Vol. 14 Fics. 22-26. Fic. 22, M. cf. macropoda, habit showing largest leaflets and petioles, ca. X 0.25. Fic. 23, M. macropoda, same for comparison, ca. X 0.25. Fic. 24, M. vestita, abnormal specimen found on UC 205105, sporocarp adnate to petiole, x 0.25. Fic. 25, M. vestita, megaspores (25b) and microspores (25a) from a normal sporo- carp, X 25. Fic. 26, M. vestita, microspores from two different sporocarps (26a, abnormal; 26b, normal), x 25. Generally in abnormal sporocarps megaspores are absent. introduction, such a thing was expected in a species with a wide distribu- tion; however, the significance of these and such other abnormal sporo- carps as have been discovered in more than one Indian species is not dis- cussed here. It is important to point out that Miss Stason’s observation (1926, p. 478) that MV. oligospora may be an “ecotype” of M. vestita may not be quite correct, for I find that the epidermal as well as soral characters are distinct in the two species (figs. 10, 11, 27, 34) insofar as they are repre- 1957] GUPTA: MARSILEA 121 sented in the collection studied. For these reasons I prefer to regard M. oligospora as a distinct species rather than just an ecotype of M. vestita. MarSILEA MACROPODA and M. cf. MAcRopopDA. Five sheets in the series studied represent M. macropoda (fig. 23) and one specimen is question- ably referred to that taxon as M. cf. macropoda (fig. 22). At the outset I should like to mention that these two look quite distinct from one an- other. The latter possesses not only larger leaves and sporocarps than M. macropoda, but probably larger than any species (figs. 1-9) repre- sented in the series here treated. The epidermal cells in WM. macropoda possess long and smooth walls (fig. 12), whereas in M. cf. macropoda the epidermal cells possess wavy cell walls (fig. 13) with stomatal frequency, however, being nearly the same in the two cases, namely nine and eight, respectively. The sporocarps are stalked and tufted at the base of the petiole in MV. macropoda (fig. 30). They are oval in shape and the biggest in size as compared with other species. On the other hand a group of only two or three sporocarps which are slightly adnate to the petiole is found in the specimens of M. cf. macropoda (fig. 38). The sporocarps in the two cases are, however, densely covered with hairs and possess both the horns, the lower one being more prominent. One of the petioles in M. macropoda bears five leaflets instead of the usual four. This is not a very unusual feature in Marsileas, having been described for M. quadrifolia in 1948 and M. aegyptiaca in 1956. MarsSILEA MOLLIS. There is only one herbarium sheet of this species, containing plants collected at two different times (1891 and 1896). The habit of this species is quite characteristic, with small hairy leaflets. There are no rhizomes or roots preserved; perhaps these are very much reduced. The epidermal cells of the leaves show wavy nature in their walls (fig. 14) and the uniformly distributed stomata show a frequency of about ten stomata per unit area. The stalked sporocarps are tufted in nature, and their exact mode of attachment is not possible to describe in the absence of the rhizomes. The hairy and ribbed sporocarps are more or less oval in shape and possess two blunt horns (fig. 28). The soral number is about 15 in the species. MARSILEA MEXICANA. The two herbarium sheets of this species (col- lected in 1888 and 1894) are badly preserved. The leaves show epidermal cells which are less wavy and possess stomatal frequency of only eight per unit area (fig. 15). The sporocarps are a bit characteristic in being somewhat narrower on the sides and swollen in the middle (fig. 29). They are ribbed and possess two blunt horns, with soral number twelve. MarSILEA FOURNIERI. The four herbarium sheets contain material from both subterrestrial and terrestrial habitats. The leaves are obovate. The epidermal cells are long and narrow with their walls wavy in nature (fig. 16). The stomatal frequency is quite high, as much as 17 per unit area. 122 MADRONO [Vol. 14 lbeb LP LP eh? |sP Ls kW be / kP be |Z Fics. 27-38. Outline diagrams showing attachment of sporocarps to the petiole and the peduncles as well as the shape of the individual sporocarps. 27, M. vestita; 28, M. mollis; 29, M. mexicana; 30, M. macropoda; 31, M. fournieri; 32, M. un- cinata; 33, M. tenuifolia; 34, M. oligospora; 35, M. mucronata; 36, M. quadrifolia; 37, M. minuta; 38, M. cf. macropoda. All about x 30. The stalked sporocarps remain tufted at the base of the petiole. They are clearly bean-shaped and possess both the horns, which are pointed. The upper horn is further characterized by being longer and curved (fig. 31). The ripe mature sporocarps are quite smooth and ribbed, otherwise the species, like M. macropoda, is profusely covered with hairs. MARSILEA UNCINATA. In the four herbarium sheets the vegetative char- acters are quite variable, as for instance the number of roots (2 to 8) on 1957] GUPTA: MARSILEA 123 the nodes and the size of the internodes. The leaves are cuneate, possess- ing epidermal cells with wavy walls (fig. 17). The stomatal frequency is thirteen per unit area. There is a single stalked sporocarp which is oval in shape and is at- tached almost at the base of the petiole. There are two horns present on the sporocarp, the upper one being pointed and curved (fig. 32). The pedicel is fully adnate to the body of the sporocarp, which is smooth and possesses 16-18 sori. MarSILEA TENUIFOLIA. The four herbarium sheets all contain material collected in an aquatic habitat. The leaves are very characteristic of the species, being oblanceolate in shape (fig. 4). The anatomy of the epider- mis shows that the epidermal cells are long and narrow. The walls of the epidermal cells are smooth (fig. 18). The distance between the two walls of the same epidermal cell is narrow in the non-stomatal region and in- creases in the stomatal region as if to house the stomata. The frequency of stomatal distribution is 6 per unit area. The single bean-shaped sporocarp is attached by a short stalk at the base of the petiole. There are two horns present; the upper one is pointed and curved downwards (fig. 33); the wall of the sporocarp is smooth and not hairy. It may be noted that there were only two sporocarps present on one specimen, while other plants were sterile. MARSILEA MUCRONATA. The four herbarium sheets were all collected from a terrestrial habitat. The leaflets are cuneate in shape. The anatomy shows epidermal cells with wavy cell walls (fig. 20). The frequency of stomatal distribution is lowest in this species, being only five per unit area. The sporocarps are usually solitary, rarely two arising from the base of the petiole. They are bean-shaped with the pedicels partially adnate to the sporocarps (fig. 35). There are two horns; both are pointed. The soral number is 16. MARSILEA MINUTA and M. QUADRIFOLIA. Both these species are found widely distributed in India and are the common ones that have been de- scribed previously in and outside India (Pande, 1923; Mahabale & Gorji, 1948; Mehra, 1938; Puri & Garg, 1953; and Williams, 1920). And asa further study of the Indian species is in hand, it is not profitable to deal with them here, particularly in view of the fact that the University of California specimens did not contain enough material of well preserved nature. The epidermal studies (figs. 19, 21) in both the species have revealed that the stomatal frequencies in these are different, being five in M. quad- rifolia and seventeen in M. minuta. Similarly the attachment of the sporo- carp is characteristic in M. quadrifolia, being adnate to the petiole and connate to the pedicel (fig. 36), while in M. minuta two to six sporocarps are slightly connate at the base of the petiole (fig. 37). The comparison with the Indian material available in my laboratory indicated minor dif- ferences, but, in view of wide distribution of these species, these minor differences are natural. 124 MADRONO [Vol. 14 CONCLUSION From the foregoing brief description of eleven American species of Marsilea, it is apparent that the genus exhibits clear variations in the veg- etative as well as in the reproductive organs, namely the sporocarps. In spite of a similar habitat, whether aquatic, subterrestrial or terrestrial, species do differ from one another, say for instance in the shape and size of their leaves (figs. 1-9) and structures of their sporocarps (figs. 27-38). The other types of variations that are often present in the same species under varied environmental conditions are distinctly different from the former differences, and are always exhibited within certain limits in that species. The latter should not be confused with the former. A thorough ecological study, therefore, becomes imperative, besides other important considerations of anatomy and cytology, not only to decide clearly the differences between two species but also to find out the range of variations within the species. Such an autecological aspect is being attempted with M. aegyptiaca here in the Botany Department at Jaswant College. One simple aspect of morphology, namely the examination of the epi- dermis of the leaflets of these eleven species of Marsilea, has demonstrated how the structure of the walls of the epidermal cells and the frequency of stomatal distribution can be of some diagnostic value. It has been possible to indicate in the present thesis, for instance, that the species pairs M. vestita—M. oligospora and M. macropoda—M. cf. macropoda are quite distinct. If only the two above characters of the epidermis are taken into consideration, the present collection can be arranged into three different categories, namely, those with wavy walls, those with less wavy walls and those with smooth walls. The structure and attachment of the sporocarps, however, provide the main clue to the correct identification of the species. The main features that are useful and essential for systematic considerations have been men- tioned in Table 2. For the present, however, the number of sporocarps in a group and the nature of attachment of these sporocarps to the petiole have been adopted as a broad basis for classification (Gupta & Bhardwa- ja, 1956). Accordingly these American species can be arranged as follows: Pedicel adnate Pedicel basal M.macropoda Sporocarps more than one, M. quadrifolia connate or free Sporocarp solitary M. oligospora M. mexicana M. vestita M. minuta M. uncinata M. mollis M. tenuifolia M. fournieri M. mucronata It must, however, be admitted that the systematics of the genus Mar- silea is a neglected subject, and it is hoped that as a result of detailed 125 MARSILEA GUPTA 1957] yueulwoud pedeys-[eag ezpeupe Jemot '2 ezeupe ATIn4 T2°0 @}eUUOD AT3U6ITS Teolseyusqns [ewsoN Tew4oN el Zz ezeupe ATIN4 IT°O eed4 aseq ey} +¥ pedeys-Teag [ewsoN TewsON 02-91 2 ezeupe AT{InN4 vI°O e}euu0y e}eupy ajeupe| peaeys-Teag TewsonN TewsoNn oT pezyulod yyoq ‘2 Atleizpsed zZ°0 eed4 eseq oy} 1y eunosqo eseq 984} Jaddn ‘}ueulwoud eyeupe| pedeys-Teag ye ezpeupe [omy [ew JON [Bw4soN v1 esow seMoT £2 ATTe1}4ed 9T°0 eed4 ATi Y46ITS A padeys-ueag TewsoN Tew4soNn eAoge se oues ‘2 eyeupe AT[N4Y ETO eed4 eseq @4} +¥ +4614 Sp4eM0} Paaungo ejeupe|] pedeys-Teag TewsoN TewsoN 8I-9T e4sow pue pajulod Alleiyseg €T°O eed4 eseq ey} +¥ esow saddn ‘2 +4614 Sp4eao} paasno AT eseq ayy -3746!1TS euo saddn edeys-ueeg e @}eUuUuUOD T ‘ i 2 O + i eseq 84} +V Tew4oN TewsoN LT pezurod yzyoq ‘2 ezeupe AT[N4 TT‘°0 Atzu6ITS eseq ay} eseq ey} yueulwoud pedeys-Teag ze ezpeUU0D ze e,Vupe TewsoN Tew4oN JoMmoT ‘Zz eyeupe ATIN4 S2°0 AT+UBITS ATVUBITS (2) e044 azpeupe Sazeuu0d Tewson Tewson 2T yuNTq ‘Zz ATTelz4ed ATTe| p4ed eseq ey} iY e}yeupe pearys-Teag TewsoNn Tew4oNn ST-¥T yuntq ‘2 ATTelz4eg TT°O uees }ON eseq 84} +¥ [ewsouqge puesqe pezuiod pedeys-ureg JO [TeWwsON Jo TewsO0N BT usoy seddn ‘2 ezeupe AT TNY 8T°0 ead4 eseq 684} +¥ seuods pure ei fuesodseboy sesods pue ey Guesodsosoin dueoouods Teoiped eTO!j}ed Jequinu susoy j$o 0} Ted!Iped ezis 0} Ted!ped 0} [Te9dlped [esos | Joqunu pue eunzen $O JPUEWYOe ZY pue edeys $O UO!}eTEY $0 UOlzeTAY Teusezuj [euse}Xxg VATISYVYN JO S3193dS NVOIYIWY 3HL NI SdYYOONOdS JHL 4O SYALOVYVHO 1V919010Hd4ON °e@ 319avl o 2 OM} Ajaues Asez1TOS Ajeued ByITOS AsezITOS AsezITos (8) peysnL pezsny poysny (2) pezsny Asez1TOS podoywoew °49 epnutw BiToOs!spenb Be} eUO WON esodso6bi To Bi TOs! nuey ezpeuyoun {Je ;usnoy epodosuoew Buso| xew ST TOW BLIPSOA “W “nw se1oeds 40 eWweN 126 MADRONO [ Vol. 14 investigation that is proceeding in the Botany Department at Jaswant College, some useful light will be thrown on the problem. Ecological, mor- phological and cytological studies of some Indian species seem to indicate a very promising field, and it may be possible in the near future to throw light on the nature of various so-called “‘ecotypes”’ possessing fertile or fluctuating type of sterile sporocarps, as a matter of fact, on the entire phenomenon of speciation in the genus Marszilea. SUMMARY Herbarium specimens representing eleven different species of Marsilea from North America have been examined, and in each case the structure of the epidermis and the sporocarps has been described. It has been shown that the species differ from each other in nature of their epidermal cells and the distribution of their stomata. So also they differ in external and internal characters of their sporocarps. The more important characters of the latter, as well as vegetative features in all these species, have been tabulated; in addition, the anatomical features of the leaves have been given in a series of photographs to indicate clearly that vegetative feat- ures, both external and internal, do contribute minor, if not major, cri- teria in identification of the various species of Marsilea. ACKNOWLEDGMENTS For the loan of the material of these American species of Marsilea, I am extremely grateful to Professor H. L. Mason, Director of the Univer- sity of California Herbarium, and especially for his great kindness in al- lowing me to remove parts of the plants from the herbarium sheets for investigation, and for permitting me to keep the entire collection at Jodhpur for more than a year. It is my pleasure to record the assistance given me by my two pupils, Shri T. N. Bhardwaja, M.Sc. and P. L. Mital, M.Sc., in preparing this paper for publication. My thanks are also due to my laboratory staff for their help in the preparation of the manuscript. Jaswant College Jodhpur, India LITERATURE CITED Aisopp, A. 1952. Experimental and analytical studies of pteridophytes. XVII. The effect of various physiologically active substances on the development of Marsilea in sterile culture. Ann. Bot. N.S. 16: 165-183. . 1953a. A comparison of the effects of 3-indolylacetic acid and 3-indolylace- tinitrile on the development of sporelings of Marsilea in aseptic culture. Jour. Exp. Bot. 5: 16-23. . 1953b. XIX. Investigations on Marsilea. 2. Induced reversion to juvenile stages. Ann. Bot. N.S. 17: 37-55. . 1953c. XXI. Investigations on Marsilea. 3. The effect of various sugars on development and morphology. Ann. Bot. N.S. 17: 449-463. . 1954. XXIV. Investigations on Marsilea. 4. Anatomical effects of changes in sugar concentration. Ann. Bot. N.S. 18: 449-461. 1957] GUPTA: MARSILEA WF . 1955. X XVII. Investigations on Marsilea. 5. Cultural conditions and mor- phogenesis, with special reference to the origin of land and water forms. Ann. Bot. N.S. 19: 247-264. Baker, J. G. 1887. Handbock of the fern-allies. London. BanDULSKA, H. 1923. A preliminary paper on the cuticular structure of certain dico- tyledonous and coniferous leaves from the Middle Eocene flora of Bournemouth. Jour. Linn. Soc. 46: 241-269. . 1926. On the cuticles of some fossil and recent Lauraceae. Jour. Linn. Soc. 47: 383-425. CuHowpuHury, N. 1937. Notes on some Indian species of Lycopodium with remarks on the distribution of the genus in India. Trans. Nat. Inst. Sci. India 1: 187-226. CHRISTENSEN, C. 1906-1934. Index Filicum. Hafniae. Frorin, R. 1931. Untersuchungen zur Stammesgeschichte der Coniferales und Cor- daitales. K. Sven. Vetensk. Akad. Handl. 10: 1-588. Gupta, K. M. and T. N. Buarpwayja. 1956. On the investigation of Indian Marsileas: their morphology and systematics. 1. Marsilea aegyptiaca Willd. with remarks on the present systematic position of Indian species. Jour. Bombay Nat. Hist. Soc. 54: 423-444. LINSBAUER, K. 1898. Beitrage zur vergleichenden Anatomie einiger tropischer Lyco- podien. Akad. Wiss. Wien, CVII. Ab. 1: 1-36. ManasaLe, T. S. and G. H. Gorji. 1948. Some observations on the sporelings and adult plants of Marsilea quadrifolia. Jour. Univ. Bombay 26: 27. Menra, P. N. 1938. Abnormal sporocarps in Marsilea minuta Linn. Proc. Ind. Acad. Sci. 8: 8-10. Miter, E. C. 1938. Plant Physiology. McGraw-Hill, New York. Mirat, P. L. 1956. On some morphological characters of American Marsileas. Un- published thesis, University of Rajputana, Jaipur. PANDE, S. S. 1923. Some observations on the biology of Marsilea. Proc. Lahore Phil. Soc. 4: 1-28. Porscy, O. 1905. Der Spaltoffnungsapparat im Lichte der Phylogenie (as quoted in Rehfous, 1917). Prat, H. 1932. L’épiderme des Graminées, Etude anatomique systématique. Ann. Sci. Nat., Bot. Série 10°, 14:117-324. Puri, V. and M. L. Garc. 1953. A contribution to the anatomy of the sporocarp of Marsilea minuta Linn. with a discussion of the nature of sporocarp in the Mar- sileaceae. Phytomorphology 3: 190-209. Reurous, L. 1917. Etude sur les stomates. Univ. d. Genéva Inst. Bot. Sér. 9¢, Fasc. Vie, these no. 605. Stason, M. 1926. The Marsileas of the western United States. Bull. Torrey Club 53: 473-478. SATAKE, Y. 1934. Systematic importance of the epidermal elements in the leaves of the Japanese Selaginellaceae. Bot. Mag. Tokyo 48: 259-278. Tomas, H. and N. Bancrort. 1913. On the cuticles of some recent and fossil Cyca- dean fronds. Trans. Linn. Soc. London. Ser. 2, Bot. 8: 155-204. WituaMs, R. G. 1920. The anatomy and morphology of Marsilea. Unpublished thesis, Cornell University, Ithaca. 128 MADRONO [Vol. 14 CHROMOSOME COUNTS IN THE SECTION SIMIOLUS OF THE GENUS MIMULUS (SCROPHULARIACEAE), II. BaRID B. MUKHERJEE, DELBERT WIENS, AND ROBERT K. VICKERY, JR. This cytological study!, which is a continuation of a previous investiga- tion of chromosome numbers in section Simiolus (Vickery, 1955), forms an integral part of a long range experimental study of the taxonomy, cytogenetics, and evolution of species in the genus Mimulus (Vickery, 1951). The counts were determined from observation of various stages of microsporogenesis. Many different techniques of fixation and staining were tried in an attempt to develop a better method than the one pre- viously used (Vickery, 1955). An effective and comparatively simple method was developed for obtain- ing chromosome counts. Buds of proper size, which varied from 1.6 to 3.1 millimeters, were killed and fixed for about two hours in a mixture of one part acetic acid to two parts distilled water. Buds that could not be studied immediately were stored in 70% ethanol. The best time for fixa- tion proved to be from 9 to 11 a. m. The anthers were dissected out of the buds and stained in strong aceto-carmine. The stain was prepared by dissolving 1 gram of carmine in 100 mls. of boiling 45% acetic acid. It was cooled and filtered before use. The anthers were placed in a drop of stain on a microscope slide and heated gently over an alcohol flame. A cover slip was added and pressed firmly with a match stick to squash the anthers. From time to time during the next half hour more stain was added, more pressure applied and more heat used. The excess stain was removed with a paper towel pad. The coverglass was sealed with a half and half mixture of beeswax and paraffin. In a few cases propionic acid was substituted, with equally satisfactory results, for acetic acid in the above schedule. The slides were examined within a day or two and the best figures found were drawn with the aid of a camera lucida (fig. 1). The method of Bhaduri and Ghosh (1954) was employed to make the slides permanent but with only moderate success. Herbarium specimens of all the cultures counted have been prepared for future reference and will be deposited in the Garrett Herbarium of the University of Utah under the culture numbers given in Table 1. The chromosome numbers were found to be n—16 for M. dentilobus Rob. & Fern. from Chihuahua, Mexico, n=15 for M. glabratus var. utahensis Penneil from southern Utah, and n—14 for the ten cultures of M. guttatus DC. from California and northern Utah. These taxa appear to form an aneuploid sequence and a geographic series linking a group of North American taxa reported as n=14 (Campbell, 1950, and Vickery, 1 The work was supported by grants from the National Science Foundation and the University of Utah Research Fund. The authors wish to thank Drs. W. W. Newby and C. M. Woolf for their helpful criticisms of the manuscript. 1957] VICKERY: MIMULUS 129 Table 1. CHROMOSOME CouNTS IN MIMULUS, SECTION SIMIOLUS n=16 M. dentilobus Rob. & Fern. Sierra Charro, Chihuahua, Mexico. Gentry 8073 (5324). n=15 M. glabratus var. utahensis Pennell. Fremont River, Bicknell, Wayne County, Utah, altitude 7100 feet, Vickery 600 (5265). n=14 M. guttatus DC. Mono Inn, Mono County, California, altitude 6420 feet, Clausen 2043 C5015): Darwin Falls, Inyo County, California, altitude 2500 feet, U.C. Her- barium 696020 (5017). Mt. Diablo, Contra Costa County, California, altitude 1000 feet, Stebbins 703 (5052). Mt. Oso, Stanislaus County, California, altitude 1000 feet, Vickery 190 (5346). Bountiful, Salt Lake County, Utah, altitude 4800 feet, Vickery 331 (5835). Mill Creek Canyon, Salt Lake County, Utah, altitude 5800 feet, Vickery 335 (5840). Alta, Salt Lake County, Utah, altitude 8800 feet, Vickery 336 (5845). Kimball Junction, Summit County, Utah, altitude 6600 feet, Vickery 341 (5856). Hailstone, Wasatch County, Utah, altitude 6300 feet, Vickery 342 (S857) Rock Creek, below Davies Resort, Duchesne County, Utah, altitude 7600 feet, Del Wiens 8/5/56 (5968). 1955) with a group of South American taxa reported as 2n—32, 48, and ca. 64 by Sugiura, Maude, and Brozek, respectively (Darlington and Wylie, 1955). However, the distinctive morphological characteristics of M. dentilobus (Grant, 1924) suggest to the authors that it is not a link in this aneuploid series and is not closely related to any of the North or South American species of Mimulus. Furthermore, interspecific crosses (Vick- ery, 1956) indicate that M. dentilobus is genetically isolated from all the other taxa of its section. Therefore, it appears to be an evolutionary side- shoot from the main group of Szmiolus species. On the other hand, the morphology of M. glabratus HBK., which is the only species of the section common to North and South America, indicates relationships to both the North and South American groups of species. An n=14 culture of MW. glabratus var. utahensis (5048) hybridizes with members of the various taxa of the two groups although the hybrids are nearly sterile (Vickery, 1956). Therefore, in view of the possible evolutionary role of M. glab- ratus aS a connecting link between the North and South American taxa 130 MADRONO [Vol. 14 2019 SHOT S052 5265 9346 Sy = SIS) 9840 59845 9856 78D 7 9968 9324 ————) 6) 10 2Omicra Fic. 1. Meiotic chromosomes of pollen mother cells of Mimulus, * 750. All draw- ings were made with the aid of a camera lucida. The numbers below the figures are the culture numbers (Table 1). The ten cultures of WM. guttatus are n=14. The one culture of M. glabratus var. utahensis (5265) is n=15. M. dentilobus (5324) is n=16. All plants are in first metaphase except 5015, 5968, and 5324 which are in second metaphase. of section Szmiolus, the aneuploid number of n=15 for the Bicknell cul- ture (5265) of M. glabratus var. utahensis is particularly interesting and significant. Further work is in progress to be sure that this count does not represent merely an aberrant individual. The extensive hybridization experiments mentioned above (Vickery, 1956) reveal crossing barriers of various degrees between the different cultures of M. guttatus. However, in no case is a culture completely iso- lated from all the others. The results of these crosses suggested to the authors that all the cultures of the races of M. guttatus would have the 1957] JONES: RUDBECKIA 131 same chromosome number. Our cytological observations confirm this idea and suggest, further, that the crossing barriers are due not only to gene differences but also to differences in chromosome structure. For example, culture 5968 has markedly smaller chromosomes than the other cultures. Probably there are cryptic structural differences in the chromo- somes of the cultures as well. In conclusion, we may report that our studies indicate that the North American M. guttatus complex of species (n=14) appears to be related to the South American M. luteus complex (x=8) by a series of aneuploid forms of M. glabratus. Work is in progress to determine the chromosome numbers of additional taxa and to determine the chromosomal homologies of the various cultures and races in order to clarify further our under- standing of the evolutionary relationships in the group. Department of Genetics and Cytology University of Utah Salt Lake City 12, Utah LITERATURE CITED Buapuri, P. N. and P. N. Guosu. 1954. Chromosome squashes in cereals. Stain Technology 29: 269-276. CAMPBELL, G. R. 1950. Mimulus guttatus and related species. El] Aliso 2: 319-335. DariincTon, C. D. and A. P. Wytte. 1955. Chromosome atlas of flowering plants. Allen & Unwin Ltd., London. 519 pp. GranT, ADELE L. 1924. A monograph of the genus Mimulus. Ann. Mo. Bot. Gard. 11: 99-388. VickKERY, R. K., Jr. 1951. Genetic differences between races and species of Mimulus. Carn. Inst. Wash. Year Book 50: 118-119. . 1955. Chromosome counts in the section Simiolus of the genus Mimulus (Scrophulariaceae). Madrono 13: 107-110. . 1956. Data on interracial and interspecific hybridizations in the section Simiolus of the genus Mimulus (Scrophulariaceae). Proc. Utah Acad. Sci., Arts, and Letters 33: 45-64. ON THE SPECIFIC DISTINCTNESS OF RUDBECKIA LACINIATA AND R. AMPLA GEORGE NEVILLE JONES Wild goldenglow, Rudbeckia laciniata L. (Sp. Pl. 906, 1753), isa rather familiar plant growing in alluvial soil in eastern United States and adja- cent Canada, ranging from Quebec to Manitoba and southward to eastern Texas and Florida. A morphologically similar plant described from Colo- rado in 1901 as R. ampla A. Nels., occurring in the western parts of the continent from Saskatchewan to South Dakota, New Mexico, Arizona, and Idaho, is less well known, and generally has been treated by contem- porary students of the western flora as a synonym of R. laciniata. There is evidence, however, on the basis of study of morphological characters, habitat, and habit, as well as geographical distribution, that 132 MADRONO [Vol. 14 these plants represent different taxa, apparently two distinct species. An alternative interpretation is that R. ampla might be treated as a “‘variety”’ or subspecies of R. laciniata, as the historical trend of human migration had led to discovery and publication of the latter taxon 148 years earlier. However, if one of them should be regarded as a subspecies of the other, the reverse process of designating R. laciniata as a subspecies of R. ampla is for certain reasons indicated as the more logical course. According to a theory of origin of the genus Rudbeckia expressed by W. M. Sharp (Ann. Mo. Bot. Gard. 22:60. 1935), the present species are descended from an ancestral group originating on the Appalachian or the Ozarkian upland regions exposed since the close of the Paleozoic, the most recent of them inhabiting the Coastal Plain from Virginia to Texas. From this it would appear that R. /aciniata and R. ampla, although mor- phologically similar, may have had different origins, the latter being much older, and the former of relatively recent origin on or near the Coastal Plain. It would be illogical to treat R. ampla as a subspecies of R. lacini- ata simply because the latter was discovered first. It seems therefore most practical to continue to designate each of them by a binary name. This procedure serves to express relationship satisfactorily, and there is no need for “new combinations” or other nomenclatural change. Some of the diagnostic characters of these two species are summarized in the following key: Heads (“disk”) 1.5-2 cm. in diameter, becoming ellipsoid and 3-4 cm. in length; rays 9-15 mm. wide; disk corollas 5 mm. long; achenes 5-6 mm. long; receptac- ular bracts (“chaff”) linear, 6-7 mm. long; pappus coroniform, the teeth short, obtuse, indistinct or none; leaves glabrous beneath, the upper surface somewhat strigose toward the apex of the leaf, the trichomes with enlarged pustular bases; basal leaves palmately ternate with broad segments and narrow sinuses; plants 1-1.5 m. tall; peduncles few, stout (1.5-3 mm. thick below the head)....R. ampla Heads (‘disk’) globose, 1-1.5 cm. in diameter; rays usually less than 1 cm. wide; disk corollas 3-4 mm. long; achenes 3-4 mm. long; receptacular bracts spatulate, 3-4 mm. long; pappus crown of 4 teeth; upper leaves glabrous above, usually finely strigose beneath varying to nearly glabrous; basal and lower cauline leaves pinnately divided into relatively narrow segments with wide sinuses; plants 1.5—3 m. tall; peduncles several, slender (about 1 mm. thick below the head) R. laciniata The relevant bibliographical references to R. ampla follow. It is not necessary to list here the extensive bibliography of R. laciniata L. Rudbeckia ampla A. Nelson, Bull. Torrey Club 28: 234, 1901; Ryd- berg, Fl. Colorado 371, 1906, Fl. Rocky Mts., 927, 1917, Fl. Pr. & PI. 836, 1932. Rudbeckia laciniata sensu A. Gray, Syn. Fl. 1(2): 262, 1878, ex parte; A. Nelson, Man. Bot. Rocky Mts. 544, 1909; Kearney & Peebles, Flowering Pl. Arizona 950, 1942, Arizona Flora 898, 1951; Davis, FI. Idaho 777, 1952; Weber, Handb. Pl. Colorado Front Range 194, 1953; Harrington, Man. Pl. Colorado 598, 1954; non L., 1753. Type locality. Colorado. The first cited specimen is Baker 699 from Pagosa Springs. 1957] ZOBEL: MEXICAN PINES 133 To those who may be familiar with these plants the habit and appear- ance of members of the two taxa are noticeably different, although, as often happens in descriptive taxonomy, some of the differences are not easily stated in objective terms. In general, however, Rudbeckia ampla is a shorter, somewhat stouter plant with larger heads on fewer and stouter peduncles, the rays averaging somewhat wider. The disk corollas, achenes, and receptacular bracts are longer, the latter, commonly known as “‘chaff,” furnishing distinctive characters that have been found to be important in the taxonomy of other species of the genus. Finally, it may be pointed out that the pappus of R. ampla is more coroniform, and the leaves are usually thicker and with different indument. In some specimens of R. laciniata the undersurface of the leaves tends to be less pubescent than in others, but almost all leaves show at least a few trichomes, particularly when viewed under the binocular microscope. It may be appropriate to note’ here that some published statements concerning R. laciniata are perhaps not as explicit as they should be. Unless modified, they may serve as a barrier to clarity of understanding of the plants described. For example, in the eighth edition of Gray’s Manual the statements referring to the elongation of the “disk” of R. laciniata, as well as that indicating length of achenes, appear to be somewhat exaggerated, and are more nearly applicable to R. ampla. Department of Botany, University of Illinois, Urbana PINES FROM NUEVO LEON, MEXICO Bruce ZOBEL AND FRANKLIN CEcH! The forest geneticist usually attempts to grow as many as possible of the different strains and species of trees in which he is interested. These may be used for two purposes: 1) to establish a breeding arboretum, the trees to be used as parents for desired crosses; 2) to establish, on a limited scale, trials of adaptability to local or special environmental conditions. Both purposes were in mind when the Texas Forest Service sponsored two pine collection trips to the state of Nuevo Leon, Mexico. In conjunction with drought resistance studies, species that can do well under severe conditions of heat and drought were especially sought. The accrual of further knowledge of the taxonomic and evolutionary position of the Mexican pines was another objective of importance. The purpose of this paper is to report on the several pine species and their many intergrading forms found growing in one of the states of northeast- ern Mexico, Nuevo Leon. Although collections were not made at all points in the state, the three areas visited were intensively studied, and 1 Silviculturist and Assistant Silviculturist, respectively, Texas Forest Service. 134 MADRONO [Vol. 14 To Monterrey NUEVO LEON Linares Galeana Iturbide Rafces TAMAULIPAS SAN LUIS Ascension Bere Escondida p Aram berr) N ‘ A ’ Zaragoza La Encantada «” Cerro de la Bandera in Sierra Madke Oriental elb Miles ) scale paved road dirt road i ee ea horseback Fic. 1. Southern end of Nuevo Leon showing major area from which collections were made. The other two collecting areas were at more northerly points in the state. complete records, specimens, and photographs were obtained for each species encountered. The collection trips were the direct result of an invitation sent by Dr. E. A. Pequeno, Director, Instituto de Investigaciones Cientificas, Uni- versidad de Nuevo Leon. Though not a forester, Dr. Pequeno has an intense interest in the research upon and wise use of Mexico’s forest resources, for he is fully cognizant of what their destructive exploitation will do to the economy of the country. 1957] ZOBEL: MEXICAN PINES 2) ‘6 Fic. 2. Vicinity of La Encantada, illustrating the type and extent of the virgin forests in this area. A true fir (Abzes) is in foreground. The first trip was made in late September, 1954, with the senior author accompanied by Mr. Ruben Rocha, a Mexican student at Agricultural and Mechanical College of Texas, who acted as interpreter. This trip was preliminary in nature and little actual material was collected, because few pines bore cones that fall. However, many yearling cones indicated the possibility of a large cone crop for the following year. It was deter- mined on this first trip that although a few species of pines matured their cones in September-October, the majority did not mature their cones until middle or late December or even in January. Accordingly, the second trip was made late in December 1955. Both authors plus Mr. Chester Rowell, taxonomist from the Biology Depart- ment of the Texas A & M College and Mr. Rocha went on the second trip. The cone crop was plentiful on most species and abundant collections were possible. The area from which most of the collections were made is in the extreme southern end of Nuevo Leon (fig. 1). As one goes into the mountains from Linares, the road which follows the stream bed rises sharply, going through the village of Iturbide and then breaking out onto a plateau near Galeana. This unimproved road then proceeds in a southerly direction down the “plateau” through the towns of Ascension, Aramberri, and (in 1954) to Zaragoza. From Zaragoza a trail leads up the mountain to the village of La Encantada, at an elevation of about 10,000 feet (fig. 2). Collections were made along this entire route wherever pines occurred, 136 MADRONO [Vol. 14 Fic. 3. Pinus arizonica var. stormiae on nearly solid rock in a very dry area near Ascension. most, however, being made up the mountain from Zaragoza and around La Encantada. The collecting area described above varied from nearly desert near Galeana, Iturbide, and Ascension (fig. 3), with less than 20 inches rain- fall, to the well-watered mountain forests around La Encantada. The higher elevation collections around La Encantada were in virgin forests, which included conifers of the genera Abies (fig. 2) Pseudotsuga, and Taxus. This is a most beautiful region, lush with vegetation. It is very similar in character to the middle-high elevations on the west side of the Sierra Nevada of California. Access is difficult, due to the poor roads and the necessity for horseback travel over considerable distances. This limited the specimens that could be collected. As far as could be deter- mined, the pines around La Encantada had not previously been collected and classified, though Martinez (1948) mentions pine specimens from the area near Zaragoza. Small collections were made from two other areas, one on Chipinque near the city of Monterrey, and the other in the Sierra Picachos near the village of Sombreretillo. This latter area is several miles to the east of the point where the Inter-American Highway from Laredo to Monterrey crosses the small range of mountains south of Sabinas Hidalgo. All timbered regions at the lower elevations have been severely depleted. A crude method of bleeding for gum (oleoresin) is used extensively in the lowland areas, and in many cases so intensively and carelessly as to kill the trees (fig. 4). Fires and over-grazing have taken a huge toll and in 1957] ZOBEL: MEXICAN PINES 137 aS Fic. 4. Method of bleeding tree for gum (oleoresin). Note “cup” chopped in tree at base of peeled face. many of the very scattered older stands, reproduction is lacking. There are indications in some areas that previously pine-covered land may degen- erate to desert scrub if present treatment continues. Conversely the high mountain, virgin forests contain considerable overmature timber that is dying out. This could well be harvested if there were any method whereby the timber could be economically transported to markets. 138 MADRONO [Vol. 14 In 1955, twenty-eight separate pine collections were made from the three areas visited. Each included: 1) ripe cones, up to a bushel from a single tree (these ‘‘green” cones were brought to Texas and the seed was extracted there) ; 2) a specimen fruiting branch, which contained mature cones, yearling cones, undamaged terminal bud, and typical foliage. These fruiting branches were pressed in the field; before this was done, however, a detailed study was made of their needles, buds, and other features, and colored photographs were taken to record foliage color, needle length and number, color and shape of unopened cone, color, position, and number of yearling cones, bud color, fuzziness, etc. In addition, needle bundles were preserved in formaldehyde-acetic acid-alcohol solution; these were later sectioned for the study of needle anatomy and morphology. Current literature (Loock, 1950; Martinez, 1945 and 1948; Dallimore and Jackson, 1948) proved inadequate to the task of identification owing to the many recent changes in Mexican pine taxonomy and to the fact that the collections showed much intergradation. In order to make satisfactory species identifications of our collections, it thus became necessary to make a thorough analysis of all characters, including cross-sections of the needles, and then refer back to original sources in the literature of the genus. SEEDLING CHARACTERISTICS In all books referred to, there is a notable absence of reference to seed- ling characters. Such characters are of importance, especially since many of the Mexican species show a tendency to have delayed height growth, i.e., a semi-grass stage of development. This tendency towards a grass stage of development might well be a key diagnostic feature, as is true for the intensively studied “southern pines” in the longleaf-loblolly-slash pine complex. A tree with a true grass stage grows one to many years putting on only very little height growth, the needles being borne near the ground and appearing as a clump of “grass.” A tree with a semi-grass stage makes only very little height growth the first year, but some stem is evi- dent; in subsequent years it may, or may not, put on rapid height growth. The seedlings of the material collected varied from nearly a true grass stage to normal height growth (fig. 5). Seedlings of all collections in this study have gone through one growing season in the nursery. Their characters will be discussed in the species list which follows. PINuS SPECIES COLLECTED IN NUEvVo LEON Many species of pines, and their intergrades, were found in Nuevo Leon, with several species commonly growing completely intermixed. Mexican pines do not display some of the usual differences held to be diagnostic of hard- or yellow-pines as contrasted with the soft- or white- pines. For example, many of the Mexican hard-pines contain five to seven or even more needles per fascicle, and five needles might be considered to be the rule rather than the exception. Conversely, the only true white- pine collected on this trip (Pinus ayacahuite var. brachyptera Shaw) had 1957] ZOBEL: MEXICAN PINES 139 Fic. 5. Seedlings of Pinus hartwegii (on left) showing tendency toward a modi- fied grass stage; seedlings of P. teocote (on right) showing normal growth in height and a basal crook in stem similar to shortleaf pine (P. echinata). many fascicles containing only three needles instead of the normally expected five. The five-needle hard-pines were most difficult to classify. Especially sO were six specimens that exhibited characters of P. montezumae Lamb., P. montezumae var. lindleyi London, P. pseudostrobus var. estevezi Mar- tinez, and P. durangensis forma quinquefoliata Martinez. Their charac- ters did not fit any taxon completely, and after careful study the conclu- sion was drawn that they were intermediate or intergrading forms. Such intermediacy might well be called the rule here, especially for the five- needled pines. Both Loock (1950) and Martinez (1948) call attention to the intergrading forms and the complexity of the classification. Four trees were found that had characters sufficiently dissimilar to previously described species so that they could not be definitely classified. These are included at the end of the species list. Three of the pine species in the list which follows have not been previously reported growing in the state of Nuevo Leon (Loock, 1950, and Martinez, 1948). In addition to the pines, collections were also made of the genera 4 dies, Pseudotsuga and Taxus. The authors at first thought that Taxus had acquired a new distribution record, until it was found that Hernandez et al. (1951) reported Taxus in the neighboring state of Tamaulipas. Many different oaks (Quercus ) were observed, as well as other hardwoods such as Madrono (Arbutus). This is an area rich in many forms of vege- tation and an extended collection trip would be most worthwhile. 140 MADRONO [Vol. 14 PINUS AYACAHUITE var. BRACHYPTERA Shaw. This species had inter- mediate characters, in many ways closely resembling P. reflexa Engelm. Needles were predominantly in fascicles of five, with many fascicles of 3, 4, and 6 being present on mature trees. Collections were made near La En- cantada at an elevation of approximately 10,000 feet. This pine was grow- ing among hard-pines, Abies, Pseudotsuga, and Taxus. The trees were very large on the better sites. Cones mature early (October and November) so no seeds were obtained. PINUS ARIZONICA var. STORMIAE Martinez. Several collections of this species were made, from near Ascension to the area around Galeana, at elevations from 5,000 to 6,000 feet. Classification was accurate. This species was growing on extremely rugged sites, with low rainfall. One collection was made on nearly pure calcareous soil, with little other vege- tation being present. Seedlings, as well as mature trees, all had needles in fascicles of three. All seedlings had delayed height growth, a semi- grass stage of development. PINUS CEMBROIDES Gord. Collections were made near Ascension, at ele- vations around 5,000 feet; they were easily classified. This species was very widespread. Needles were mostly in fascicles of 2, a few fascicles having 3. Cones mature early and no seeds were collected. PINUS HARTWEGII Lindl. Several collections were made of this species, most of which were easily classified. One collection had characters very similar to P. rudis. All collections were made near La Encantada, at ele- vations around 10,500 feet, in nearly pure stands. Needles were predom- inantly in fascicles of 5, some of 4. Seedlings all had their needles in fas- cicles of 3, and had a definite semi-grass stage (fig. 5). The seedlings looked somewhat like those of P. pseudostrobus, but were coarser. PINUS PSEUDOSTROBUS Lindl. Two collections of this species were made, one fitting the classification well, the other having affinities with P. hart- wegi. Collection was near La Encantada at elevations around 10,000 feet. These trees were growing in a mixed pine and fir stand, just below the pure P. hartwegit. Needles were mostly in fascicles of 5, with some of 6 and 7 present. Seedlings had needles in fascicles of 3, 4, and 5, and had a definite semi-grass stage similar to P. hartwegit. PINUS PSEUDOSTROBUS var. ESTEVEZI Martinez. This species was col- lected only west of Iturbide at an elevation of 5,000 feet. It was fairly easily classified, though all characters did not exactly fit. It was growing under relatively droughty conditions. Needles were in fascicles of 5, with some 4 and 6. Seedlings had needles in fascicles of 5. Unlike P. pseudo- strobus, height growth of the seedlings was normal. PINUS PSEUDOSTROBUS forma PROTUBERANS Martinez. It was difficult to classify this species accurately since no foliage was collected. Cones were collected at dusk from three trees, all appearing to be the same until cones were closely examined. Two of the trees remain unclassified, having needle anatomy like P. hartwegi, foliage like P. rudis, and cones similar to P. pseudostrobus var. estevezi. The third tree had cones exactly like P. 1957] ZOBEL: MEXICAN PINES 141 pseudostrobus forma protuberans. Collection was made near Ascension, at an elevation of 6,000 feet. Seedlings had a semi-grass stage, with needles in fascicles of 3,4, and 5. This species has never been reported in Nuevo Leon, only from the more southern states. PINUS MONTEZUMAE Lamb. Several collections were made of this spe- cies, and classification was from very exact to doubtful. Two collections were very similar to P. montezumae lindleyi and also to P. pseudostrobus. Specimens were obtained near Zaragoza, Escondida, Ascension, Iturbide, and in the Sierra Picachos near Sombreretillo, ranging in elevation from 3,500 to 6,000 feet. Needles were in fascicles of 5, with an occasional 4. The needles of seedlings were in fascicles of 3, 4, and 5, the predominant number varying from tree to tree. Seedlings had extra long needles and had normal height growth (no grass stage). This was different from seedlings obtained from seed previously sent to us as P. montezumae, all of which had a pronounced grass stage. PINUS MONTEZUMAE var. LINDLEYI Loudon. Classification of this spe- cies is only partially satisfactory, some characters not fitting too well. It was collected near Zaragoza at 5,000 feet elevation. Needles were in fascicles of 5, with some of 4. Seedlings had needles in fascicles of 5 and had normal height growth. This species has not previously been reported in Nuevo Leon, only from the states to the south, including Hidalgo, Vera Cruz, and Querétaro. PINUS TEOCOTE Schl. and Cham. Of the several collections of this spe- cies, only one was in doubt, because of the needle anatomy. Collections were made near La Encantada, Ascension, and at Chipinque, near Mon- terrey. Elevations ranged from 4,500 feet to 10,500 feet. Needles were short, always in fascicles of 3. Seedlings had all their needles in fascicles of 3, and had normal height growth (fig. 5). All had an unusual crook at the base of the stem, similar to that found in shortleaf pine (P. echinata Miller). Just as for P. montezumae, seed sent to us under the name of P. teocote had a grass stage, while none of our own collections showed this feature. PINUS DURANGENSIS forma QUINQUEFOLIATA Martinez. Three collec- tions were made of this species, none of which were classified with exact- ness. Characters were found similar to P. pseudostrobus estevezi, P. montezumae and P. pseudostrobus. This species has not previously been reported in Nuevo Leon, being found to the west in the states of Chihua- hua and Durango. Our collections were made at La Encantada, Chipinque, and Sombreretillo. Needles were in fascicles of 5, with a few 4, and occa- sionally 6. Seedlings from two of the trees had needles in fascicles of 5, while one (from Sombreretillo) had them of 3, 4, and 5. Seedlings were tall (no grass stage), with long needles. Pinus RupiIs Endl. One poorly classified collection was made of this species, the tree having characters like P. montezumae lindleyi and P. pseudostrobus. Collection was made at La Encantada, at 9,500 feet eleva- tion. Needles were in fascicles of five. Seedlings were tall and had needles in fascicles of three, four, and five. MADRONO 142 1957] ZOBEL: MEXICAN PINES 143 The following Pinus collections from Nuevo Leon remain unclassified: TREE No. 12 (fig. 6). Growing on a very dry site near Ascension, with cacti and a few xerophytic shrubs. Surprisingly vigorous for a tree grow- ing on such a poor site. Foliage with needies all in fascicles of five, about 20 cm. long, thin, flexible, slightly drooping; resin ducts two to four; endodermal outer wall thickened; hypodermis with moderate intrusions into chlorenchyma; two vascular bundles touching but distinct. Cone 8 cm. long; peduncle about 2 cm. long; color greenish black when unopened, changing to dark brown with drying; apophysis strongly keeled, somewhat reflexed on dorsal part of cone; umbo very small, prickle not sharp, not deciduous. Seedling with needles in fives, thin; having normal height growth. TREE No. 16 (fig. 6). Collected only in one small valley above Zara- goza, growing intermixed with oaks at 6,000 feet elevation. Given by us the name “‘blue pine” due to the very bluish color of the foliage. Appearing to belong to the P. rudis-P. hartwegi complex, with intermediate char- acters. Foliage with needles predominantly in fives, about 18 cm. long, very stiff, twisted, glaucous blue-green; resin ducts seven to ten, medial; endodermal walls thin; hypodermis intruding slightly into chlorenchyma; two vascular bundles separated by several cell thicknesses of transfusion tissue. Cone about 10 cm. long, purplish black when unopened, olive brown when dried, asymmetric; apophysis flat to somewhat raised; umbo grey, sunken, very small. Seedling with needles in bundles of three, coarse; with modified grass stage. TREE No. 25 (fig. 6). Growing in creek bottom, very rocky, Sierra Picachos, near Sombreretillo. Foliage with needles in threes, fours, and fives, predominantly fives, about 23 cm. long, moderately heavy, droop- ing, blue-green; resin ducts three to six, medial; endodermal outer walls thickened; two vascular bundles close together, not distinct; hypodermis very thick with small intrusions. Cone (old, hanging on tree; no fresh cones) about 5 cm. long; peduncle about 1.5 cm. long; apophysis raised with flat face, slightly keeled; umbo flat to sunken. Seedling none. TREE No. 28 (fig. 6). From Sierra Picachos, near Sombreretillo. Needle number distinguishing this tree from P. durangensis and similar species. Foliage with needles predominantly in threes, few fours and fives, around 25 cm. long, somewhat pendent, moderately thin; resin ducts three and four, medial and one internal; endodermal outer walls variable, thick to slightly thick; two vascular bundles touching but distinct ; hypodermis thick, with considerable intrusions. Cone about 9 cm. long, black to purplish green unopened, drying to light olive brown, somewhat asym- metrical; apophysis on dorsal side considerably reflexed, strongly keeled; umbo raised, ashy gray, with recurved, short, stout prickle. Seedling with needles in threes and fours, occasionally fives, thick; with normal height growth. 144 MADRONO [Vol. 14 ACKNOWLEDGEMENTS The authors wish to thank all those who made these trips to Mexico possible and profitable. Dr. E. A. Pequeno is to be especially recognized for suggesting the trips, for accompanying the authors on both trips, and for helping to defray expenses incurred while in Mexico. He is to be com- mended for his interest in the natural resources of Mexico. We hope his conservation efforts will be successful. Mr. Erasmo Cerda of Aramberri hosted both collection parties, and supplied housing, food, horses and guides necessary for the trip to La Encantada. Mr. Chester Rowell helped considerably on the second trip with his knowledge of both the Mexican flora and language, as well as helping greatly on the photography. The Medellin brothers acted as guides from Zaragoza, and welcomed the parties to La Encantada. Their help and friendliness certainly aided in making the North American visitors feel welcome in Mexico. The several others that went as helpers or students on the trips were indeed appreciated. Everyone present contributed something to help enrich the value of the collections. School of Forestry, Forest Genetics, North Carolina State College, Raleigh and Forest Genetics Laboratory, Texas Forest Service, College Station, Texas LITERATURE CITED DaLimoreE, W., and A. BrucE Jackson. 1948. A handbook of Coniferae including Ginkgoaceae. Edward Arnold & Co., London. 682 pp. Haritow, Wirti1Am M., and Ertwoop S. Harrar. 1941. Textbook of dendrology. McGraw-Hill Book Company, Inc., New York. 541 pp. HERNANDEZ X., Err1AM, Howarp Crum, Wo. B. Fix, and A.J. SHarp. 1951. A unique vegetational area in Tamaulipas. Bull. Torrey Club 78:458-463. Loock, E. E. M. 1950. The pines of Mexico and British Honduras. Bull. No. 35, Dept. For. Union of South Africa. 244 pp. MartTinEz, MaxtmMino. 1945. Las Pinaceas Mexicanas. Vol. I. Mexico. 345 pp. . 1948. Los Pinos Mexicanos. 2nd ed. Mexico. 361 pp. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication should not exceed an estimated 20 pages when printed unless the author agree to bear the cost of the ad- ditional pages at the rate of $15 per page. Illustrative materials (includ- ing “typographically difficult” matter) in excess of 30 per cent for papers up to 10 pages and 20 per cent for longer papers are chargeable to the author. Subject to the approval of the Editorial Board, manuscripts may be published ahead of schedule, as additional pages to an issue, provided the author assume the complete cost of publication. Shorter items, such as range extensions and other biological notes, will be published in condensed form with a suitable title under the general heading, “Notes and News.” Institutional abbreviations in specimen citations should follow Lanjouw and Stafleu’s list (Index Herbariorum. Part 1. The Herbaria of the World. Utrecht. Second Edition, 1954). Articles may be submitted to any member of the Editorial Board. MADRONO A WEST AMERICAN JOURNAL OF BOTANY A quarterly journal devoted to the publication of botanical re- search, observation, and history. Back volumes may be obtained from the Secretary at the following rates: Volume I, 1916-1929 . . . . . $5.00 Volume II, 1930-1934. . . . . 5.00 Volume ITI, 1935-1936 . . . . 5.00 Volume IV, 1937-1938 . . . . 5.00 Volume V, 1939-1940. . . .. . 5.00 Volume VI, 1941-1942 . . . . 5.00 Volume VII, 1943-1944 . . . . 5.00 Volume VIII, 1945-1946 . . . . 5.00 Volume IX, 1947-1948 . Se tas 6.00 Volume X, 1949-1950. . . . . 7.00 Volume XI, 1951-1952 . . . . 8.00 Volume XII, 1953-1954 . . . . 8.00 Volume XIII, 1955-1956 . . . . 8.00 Single numbers, Volumes.) and’2. 2° 2 ae .50 Succeeding numbers, (except Volume XI, No.2). . 1.00 Volume XI, No.2 007. ee 1.50 The subscription price of MADRONo is $4.00 per year. If your institution does not now subscribe to MADRONO, we would be grateful if you would make the necessary request. Since there is no extra charge for institutional subscriptions, an individual may hold membership in the California Botanical Society on the basis of his institution’s subscription. Address all orders to: G. THoMAs Rossins, Corresponding Secretary Department of Botany University of California, Berkeley 4, California PRONO VOLUME 14, NUMBER 5 JANUARY, 1958 Contents PAGE PINUS OAXACANA, A NEW SPECIES FROM MEXICco, N.T. Mirov 145 CHROMOSOME COUNTS IN SECTION ERYTHRANTHE OF THE GENUS MIMULUS (SCROPHULARIACEAE), Robert K. Vickery, Jr., Barid B. Mukherjee, and Delbert Wiens 150 Two NEw SPECIES OF PENSTEMON IN COLORADO, C. William T. Penland 153 PECULIARITIES OF THE COLUMBIA RIVER GORGE FLorRA, LeRoy Detling 160 THE GENUS COLLINsIA. III. THE SIGNIFICANCE OF CHIASMATA FREQUENCIES AS A CYTOTAXONOMIC Toot, E. D. Garber Ei A WEST AMERICAN JOURNAL OF BOTANY > SSPare PUBLISHED QUARTERLY BY THE CALIFORNIA BOTANICAL SOCIETY ie MADRONO A WEST AMERICAN JOURNAL OF BOTANY Entered as second-class matter at the post office at Berkeley, California, January 29, 1954, under the Act of Congress of March 3, 1879. Established 1916. Subscription price $4.00 per year. Published quarterly and issued from the office of Madrofio, Herbarium, Life Sciences Building, University of California, Berkeley 4, California. BOARD OF EDITORS HERBERT L. Mason, University of California, Berkeley, Chairman EpcarR ANDERSON, Missouri Botanical Garden, St. Louis. LymMAN BENSON, Pomona College, Claremont, California. HERBERT F. COPELAND, Sacramento College, Sacramento, California. Joun F. Davipson, University of Nebraska, Lincoln. Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts. Mitprep E. Matus, University of California, Los Angeles 24. Marion Owneey, State College of Washington, Pullman. IrA L. Wiccins, Stanford University, Stanford, California. Secretary, Editorial Board — ANNETTA CARTER Department of Botany, University of California, Berkeley. Business Manager and Treasurer—WINSLow R. Briccs Department of Biology, Stanford University, Stanford, California CALIFORNIA BOTANICAL SOCIETY, INC. President: John M. Tucker, Department of Botany, University of California, Davis, California. First Vice-president: James R. Sweeney, San Francisco State Col- lege, San Francisco, California. Second Vice-president: Reid V. Moran, San Diego Natural History Museum, San Diego, California. Recording Secretary: Mary L. Bow- erman, Department of Botany, University of California, Berkeley, California. Cor- responding Secretary: Jean H. Langenheim, Department of Botany, University of California, Berkeley, California. Treasurer: Winslow R. Briggs, Department of Biol- ogy, Stanford University, Stanford, California. 1958] MIROV: PINUS 145 PINUS OAXACANA, A NEW SPECIES FROM MEXICO N. T. Mrrov In Oaxaca and adjacent states of Mexico is found a pine with very long, prominent projections of the apophyses of the cone scales. Though pre- viously identified as a variety of Pinus pseudostrobus Lindl., this pine differs from P. pseudostrobus especially in the cones and in the chemical composition of the turpentine, and it is here published as a new species." Pinus oaxacana Mirov, sp. nov. P. pseudostrobus var. apulcensis Shaw pro parte, as to two specimens cited, not as to type, Pines of Mexico, p. 19, 1909. Non P. apulcensis Lindley, Edwards’ Bot. Reg. 25, Misc. 63, 1839. P. pseudostrobus var. oaxacana Martinez as to description, not as to type, Las Pinaceas Mexicanas 1:195. 1945. Oaxaca pine. Pinus subgenus D1- ploxylon Koehne. Arbor 20-30 m. alta, ramulis vernis uninodalibus, glau- cis; folia 5 in fasciculo, 20-33 cm. longa, serrulata, tenuissima, flexilia, pendula; stomata dorsalia pleurumque 5-7 seriebus et stomata ventralia 3—5 seriebus untrinque; hypodermis 2—4 seriebus cellularum, uniformis vel multiformis; ducti resinosi mediani, 2-4; endodermis cum septis ex- terioribus cellularum crassis; fasciculi vasculares 2, approximati; vaginae ca. 28-18 mm. longae, persistentes; strobili subterminales, 1-3, subses- siles, 10-14 cm. longi, clausi ca. 6 cm. lati, aperti ca. 9-11 cm. lati, ovoidei vel conici, acuti, leviter asymmetricales vel obliqui, ad maturationem aperientes, decidui super squamas infimas; apophyses ca. 12—20 mm. latae et ca. 8-12 mm. altae, rhomboideae, crassae, carinatae, cum projectione prominenti elongata 5-22 mm. longa, basi 5-12 mm. lata et 3-8 mm. crassa, pyramidali vel conica, dura, complanata, recta vel curvata et re- flexa; umbo in parte exteriore projectionis in puncto brevi terminens; pro- jectiones inaequales, ad latum abaxialem strobili longiores; semen 6—7 mm. longum, obovoideum, atro-brunneum, cum ala separabili brunnea ca. 20 mm. longa et 8-9 mm. lata. Resina terebintha n-heptane, 21 per centum; dextro- et dextro, laevo- a-pinene, 51 per centum; laevo- et dextro, laevo-limonene, 15—16 per centum; n-undecane, 1.3 per centum; et sesquiterpene, longifolene, 7.5 per centum componitur. Tree 20-30 m. tall, the spring shoots uninodal, glaucous; leaves 5 in a fascicle, 20-33 cm. long, serrulate, very slender, flexible, drooping; dorsal stomata mostly 5—7 rows and ventral stomata 3—5 rows on each side; hy- podermis of 2—4 layers of cells, uniform or multiform; resin ducts medial, 2-4; endodermis with outer cell walls thick; vascular bundles 2, close together; sheaths about 28-18 mm. long, persistent; cones subterminal, 1-3, subsessile, 10-14 cm. long, ovoid or conic, acute, slightly asymmet- rical or oblique, opening at maturity, deciduous above lowest scales; scales 1 Dr. Elbert L. Little, Jr., of the United States Forest Service, assisted in prepar- ing the Latin and English descriptions and in checking the nomenclature. Maprono, Vol. 14, No. 5, pp. 145-176. Jan. 27, 1958. Be bid 146 MADRONO [Vol. 14 with apophyses ca. 12-20 mm. broad and 8-12 mm. high, rhomboidal, thick, keeled, the apophyses with projections prominent, elongate, un- equal (those on abaxial side of cone longer), 5—22 mm. long, 5-12 mm. wide at base, 3-8 mm. thick at base, pyramidal to conic, hard, flattened, straight or curved and reflexed, the umbo on outer part of each projection ending in a short point; seeds 6—7 mm. long, obovoid, dark brown, with detachable brown wing ca. 20 mm. long and 8—9 mm. wide. The turpentine is composed of n-heptane, 21 percent; d, dl-«-pinene, 51 percent; 1, dl-limonene, 15—16 percent; n-undecane, 1.3 percent; and a sesquiterpene, longifolene, 7.5 percent.” A herbarium specimen which serves as a voucher for the turpentined trees was collected near Rancho Nuevo, 65 kilometers southwest of San Cristobal de Las Casas, Chiapas, Mexico (Mirov, in 1951) and is deposited at the Institute of Forest Genet- ics, Placerville, California. Holotype. Near La Parada, Oaxaca, Mexico, altitude 7,500—9,000 feet, August 18, 1894, E.W. Nelson 985 (US 398558). A slightly reduced draw- ing of the cone of the holotype which is about 13 cm. long, was reproduced by Shaw (1909, pl. 12, fig. 8; 1914; pl. 24, fig. 214). Another cone illus- trated by Shaw (1909) is: E.W. Nelson 2539 (US 398583), Miahuatlan, Oaxaca. Under the name P. pseudostrobus var. oaxacana Martinez, detailed de- scriptions and good illustrations of P. oaxacana were published by Mar- tinez (1945, pp. 195-201; 1948, pp. 202-9) and by Loock (1951, pp. 161— 164). In addition, both of these writers recorded it as occurring in the states of Oaxaca, Mexico, Puebla, Guerrero, Veracruz, and Chiapas. In his treatment of the pines of Mexico, Shaw (1909) included this taxon as a variety of P. pseudostrobus Lindley (Edwards’ Bot. Reg. 25, Misc. 63, 1839), basing his concept on Lindley’s P. apulcensis (ibid.) and characterizing it as having ‘“‘a greater or less prolongation of the apo- physes.”’ Shaw cited and illustrated three specimens: £.W. Nelson 985, La Parada, Oaxaca; E.W. Nelson 2539, Miahuatlan, Oaxaca; and Pringle 8788, Eslava, Distrito Federal. He assumed that all these specimens, as well as a cone collected by Hahn in 1866 at Cofre de Perote, Veracruz, and in the Museum d’Histoire Naturelle in Paris, were the same as P. apul- censis Lindl. from Apulco, Hidalgo. Martinez (1945, pp. 168-201; 1948, pp. 184-209) recognized within P. pseudostrobus six variations (the typical variety, four named varieties and one form), differing chiefly in cone scale characters. Two of these named varieties and the form are pertinent herein. He showed clearly that the two Nelson Oaxaca collections, cited by Shaw as P. pseudostrobus var. apulcensis, belong to another taxon which he (Martinez) illustrated and described (in Spanish, loc. cit.) as P. pseudostrobus var. oaxacana. Unfortunately, Martinez failed to cite a type specimen, to publish a Latin * This information was obtained from one sample, in which oleoresin from 25 trees was combined. In other samples the percentages may be different. Of significance is the presence of large quantities of a paraffin hydrocarbon, normal heptane. 1958] MIROV: PINUS 147 Fic. 1. Cone of P. pseudostrobus (left) ; cone of P. oaxacana (right). Both ca.x 14. Photo courtesy California Forest and Range Experiment Station, U.S. Forest Service. diagnosis, or to designate the synonym P. pseudostrobus var. apulcensis (Lindl.) Shaw as applicable to his material only “in part.” Therefore, although his intent was clearly otherwise, the name P. pseudostrobus var. oaxacana of Martinez must be referred to the type specimen (Hartweg in 1838, Apulco, Hidalgo) upon which both P. apulcensis Lindl. and P. pseu- dostrobus var. apulcensis (Lindl.) Shaw are based. Hence, it has been necessary to present a formal description and typification of the taxon represented by the Nelson Oaxaca specimens. To the pine from Apulco (P. apulcensis Lindley), Martinez (1945, p. 192; 1948, p. 199) assigned the name P. pseudostrobus var. apulcensis, an epithet which was super- fluous when published, inasmuch as Shaw had already published this com- bination in 1909. Under P. pseudostrobus var. apulcensis, Shaw cited a third collection with least prominent apophysis, Pringle 8788 from Eslava, Distrito Fed- eral. This collection had been distributed as P. pseudostrobus but had the synonym P. protuberans Roezl also on the printed label. Pringle $788 is therefore referred to P. pseudostrobus forma protuberans Martinez (1945, p. 184; 1948, p. 192), which was characterized by the protuberant apo- physis ending in a sharp point and which was recorded from Eslava as well as other localities. However, it becomes desirable to establish the identity of the century- old prior species P. protuberans Roezl (Cat. Grain. Conif. Mex. 27, 1857) from 9,000—10,000 feet altitude near Contreras, Distrito Federal. Bene- 148 MADRONO [Vol. 14 dict Roezl (1824-85), a Czech plant collector and horticulturist, in 1857 and 1858 published names of nearly one hundred new species of Pinus, mostly from the vicinity of Mexico City, in two commercial catalogs of Mexican conifers. The original descriptions of Roezl’s species were repub- lished and translated by other authors, of which the following may be cited for P. protuberans: German, condensed, by Otto (Hamburg. Gart. Blumenzeit. 13:408. 1857); English by Gordon and Glendinning (Pine- tum 259. 1858); Latin by Schlechtendal (Linnaea 29:348. 1858); and the original French by Carriére (Traité Gén. Conif. Ed. 2, 522. 1867). Actually, Roezl characterized the five members of his Section VIII of Pinus as having five long needles and very prominent apophyses and pro- tuberances, characters which indicate close relationship with the P. pseu- dostrobus complex and with P. oaxacana. Besides P. protuberans, this section had three other new species, P. angulata Roezl, P. exserta Roezl, and P. heteromorpha Roezl, and also P. rudis Endl. (a misapplication of that name). Gordon (Sup. Gord. Pinetum 70, 1862; Pinetum Ed. 2, 319, 1875) ex- amined Roezl’s specimens and accepted P. protuberans Roezl, giving the other species of Roezl’s Sect. VIII as synonyms under P. protuberans. Carriere (7bid.) also accepted P. protuberans Roezl, quoted the original description, and published a longer description of Roezl’s dry specimens and living young plants. He reduced two species to varieties | P. protuber- ans angulata and P. protuberans exserta|, and retained P. heteromorpha Roezl as a species. Parlatore [in DC., Prodr. 16(2):401—402. 1868] re- duced P. protuberans and P. heteromorpha to synonymy under P. pseudo- strobus Lindl. and placed P. angulata and P. exserta under P. montezumae Lamb. In a summary of 82 new species of Pinus in Roezl’s catalog of 1857, Shaw (1909, p. 3 and table) cited the above and other references, noted that many of Roezl’s specimens had been lost, and concluded that there was not a single valid species among the six or seven pines represented. The epithets P. protuberans Roezl and P. pseudostrobus forma pro- tuberans apparently refer to the same entity. The cone scale of P. pro- tuberans has a protuberance, while that of P. oaxacana has a longer, more prominent projection. Therefore, because of the morphological and geo- graphical differences, this pine from Oaxaca is not referable to any of Roezl’s species. Pinus oaxacana differs from typical P. pseudostrobus by the prominent projections of the apophyses of the cone scales. In the general part of his Los Pinos Mexicanos (Martinez, 1948, p.37), Martinez showed drawings of the cone scales of his variety oaxacana and of P. coultert D. Don as examples of protuberant apophyses. When I collected oleoresin of Oaxaca pine in 1951 in Chiapas, Mexico, I found typical P. pseudostrobus and P. oaxacana growing together and I was much impressed by the difference in the cones. The heavy cones of Oaxaca pine resembled the cones of Shaw’s group Macrocarpae (1914) more than the cones of typical P. pseudostrobus. 1958] MIROV: PINUS 149 Shaw’s group Macrocar pae consists of Pinus torrevana Parry, P. coul- teri D. Don and P. sabiniana Dougl. In my opinion P. jeffreyi Murr. also is closely related to this group since, among other considerations, it crosses naturally with P. coulteri. Besides the morphological and genetic affinities of these pines, they all have common biochemical characters. Their tur- pentines all contain aliphatic hydrocarbons, either normal heptane C7H1¢ or normal undecane C,,H», or both (Table 1). Turpentine of Pinus pseudostrobus was analyzed by Iriarte (1946), and was found to contain over 90 percent of d-a-pinene and a small quan- tity of an unidentified sesquiterpene. Turpentine of P. oaxacana has an entirely different composition. It contains: n-heptane, 21 percent; d, dl- a-pinene, 51 percent; 1, dl-limonene, 15—16 percent; n-undecane, 1.3 per- cent; and a sesquiterpene, longifolene, 7.5 percent (Iloff and Mirov, 1953). Chemically P. oaxacana has much more in common with the pines of the group Macrocar pae, including P. jeffreyi, than with P. pseudostro- bus (Table 1). TABLE 1. OCCURRENCE OF SOME CHEMICAL SUBSTANCES IN THE TURPENTINES OF SEVERAL PINES oO = x o o ao} o q a a fs aie iee foe” Won ee Se 8 ~ CB) o {o) SQ = Ou. (a = aS) a S cS o om fo) i=) oD oO i a ee = qt 84s. Ge” Si = b = = = rs oO P. jeffreyi + fe P. sabiniana + Ae a — a8 P. coulteri +. + ne + + + P. torreyana ~ eee + + + + ae P. oaxacana + a + + _ a P. pseudostrobus — + 52) sets ? The proposal to elevate Oaxaca pine to a specific rank is based on con- sideration of both morphological and biochemical characteristics of the pine. The use of biochemical characters for taxonomic purposes is gaining more and more ground among taxonomists, especially in view of the bril- liant research on Australian trees by Penfold (1935) and his co-workers. Among the pines, Pinus je ffrevi is an outstanding example of the validity of biochemical characters in taxonomy. This pine has been considered by some botanists for a long time as a variety of P. ponderosa (Shaw, 1914). Lately, however (and at least, partly because of profound chemical differ- ences of the two pines) P. jeffreyi has been reinstated to its original status as a valid species. Pinus oaxacana apparently crosses naturally with P. pseudostrobus. It is entirely possible that great variability in the structure of the cone scales within the P. pseudostrobus complex has been caused by hybridization 150 MADRONO [Vol. 14 between these two pines. Pinus oaxacana also probably crosses with some varieties of P. montezumae, but consideration of such behavior is beyond the scope of this paper. Institute of Forest Genetics, California Forest and Range Experiment Station, U.S. Department of Agriculture, Forest Service, Berkeley, California. REFERENCES Ivorr, P.M. J., and N. T. Mirov. 1953. Composition of turpentines of pines. XVI. A report on Pinus oocarpa, and P. pseudostrobus var. caxacana from Chiapas and P. cooperi from Durango. Jour. Am. Pharm. Assn., Sci. Ed. 42:46-49. TRIARTE, JOSE. 1946. Estudio de los aguarrases mexicanos. Quimica (Mexico) 4:117- 119. Loock, E. E. M. 1951. The pines of Mexico and British Honduras. Union South Africa Dept. Forestry Bull. 35, 244 pp., illus. MartinEez, MAxImINno. 1945. Las pinaceas mexicanas. Vol. 1, 345 pp., illus. Mexico. . 1948. Los pinos mexicanos. Ed. 2, 361 pp., illus. Edicionas Botas. Mexico. Mirov, N. T. 1954. Composition of turpentines of Mexican pines. Unasylva 8:167- 1/3. PENFOLD, A. R. 1935. The physiological forms of the eucalypts, as determined by the chemical composition of the essential oils and their influence on the botanical nomenclature. Australian Jour. Pharm. 16:168-171. SHAW, GEORGE RUSSELL. 1906. The pines of Mexico. Publ. Arnold Arboretum No. 1, 29 pp., illus. . 1914. The genus Pinus. Publ. Arnold Arboretum No. 5, 96 pp., illus. CHROMOSOME COUNTS IN SECTION ERYTHRANTHE OF THE GENUS MIMULUS (SCROPHULARIACEAE)! RoBERT K. VICKERY, JR., BARID B. MUKHERJEE, AND DELBERT WIENS Brozek (1932) of Charles University, Prague, has determined the chromosome numbers of three horticultural plants of Mimulus cardinalis to be 2n = 16. These counts were made in connection with his investiga- tion of the genetics of flower color in this species. The senior author also has carried on work on the inheritance of flower color in the M. cardinalis complex (Vickery and Olson, 1956). In addition he is undertaking a bio- systematic study of the group. These investigations have necessitated a survey of the chromosome numbers of both the horticultural populations and the cultures of the wild races being used in these two studies. Her- barium specimens of all the cultures counted are deposited in the Garrett Herbarium of the University of Utah under the culture numbers given in Table 1. 1 The work was supported by grants from the National Science Foundation and the University of Utah Research Fund. The authors wish to thank Drs. W. W. Newby and C. M. Woolf for their helpful criticisms of the manuscript. 1958] VICKERY, MUKHERJEE, WIENS: MIMULUS 151 TABLE 1. CHROMOSOME COUNTS IN MIMULUS, SECTION ERYTHRANTHE A. CULTURES OF NATIVE SPECIES n=8 M. lewisit Pursh Porcupine Flat, Mariposa County, California, altitude 8000 feet, Sept. 17, 1948, Hiesey (5032). Above Alta, Salt Lake County, Utah, altitude 8700 feet, Vickery 207 (5875). n=8 M. verbenaceous Greene Narrows Trail, Zion National Park, Washington County, Utah, altitude 4400 feet, April, 1956, Wzens (5264). Bright Angel Creek, near Phantom Ranch, Grand Canyon National Park, Coconino County, Arizona, altitude 2000 feet, Nov. 4, 1954, Jackson (5924). n=8 WM. cardinalis Dougl. Beaver Creek, Siskiyou County, California, altitude 1800 feet, Hiesey 555 (5031). B. CULTIVATED CoLor FormMs n=8 M. cardinalis Dougl. 5077. Royal Botanic Garden, Edinburgh, Scotland. 5078 Regional Park Botanic Garden, Berkeley, California. 5308 Botanic Garden, Halle, Germany. 5309 Botanic Garden, Basel, Switzerland. 5310 Botanic Garden, St. Gallen, Switzerland. 5311 University Botanic Garden, Liége, Belgium. 5312 Botanic Garden, Antwerp, Belgium. 5313 Botanic Garden, Wageningen, Netherlands. 5315 University Botanic Garden, Brno, Czechoslovakia. 5316 Botanic Garden, Cluj, Roumania. 5318 Botanic Garden, Copenhagen, Denmark. Brozek made his counts from the tips of adventitious roots, fixing them in Navashin’s fluid and staining them with either Heidenhain’s iron haematoxylin or Cajal’s magenta and picro-indio-carmine stain (Brozek, 1932). We obtained good results with the method previously described (Mukherjee, Wiens, and Vickery, 1957) which employs fixation of the buds in acetic or propionic acid followed by squashing and staining in either aceto-carmine or propio-carmine. The sources of the horticultural color forms and of the wild cultures which we have studied are given in Table 1. The chromosome number appears to be n = 8 in all cases (fig. 1) although the preparations of M. verbenaceous (5264) showed occasional cells with apparently 9 or 10 chromosomes instead of the usual 8. The chromosomes of the 5077 cul- ture of M. cardinalis were noticeably larger than the average for other races of the species (fig. 1). Those of the 5875 culture of M. lewisii were appreciably smaller than those of any other culture of the complex which we have examined. These size differences in the chromosomes suggest that there may be structural differences which in turn may account for the bar- riers to hybridization which we have observed when crossing M. cardinalis and M. lewisii (Vickery, 1956). Initially, all crosses produce vigorous F 152 MADRONO [Vol. 14 CI EI@Y SOT 5078 oe 5S )0)9 5311 Sige & 5312 63s ope ais 5316 OC a 5318 562725 5924 5 310 or O 10 20 micra Fic. 1. Meiotic chromosomes of pollen mother cells of Mimulus, * 750. Chromo- some numbers of all the cultures are n = 8. Camera lucida drawings were made in all cases. The pollen mother cells of cultures 5077, 5264, 5308, 5309, 5318, and 5875 are in first metaphase whereas all the others are in second metaphase, but the two figures of the second metaphase were drawn only if the counts were clear in both nuclei. populations but frequently the F. and Fs generations are marked by decreased fertility. The crosses between M. verbenaceous and M. cardt- nalis and between M. verbenaceous and M. lewisiu yield vigorous, fertile F, hybrids, also. Combinations of M. cardinalis or M. lewis with species of section Paradanthus consistently failed (Vickery, 1956). Mimulus moschatus 1958 | PENLAND: PENSTEMON 153 Doug., M. primuloides Benth., and M. bioletti Kastw. were used to repre- sent section Paradanthus. Six different reciprocal crosses were attempted using an average of five flowers each. A Paradanthus intrasectional com- bination of M. moschatus and M. floribundus Dougl. produced vigorous but sterile F, hybrids. We have obtained only one chromosome count for this section: M. moschatus, n = 16 (unpublished). On the basis of these genetic and cytological results we believe that the most natural taxonomic treatment is to group the three taxa, M. cardi- nalis, M. verbenaceous, and M. lewisii, in section Erythranthe. This treat- ment follows that of Pennell (1951) rather than that of Grant (1924) where MW. lewisti is placed in section Paradanthus. Department of Genetics and Cytology University of Utah Salt Lake City 12, Utah. LITERATURE CITED Brozex, A. 1932. Mendelian analysis of the “red-orange-yellow” group of flower- colors in Mimulus cardinalis hort. Preslia 11:1—10. GranT, A. L. 1924. A monograph of the genus Mimulus. Ann. Mo. Bot. Gar. 11:99- 388. MUKHERJEE, B. B., D. Wiens, and R. K. VICKERY, JR. 1957. Chromosome counts in section Simiolus of the genus Mimulus (Scrophulariaceae) , II. Madrono 14:128- 131. PENNELL, F.W., 1951 zm Illustrated Flora of the Pacific States by Leroy Abrams. Stanford Univ. Press. Vol. II], pp. 688-731. VickERY, R. K., Jr. 1956. Data on intersectional hybridizations in the genus Mimu- lus (Scrophulariaceae). Proc. Utah Acad. Sci., Arts, and Letters 33:65-71. , and R. L. Orson. 1956. Flower color inheritance in the Mimulus cardinalis complex. Jour. Heredity 37:195-199. TWO NEW SPECIES OF PENSTEMON IN COLORADO C. WILLIAM T. PENLAND Since the treatment of Penstemon for Harrington’s ‘‘Manual of the Plants of Colorado” was prepared, additional collections and study have made it necessary to recognize the following two new species for the state. Penstemon harringtonii sp. nov. Herba perennis, glabra, 3— dm. 7 alta; caulibus erectis, simplicibus, glaucis; foltis glaucis, integris, crassis, obtusis vel acutis, mucronatis, basalibus spathulatis vel oblanceolatis, 5—7 cm. longis, 1.5—2.5 cm. latis, caulinis ceteris parvioribus, sessilibus, obo- vatis, elliptico-ovatis vel cordato-amplexicaulibus, ad inflorescentiam versus gradatim reductis; thyrso cylindraceo, angusto, interrupto, 5—10- fasciculato; calyce 5-9 mm. longo, lobis ovato-lanceolatis, acutis vel acuminatis, scarioso-marginatis; corolla 18-24 mm. longa, coerulea (vel rosea), bilabiata, fauce ampliata; staminibus didynamis, inferioribus 154 MADRONO LVol. 14 corollae faucem multo excedentibus; antheris 2.5-3 mm. longis, glabris, sagittatis; loculis antherarum de apice ad basim dehiscentibus, conflu- entibus, haud explanatis; filamento sterili dilatato (usque ad 1—-1.5 mm. latitudine), pilis 1-2 mm. longis, dense aurato-barbato; capsulis ignotis.! Glabrous perennial, 3—7 dm. tall; stems one to few, simple, erect, glau- cous below, often purplish, especially above; leaves entire, thickish, glau- cous, mucronate, the basal 1.5—2.5 cm. wide, 5—7 cm. long, spatulate to oblanceolate, obtuse to acute, the lower cauline 1—2 cm. wide, 2—5 cm. long, obovate, elliptic or ovate, acute, the upper cauline gradually reduced, ovate to cordate-clasping; inflorescence cylindrical, of 5-10 rather lax- flowered fascicles, one-third to one-half or more the height of stem; peduncles surpassing bracts, especially above; bracts all acute, scarious- margined, the lower broadly sessile-clasping, the upper reduced and incon- spicuous, longer than wide; calyx 5—9 mm. long, the lobes ovate- to lance- acuminate, striate, narrowly scarious-margined; corolla 18—24 mm. long, pale to deep coerulean blue, often lilac-tinged on tube, or strongly pinkish-lilac throughout, the throat ampliate-funnelform, rather distinctly bilabiate, the lower lobes divergent but scarcely reflexed, the upper arched-erect ; stamens conspicuously didynamous, the upper pair attached to corolla at its very base, the lower pair attached along corolla tube for about one-half their lengths, the free part then curved ventrally in the throat and well-exserted outward and upward to bring anthers opposite middle of orifice of corolla; anthers 2.5—3 mm. long, glabrous, sagittate, attached to filament at their middle or below, the connective wide, the sacs conjoined for one-half or more of their lengths, dehiscing throughout but not explanate, remaining parallel but curved; staminode glabrous to middle, then abruptly densely bearded with golden yellow hairs, those on the upper surface 1—2 mm. long, those on the lower surface prominent, but shorter and less dense, the staminode widened distally to 1-1.5 mm., rounded at tip and curved ventrally, usually a little exserted from orifice of corolla tube; mature capsules not seen. Type. Abundant in sagebrush, altitude about 8000 feet, 3-5 miles northwest of Green Mountain Dam, Grand County, Colorado, 19 June 1952, Penland 4296 (COCO; isotypes, COLO, CS, GH, NY, RM, UC, US). Other known collections. CoLorApo. Eagle County: rather dry slope 5 miles east of Wolcott, elevation 7200 feet, 7 June 1951, H. D. Harrington 4935 (COCO, CS). Routt County: pinyon, cedar, sage association, elevation 7500 feet, 2 miles north of McCoy, Highway 131, 29 June 1951, M. & C. Norton sn. (COCO, CS). Penstemon harringtonii is readily recognized both in the field and in the herbarium by its two well-exserted stamens (fig. 1), a character that appears remarkably constant. Judged on the basis of its anthers it is most closely related to P. cyvathophorus Rydberg, and it is therefore referred to the section Coerulei (as defined by Pennell, 1935). It differs 1 The assistance of Mr. Robert M. Ormes in preparation of Latin diagnoses is gratefully acknowledged. 1958 | PENLAND: PENSTEMON 155 Fic. 1. Penstemon harringtonii. Habit, ca. X ¥3; anthers, x 5; staminode, ca. x 1%; flower, x 1. from P. cyathophorus in its larger flowers (18-24 mm. long as compared to 9-15 mm.), in having only two instead of its four stamens strikingly exserted, in its longer anthers (2.5—3 mm. as compared with 2 mm.), and in its strongly reduced bracts, which are mostly longer than broad (the reverse is true of P. cyathophorus). In general appearance, however, P. harringtonii is more like P. osterhoutii Pennell (same section), to which it 156 MADRONO [Vol. 14 was first referred, than it is like P. cyathophorus. Although the flowers average 14-20 mm. long in P. osterhoutiu, thus approaching those of P. harringtonu, the anther sacs are widely divaricate and not at all sagittate as in P. harringtonu, and the stamens are not or scarcely exserted in P. osterhoutiu. These closely related species may be keyed as follows: Anthers sagittate with parallel sacs; 2 or all 4 stamens well-exserted. Flowers 18-24 mm. long; only 2 stamens exserted; anthers 2.5—-3 mm. long; bracts mostly Jonger. Chan sbr@ad tise re eee rc eee P. harringtoni Flowers 9-15 mm. long; all 4 stamens exserted; anthers 2 mm. long; bracts mostly broader-than long tec 50a ees Ds eee ee ee P. cyathophorus Anthers with sacs divaricate; stamens not or scarcely exserted.............. P. osterhoutu The three cited collections of P. harringtonii represent a distribution of this plant on both sides of the Colorado River drainage system, and, to the south, on both sides of the Gore Range (Eagle and Blue rivers). The known range of P. cyathophorus (Grand and Jackson counties in Colo- rado, and in adjacent southern Wyoming) is to the north of that of P. harringtoniu (Eagle, Grand, and Routt counties, Colorado), although at one place in Grand County the two species occur within 2—3 miles of each other. Penstemon osterhouti, the other species closely related to P. har- ringtonti, is common to the west of P. karringtonii in Eagle and Garfield counties, but the ranges of the two do not overlap. Moreover, P. karring- tonii occurs at elevations of 7200-8000 feet, while P. osterhoutu has not been collected above an elevation of 6600 feet. The unusually interesting Penstemon acaulis Wms. was first collected in Sweetwater County of southwestern Wyoming. Later it was found in adjacent Daggett County in Utah. So far as is known it has not been reported from south of the Uinta Mountains in Utah, or from the north- western corner of Colorado. In 1951 the writer collected a Penstemon which was at first presumed to be only a more robust form of P. acaulis. It was found a few miles east of Elk Springs, south of the Yampa River in Moffat County, Colorado, a locality approximately eighty airline miles southeast of the nearest known station for P. acaulis. In order to secure better diagnostic material this area was again visited in 1952 and 1953. In those years two additional stations for the plant were discovered; both of these were north of the Yampa River, between Greystone and the Little Snake River. The collections and all material observed in the field are quite uniform in aspect, and continued study has led to the proposal of the following new species, whose closest relative is clearly P. acaulis. Penstemon yampaensis sp. nov. Herba caespitosa perennis, 3 cm. vel minus alta; rhizomatibus ramosis; foliis 15-30 mm. longis, 2-4 (—5) mm. latis, oblanceolatis, acutis, cinereis, spiculato- vel papillato-pube- scentibus, plus minusve viscidis; calyce 5-9 mm. longo, viscido-pube- scente, lobis acuminatis, inferne scarioso-marginatis; floribus 2—4 in ramo terminali; corolla 15-18 mm. longa, pallide roseo-purpurea (vel demum coerulea), extus glanduloso-pubescente, basi faucis rotundatae haud pli- 1958] PENLAND: PENSTEMON Si, Fic. 2. Penstemon yampaensis (Penland 4415). About X 1. catae aliquantulo ventricosae aureo-barbata, lobis posterioribus quam ceteris brevioribus; staminibus inclusis; loculis antherarum 1—1.5 mm. longis, glabris, confluentibus, haud late explanatis; filamento sterili paulo exserto, insuper aureo-barbato; capsulis fere globosis, 4 mm. longis, glabris; seminibus 2-4, nigris, lunatis, rugosis. Caespitose, essentially acaulescent perennial, spreading from branching rootstocks and in the open forming loose mats up to one foot in diameter, 3 cm. or less in height; leaves cineraceous, 15-25 (—30) mm. long, 2—4 (—5) mm. wide, oblanceolate, acute but not mucronate, scabro-pubescent with low, blunt, papilliform hairs (common on upper portion) or longer, straight or somewhat recurved, spiculate hairs (common toward base and petiole), the foliar hairs not gland-tipped but evidently viscid; mid-rib and two or more lateral veins prominent on dried specimens; flowers usually 4 (2-6) on each of the very short ultimate branches, often exceeded by the foliage; calyx 5—9 mm. long, the lobes long-acuminate, rather densely viscid-pubescent, narrowly scarious-margined below; cor- olla 15-18 mm. long, lilac, or with strong bluish tinge when older (?), on dried specimens usually blue, glandular-pubescent externally, golden- (to whitish-) bearded in throat below, the two upper lobes arched-erect, but little shorter than the less divergent three lower lobes (which may reach a length of 5 mm.), the throat moderately ampliate above short tube, some- 158 MADRONO [Vol. 14 what ventricose, not at all plicate; stamens with longer pair reaching limb of corolla, the anther sacs 1-1.5 mm. long, ovate to oblong-ovate, glabrous, minutely denticulate along suture, divaricate, dehiscent confluently, not explanate; the staminode exserted from throat, bearded all along its dorsal surface with golden hairs, a little widened distally (to as much as 0.75 mm.), the apex emarginate; capsules nearly globose, about 4 mm. long, glabrous at maturity; seeds lunate, 2-3 mm. long, black, rugose and puncticulate, 2—4 per capsule. Type. Sandy, calcareous knoll, scattered grass and sage; 5.5—6 miles east of Elk Springs, Moffat County, Colorado, 9 June 1953, Penland 4415 (COCO; isotypes, COLO, GH, NY, RM, UC, US). All other collec- tions are from Moffat County, Colorado: type locality, 26 June 1951, Penland 4236 (topotype, COCO); type locality, 20 June 1952, Penland 4311 (topotypes, COCO, COLO, CS, GH, NY, RM, UC, US); 20 miles south of Vermillion Creek Bridge, 22 June 1952, Penland 4326 (COCO, UC, US); 3-4 miles south of Greystone, 10 June 1953, Penland 4426 (COCO, NY, RM). A key serving to separate P. acaulis (fig. 3) and P. vampaensis (fig. 2) follows: Leaves essentially linear, 1-1.5 mm. wide, 10-20 mm. long; flowers 1 (—2) per ulti- mate floriferous shoot; corolla 14-16 mm. long, blue.........0000...0220000.02.- P. acaulis Leaves oblanceolate, 2-4 (—5) mm. wide, 15-25 (—30) mm. long; flowers usually 4 (-6) per ultimate floriferous shoot; corolla 15-18 mm. long, lilac to bluish- Va ie eee oie ie ee eer ee eee eee ee ee P. yam paensis The foliar pubescence of these two species is essentially the same, though more of the spiculate hairs seem to be present in P. yampaensis, especially toward the leaf base. Penstemon acaulis was assigned to the section Caespitost by Williams in his original description of this species (1934). Keck (1937) also assigned it to this group but changed the rank of the group to a subsection of the section Evicopsis. Keck pointed out, however, that in P. acaulis, “The character of its pubescence is unique in this group of species.”’ There are two probable reasons for this unique- ness: 1) the papillate hairs, besides being short and blunt, are very broad at the base (commonly 44—73 microns, as compared to 14—25 microns for hairs of P. caespitosus Nutt., P. crandallii A. Nels., P. retrorsus Payson and P. abietinus Pennell, all of which Keck also places in the Caespitost ; 2) the external walls of these hairs in P. acaulis and P. yampaensts are smooth or nearly so, while those of the four species named just above are denticulate (as observed in boiled or cleared material, at 100 * magni- fication). In this connection it is of interest that the hairs of P. moffatu Eastw. (section Aurator) are also blunt and smooth-walled. Also, the apex of some leaves of P. moffatii may even show the papillate condition of P. acaulis, but the longer, blunt, recurved type of hair is typical for the species. The midrib of leaves of dried specimens of both P. acaulis and P. yam- paensts is prominent nearly throughout; in addition, P. yampaensis shows 1958 | PENLAND: PENSTEMON 159 Fic. 3. Penstemon acaulis (Penland 4419). About 1. two or more prominent lateral veins. This type of veining is certainly rare in the Caespitosi. Although the midrib may be prominent, as in P. crandalliu, this condition is found only in the petiolar portion of the leaf. In fact, I have found a leaf situation similar to P. yam paensts in only one other case, a specimen from the Uinta Mountains of Utah which I have doubtfully referred to P. abietinus. Two other collections of P. abietinus show leaves agreeing with the other Caespitosz. (I have not studied leaves of P. thompsoniae (Gray) Rydb., however.) The midrib of P. vam paensts is not only prominent but is also relatively large (sometimes over 400 microns in diameter in cleared material, as compared with under 200 microns in P. caespitosus). Probably correlated with the presence of large lateral veins in P. yampaensis is the fact that a number of leaves have been found showing 1—4 small teeth near the apex. The non-plicate character of the corolla throat certainly poses a ques- tion as to the assignment of P. acaulis and P. yampaensis to the subsection Caespitosi, where their caespitose nature would apparently place them. In all the other six species of the subsection (Keck, loc. cit.), the 2-ridged and commonly laterally flattened, plicate throat is a constant and diag- nostic feature. Nor do P. acaulis and P. yampaensis fit into the other two subsections of Evicopsis (viz., the Linarioides and the Laricifola). It seems to me that their affinity is rather with the section Aurator, through such species as P. nanus Keck, P. dolius Jones, P. pumilus Nutt., and P. 160 MADRONO [Vol. 14 moffatu Eastw. Perhaps the least that can be said for them is that they help to emphasize more strongly the relation between the sections Aurator and Evicopsis, a point which was elucidated by Keck. Colorado College, Colorado Springs, Colorado LITERATURE CITED Keck, D. D. 1937. Studies in Penstemon IV. The Section Ericopsis. Bull. Torrey Club 64:357-381. PENNELL, F.W. 1935. The Scrophulariaceae of eastern temperate North America. Monog. Acad. Phila. 1:i-xiv, 1-650. WitiiAmMs, L. 1934. Field and herbarium studies, III. Ann. Mo. Bot. Gard. 21:345- 346. PECULIARITIES OF THE COLUMBIA RIVER GORGE FLORA LERoy E. DETLING The gorge cut through the Cascade Range by the Columbia River as it flows westward between Washington and Oregon has long been recognized as a place of peculiar botanical interest. If one studies the distribution of the plant species found there, he is soon impressed by the large number that are either endemic to the area or occur as isolated populations sig- nificantly removed from their normal range. My interest in the history of the various elements of the Pacific Northwest flora led me to speculate upon the meaning of these peculiar distributional patterns. It seemed pos- sible that an analysis of them might furnish clues to the role the Columbia Gorge has played in the migrations of vegetation in the past. Its physiog- raphy and geographical relationship to adjacent vegetation areas would in themselves lead one to believe that it might have served as a pathway for the migration of many types of organisms. With this in mind, I have spent considerable time during the past few summers studying the local distribution of those plant species occurring within the Gorge. This field work has subsequently been supplemented with herbarium studies on their wider distribution. The present paper is the result of this work. The species listed herein by no means constitute a complete flora of the Columbia Gorge, although I have tried to make the lists as inclusive as possible. The collections of other botanists who have collected extensively here have also been studied, particularly those of Howell, Gorman, Hen- derson, Suksdorf, and Sheldon. Nevertheless, further search would un- doubtedly reveal more species which might be included in the roster. How- ever, I am confident that the list is sufficiently extensive to give us a good general picture of the significant features of the Gorge flora as a whole, its distribution in the various habitats, and its relation to outside popu- lations, which was the chief purpose of this investigation. Any additions to the number of species will not increase the accuracy of a study made from this particular viewpoint. 1958] DETLING: COLUMBIA RIVER GORGE 161 PHYSIOGRAPHY AND CLIMATE OF THE GORGE The Cascade Range has been formed by a combination of gradual uplift since the Miocene epoch and of volcanic action largely during the Pliocene and Pleistocene. During this process the Columbia River, without materi- ally altering its course, kept pace with the change by cutting its bed deeper and deeper, thus forming the great gash through the mountains which we know as the Columbia River Gap, or in a more restricted sense the Co- lumbia River Gorge. Lateral erosion has been much slower than that car- ried on by the main stream, resulting in a deep canyon with precipitous walls, extending for some thirty-five miles. The small streams which flow into the river have cut their beds back into this canyon wall sometimes as much as a half mile, forming deep, narrow, lateral gorges into which the streams frequently plunge as waterfalls from a hundred to several hun- dred feet high. Erosion has been less effective on the south wall of the canyon than on the north, and it is here that the most spectacular cliffs are to be seen. One of the most striking of these is the north face of St. Peter’s Dome, which rises abruptly for about twenty-five hundred feet. Some of the cliffs have their bases almost at the river’s edge, while others rise as much as a half mile or more back from the water. On the north wall of the Gorge the bases of the sheer drops are usually farther back from the river as well as higher up. A noticeable feature of both walls of the canyon is the frequent stair-step arrangement, apparently due to differences in hardness of the various strata of rock. It is of interest, but perhaps of no further consequence to us, that these steps, and therefore the shelves upon which much of the vegetation is growing, are roughly about six hundred feet one above the other. Thus there are a number of places where a shelf occurs about six hundred feet above the river, with other shelves above this at elevations of approximately twelve hundred and eighteen hundred feet. At most places the eighteen-hundred-foot shelf marks the top of the most abrupt cliffs, and the rise above that, while steep, is nevertheless a distinct slope, usually forested. For an area so limited in extent the Columbia Gorge has an extremely varied climate. This is due in part to the depth of the canyon itself, in part to the influence of two entirely different climates at either end of the canyon, and in part to the general east-west orientation of the gap. The curve of annual precipitation through the Gorge follows rather closely that of any transect extending from the Willamette Valley east- ward across the Cascade Range to the central Oregon plateau. Although the level of the river bed rises comparatively little as one follows it east- ward through the canyon, yet the elevation of the mountains that press in on either side is such as to raise the moisture-laden air masses from the Pacific Ocean sufficiently to cause a heavy precipitation over the central sector of the Gorge. The mean annual precipitation at Vancouver of 37 inches is typical for points along the lower Willamette and Columbia riv- ers. However, at Mount Pleasant, situated close to the western portal of 162 MADRONO [Vol. 14 the Gorge, this figure has risen to 57 inches. Rainfall increases eastward to the vicinity of Cascade Locks where an average of approximately 75 inches is measured. Upstream from here there is a sharp drop in the annual mean. At Hood River and White Salmon about 30 inches are re- corded, and at Lyle 25 inches. Recorded data on temperatures in the Gorge do not tell us a great deal. Between Mount Pleasant and Hood River there is a gradual drop of 4.7°F. in the mean January temperature. In the same distance the mean July temperature rises 2.3°F. One striking feature of the Gorge climate does not appear in the rec- ords. When atmospheric pressure differentials are favorable, masses of air move from eastern Oregon and Washington westward through the Gorge. In winter this situation frequently results in cold winds which at times reach very high velocities. When this occurs in conjunction with rainfall the result is often a very serious “‘silver thaw,” 1.e., a rain which freezes as it touches the ground, or trees and other vegetation. These freezing rains sometimes last for several days, and the breakage of trees and de- struction of other types of plant life may be quite serious. A parallel situ- ation frequently occurs during the summer when hot, dry east winds blow down through the gap, sometimes for several days at a time. The effect that such winds might have upon the vegetation is obvious. As might be expected in a canyon with an east-west orientation, the local climate of the north wall is quite different from that of the south wall. This is clearly reflected in the difference in the general vegetation of the two sides. The flora of the dry, warm area of central Oregon and Washington extends much farther westward on the well-insolated south- facing slopes than on the north-facing wall, which is in deep shadow much of the time. One the other hand, the Douglas fir forest and its associates follow the cool north-facing slopes entirely through the Gorge to the Hood River Valley. The significance of the Columbia Gap from the standpoint of such a climatic and vegetational barrier as the Cascade Range may be realized when we recall that it is the only point between the Fraser River in British Columbia and the Klamath River in northern California where the axis of the range has been cut through in such a way as to interrupt the con- tinuity of all the life zones above the Transition. In this case the cutting has extended almost to sea level. FLORA OF THE GORGE To understand fully the flora of the Gorge and the implications of the distribution patterns of many of its species we must look first at the type of vegetation now existing in the adjacent lowlands at either end. To the west the Gorge opens out upon what I have elsewhere called the Puget Area (Detling, 1948). This vegetation area occupies the broad val- ley between the Cascades and the Coast Range, extending from the south- DETLING: COLUMBIA RIVER GORGE 163 1958] a a SSS SS SS = eee \) _ —_—_—_—_—— J > Wn e SSN Gf ; Li“ . \W \v' ia = rip SS = Zz Pe A a y" J Q Se L ) & > wy ly, all, » — EZ Tye = ot YSM PIC é %, = > > LJ “My 0 = WV: 1f S Me, AY =~ ce = x Ney > ww }) a fi = Wee "EWS Sy x = = S ©O ° = > ay. = ~ 7, = =A My % <@W/D ' G = wv J ~ % Mh ie MO ; . We Oo q Mm Ss gee Le = Bie wt 7/) : wis uy P ES My , z = 4 By. z \! w aN 4 ZZ ly ty Ss Xo: ie ; = W S a s, — = . “7 4 ‘iS Zz Wy 2 Zs z a =S ee Nj" wy oo < 2 “1 lltZ, ss Ch ee O SADDLE \g =, om Moy 24S @ = Bigs mG DUS a = »< Mi Zs, << ~ = = = SS zg . L PAY) n V. 6) pee = = ca> “ Tres => x * G6 OW, E 2 VY — A g SS PS LK RK. S Mi S = ~S > 5 = we = ; S = =» ZI = Le = = : Vv Be = es = Bs = EQ - ZI is ly ZO = Oo /I\\ ge Sq 0 — = = S es = wy, 32 S 2 ae = , Be Oo ee _ = wey S ~S Wy - = SS wy MI, =z = . iy, = Z, ( >) c bs, AIGY = = Z, Zz = oa < oe Z = ly Yan a - \ \ = “IW — “ayys S Z AR E 7 . = Zs, = Z, \( Ss Ss Wi) - SI diy Z ANN - = Sees) TIN canes Ss A mg , a TT \Y \ ce = yy) Ww = — - = _ iN = , = owl = : N af wo ER waE” G oO Tw - = “ ; (Ny A ’ we = “ Uf - 5 ._ = INS . ~ = as =, . ws ¢ \ > trq\\v = " av «WW Ie Sc = = N >= = = AT) = 3B : ” a SW) ~~ = ow Q > a a => x Z NW N Ring es Sy S “yw 80 MILES 1A i < \ 7 aw > KO = \ = Nw = xs § oO St Re > Aw 2 WN 2 g , 5 “qe NNW | ae = Ss = {\M \\ WZ) = wD a et hast = . 7, As ae Se ea ee > N —— ee oe an )\) Fic. 1. Main geographic features of the Pacific Northwest related to the Columbia Gorge flora. 164 MADRONO LVol. 14 ern limits of the Willamette watershed in Oregon northward to Puget Sound and southern British Columbia. From a physiographic standpoint it coincides more or less with what the geomorphologists call the Puget Trough. The vegetation of the area is basically a Douglas fir forest, with an intermingling of oak-madrone woodland and grassland at lower levels and on the valley floor. Annual precipitation is high and neither summer nor winter temperatures are extreme. These conditions result in a dense forest cover, especially in the foothills, with an understory of vine maple (Acer circinatum), salal (Gaultheria shallon) , low Oregon grape (Maho- nia nervosa), and red huckleberry (Vaccinium parvifolium), along witha wealth of herbaceous species. Eastward the Gorge widens out and merges with the Columbia Area, again to use a name adopted by the author (op. cit.). This area occupies the valley of the Columbia River in north-central Oregon and south- central Washington, and such tributary valleys as those of the Yakima, John Day, and lower Deschutes rivers. The Hood River Valley comprises its westernmost extension. Annual precipitation is light throughout the area (13 inches at The Dalles). Winter temperatures are frequently very low, while the summers are hot. Apparently the region was originally a grassland, but there has been an invasion of sagebrush (Artemisia species) where the land has been overgrazed. Stands of ponderosa pine (Pinus ponderosa) are not uncommon on the hills where soil, moisture, and tem- perature conditions are favorable. 1. SPECIES OF WIDE DISTRIBUTION. For purposes of analysis I have found it convenient and useful to divide the Columbia Gorge species into groups according to their present general distribution. The first of these comprises a relatively large number of species (70, which is 34.0 per cent of the total of 206) which are widely distributed and occur at lower and middle elevations on both sides of the Cascade Range. From the stand- point of their origin this group can be divided again into two sub-groups —one whose members have probably come into our region from the north, the other evidently having originated to the south. In trying to determine where any given species has originated we may use several criteria. It is admitted that the use of any of these criteria alone would be unsound, but when they lend mutual support, and especially when applied to rela- tively recent migrations (from the standpoint of geological time), they probably present a fairly accurate idea of the direction of migration of most of the components of this flora. The criteria that seem most applica- ble to the problem at hand are: (a) The present maximum concentration of individuals of a species is likely to be somewhere near the area from which it has radiated in its migrations, and such concentration will prob- ably be found in environmental conditions similar to those under which it originated. (b) A species is more likely to have come from a region where obviously close relatives are now located, than from a region in which such relatives are absent. (c) When a species typically occurs as a member of a definite species association, the area of origin of whose other 1958] DETLING: COLUMBIA RIVER GORGE 165 members can be postulated, we may assume it is safe to assign to it that same origin. Judged on these bases the northern element of this first group consists of those species whose centers of distribution are to the north of the Columbia River; their present range frequently extends as far north as Alaska. At our latitude they are plants of shaded woodlands and stream banks, mostly associates of the coniferous forests. As might be expected from the dryness and high summer temperatures of the Columbia and Deschutes areas, the species tend to be absent from these two areas, but are common in the foothills and lower mountain slopes west of the Cas- cade crest and again in the Blue Mountains and in the Rocky Mountains of Idaho and Montana. It is highly probable that most of them migrated into our region by following the Cascade and Rocky Mountain ranges southward. The following species make up this sub-group: Adiantum pedatum L. Mimulus guttatus DC. var. aleuticum Rupr. Physocarpus capitatus (Pursh) Ktze. Aquilegia formosa Fisch. Poa gracillima Vasey Cerastium arvense L. Poa nervosa (Hook.) Vasey Chimaphila menziesii (R.Br.) Spreng. Potentilla glandulosa Lindl. Chimaphila umbellata (L.) Nutt. Pseudotsuga menziesii (Mirb.) Franco var. occidentalis (Rydb.) Blake Pteridium aquilinum (L.) Kuhn Circaea pacifica Asch. & Mag. var. pubescens Underw. Cirsium edule Nutt. Rosa gymnocarpa Nutt. Clintonia uniflora (Schult.) Kunth. Rosa nutkana Presl Cornus stolonifera Michx. Rubus parviflorus Nutt. Elymus glaucus Buckl. Sambucus glauca Nutt. Epilobium angustifolium L. Sedum spathulifolium Hook. Festuca occidentalis Hook. Selaginella douglasii (Hook. & Grev.) Festuca rubra L. Spring. Festuca subulata Trin. Sisyrinchium idahoense Bickn. Galium aparine L. Smilacina racemosa (L.) Desf. Geranium bicknellii Britt. Smilacina sessilifolia (Baker) Nutt. Heracleum lanatum Michx. Symphoricarpos albus (L.) Blake Heuchera micrantha Dougl. Thalictrum occidentale Gray var. pacifica R.B.L. Thuja plicata Donn. Linnaea borealis L. Tiarella unifoliata Hook. var. americana (Forbes) Rehd. Trientalis latifolia Hook. Lupinus polyphyllus Lindl. Trillium ovatum Pursh Lupinus rivularis Doug]. Viola glabella Nutt. Melica subulata (Griseb.) Scribn. The species comprising the southern element of this first group have their present centers of distribution to the south of the Columbia River. None extends as far north as Alaska, although some reach southern Brit- ish Columbia in the dry region east of the Cascades. West of the moun- tains they are most typically associated with the oak-madrone woodland, and when they occur at middle and higher elevations it is on exposed places with shallow soil, strong insolation, or other factors favoring a xeric flora. The members of this sub-group display less tendency to avoid the Columbia and Deschutes areas. They have probably arrived at our lati- 166 MADRONO [Vol. 14 tude by way of the system of valleys west of the Cascades or by way of the valleys and plateaus adjacent to the Great Basin east of this range. Adenocaulon bicolor Hook. Gilia capitata Hook. Allium acuminatum Hook. Holediscus discolor (Pursh) Maxim. Amelanchier florida Lindl. Koeleria cristata (L.) Pers. Apocynum pumilum (Gray) Greene Lilium columbianum Hans. Asarum caudatum Lindl. Lotus douglasii Greene Bromus vulgaris (Hook.) Shear Madia gracilis (Smith) Keck Castilleja hispida Benth. Microsteris gracilis (Dougl.) Greene var. hispida Pinus ponderosa Dougl. Crocidium multicaule Hook. Prunus demissa (Nutt.) Dietr. Cryptantha hendersonii (Nels.) Piper Rhamnus purshiana DC. Delphinium menziesii DC. Sedum douglasii Hook. Eriophyllum lanatum (Pursh) Forbes Trifolium microcephalum Pursh Erysimum capitatum (Dougl.) Greene Zygadenus venenosus Wats. 2. PuGET AREA ELEMENT. Another large group of species represented in the Gorge (54 in number and 26.2 per cent of the total) differs from the previously listed one in being restricted to the west side of the Cascade axis. They are, for the most part, common and widely distributed in the Puget Area and form a considerable part of its basic flora. Acer circinatum Pursh Penstemon ovatus Dougl. Acer macrophyllum Pursh Penstemon serrulatus Menz. Achlys triphylla (Smith) DC. Phacelia nemoralis Greene Alnus oregona Nutt. Philadelphus lewisii Pursh Anemone deltoidea Hook. var. gordonianus (Lindl.) Jeps. Arnica amplexicaulis Nutt. Plectritis congesta (Lindl.) DC . Aruncus silvester Kostel. Polypodium vulgare L. var. acuminatus (Dougl.) Jepson var. occidentale Hook. Asplenium trichomanes L. Polystichum munitum (Kaulf.) Presl Brodiaea coronaria (Salisb.) Engler Prunus emarginata (Dougl.) Walp. Campanula scouleri Hook. var. mollis (Dougl.) Brew. Cimicifuga elata Nutt. Pyrola bracteata Hook. Cornus nuttallii Aud. Rhododendron macrophyllum D. Don. Corydalis scouleri Hook. Rhus diversiloba T. & G. Corylus californica (A. DC.) Rose Ribes bracteosum Dougl. Delphinum oreganum How. Ribes sanguineum Pursh Dicentra formosa (Andr.) DC. Romanzoffia suksdorfil Greene Dryopteris arguta (Kaulf.) Watt. Rubus spectabilis Pursh Fragaria bracteata Hel. Sambucus callicarpa Greene Gaultheria shallon Pursh Stachys emersonii Piper Hydrophyllum tenuipes Hel. Struthiopteris spicant (L.) Weis. Iris tenax Dougl. Tellima grandiflora (Pursh) Dougl. Mahonia aquifolium (Pursh) Nutt. Tiarella trifoliata L. Mahonia nervosa (Pursh) Nutt. Tolmiea menziesii (Pursh) T. & G. Maianthemum bifolium DC. Tsuga heterophylla (Raf.) Sarg. var. kamtschaticum (Gmel.) Jeps. Vaccinium parvifolium Smith Melica harfordii Boland. Valeriana sitchensis Bong. Montia parvifolia (Moc.) Greene var. scouleri (Rydb.) Piper Oplopanax horridum (Sm.) Miq. Vancouveria hexandra (Hook.) Morr. & Oxalis oregana Nutt. Dene. Oxalis trilliifolia Hook. 1958] DETLING: COLUMBIA RIVER GORGE 167 3. CoLUMBIA AREA ELEMENT. These species (27 in number, 13.2 per cent of the total) are typical members of the plant associations found in the Columbia Area, at least in its western portion, i.e., near the upper end of the Gorge. Some of them range beyond the confines of the area, occur- ring at relatively low elevations and in situations ecologically similar to those in the Columbia Area. When these occur west of the Cascades it is normally in the Rogue River Valley and southward, but not in the Puget Area. *Anemone oregana Gray Arnica cordifolia Hook. Bromus tectorum L. Castilleja hispida Benth. var. acuta Penn. *Crepis barbigera Leib. Dicentra cucullaria (L.) Bernh. var. occidentalis (Rydb.) Peck *Dodecatheon poeticum Hend. Eriogonum sphaerocephalum Benth. var. tenue (Small) Piper Hackelia diffusa (Lehm.) Johnst. Helianthus cusickii Gray *Hydrophyllum capitatum Dougl. var. thompsonii (Peck) Const. Iliamna rivularis (Dougl.) Greene Lewisia rediviva Pursh *Lupinus leucopsis Agardh var. bingensis (Suks.) C.P.Sm. Microseris nutans (Geyer) Schultz-Bip. *Penstemon barrettae Gray Penstemon richardsonii Dougl. Penstemon subserratus Penn. Philadelphus lewisii Pursh var. lewisii Phlox speciosa Pursh Poa ampla Merr. Prunus emarginata (Dougl.) Walp. var. emarginata Senecio integerrimus Nutt. var. exaltatus (Nutt.) Cronq. Silene douglasii Hook. Spiraea lucida Dougl. Trifolium macrocephalum (Pursh) Poir. Luina nardosmia (Gray) Cronq. var. glabrata (Piper) Cronq. The six species preceded by an asterisk are of especial interest in that they occupy a very restricted region. They have their centers of distribu- tion near the east end of the Gorge and are strictly endemic to this part of the Columbia Area. In the Gorge they usually occur toward the east end and, as might be expected, in fairly dry situations. Since the Lupinus and Penstemon species referred to here do not actually occur west of Mosier, it is doubtful whether they should be considered in our discussion of the Gorge flora. I list them merely because they form part of a flora which does have a definite relationship to the Gorge. 4. ROGUE AREA ELEMENT. A significant element of the Columbia Gorge flora (17 species, or 8.3 per cent) is typically associated with the oak- madrone woodland of the Rogue Area of southwestern Oregon (op. cit.), its components having their centers of distribution either there or farther to the south. However, they occur again rather frequently in the Colum- bia Area, especially in that portion near the upper end of the Gorge. These species rarely occur otherwise east of the Cascades. In the Puget Area they commonly appear as components of some rather striking islands of xeric vegetation growing on exposed rocky points in the Douglas fir belt (Detling, 1953). Their peculiar distribution pattern suggests the prob- ability that during a warm, dry period they migrated northward from the 168 MADRONO [Vol. 14 Rogue into the Puget Area, then eastward through the Columbia Gorge into the Columbia Area. During a succeeding cooler and moister phase of the climatic cycle these species have been much restricted in their occur- rence in the Puget Area, and probably in the Gorge and in the Columbia Area as well. Leptotaenia dissecta Nutt. Lotus micranthus Benth. Lupinus bicolor Lindl. Balsamorhiza deltoidea Nutt. Collinsia grandiflora Dougl. Collinsia parviflora Dougl. Comandra umbellata (L.) Nutt. Lupinus laxiflorus Doug]. Dichelostemma pulchellum (Salisb.) Phacelia linearis (Pursh) Holz. Hel. Quercus garryana Dougl. Eriogonum compositum Dougl. Senecio harfordii Greenm. var. pilicaule St. J. & War. Triteleia grandiflora Lindl. Godetia amoena Lilja var. howellii (Wats.) Hoover Godetia quadrivulnera (Dougl.) Spach Viburnum ellipticum Hook. 5. BOREAL ELEMENT. A group of 31 species (15.1 per cent of the total) occurs normally at fairly high elevations in the Cascades, but is of interest here because its members are found rather abundantly at or near the bot- tom of the Gorge. Typically Canadian or Hudsonian Zone species, they usually range above 4000 feet. Some of them occur only in the Cascades, while others are found also in the mountain ranges both to the west and to the east of the Cascade axis; however, with only two or three excep- tions they do not descend to the valley and plateau levels on either side. In the Gorge these species are nearly always found below the 1600-foot level, i.e., they are limited largely to the steep bluffs and cool lateral can- yons, mostly on the south side of the river. In each case they are isolated from the main body of the population by a broad band of forested and less precipitous terrain in which I have so far found no record of their occurrence. Menziesia ferruginea Hook. var. glabella (Gray) Peck Mitella trifida Graham Penstemon nemorosus (Dougl.) Trautv. Penstemon rupicola (Piper) How. *Phlox diffusa Benth. *Acer glabrum Torr. subsp. douglasii (Hook.) Wesml. *Alnus sinuata (Regel) Rydb. *Antennaria racemosa Hook. Arabis furcata Wats. Arctostaphylos uva-ursi (L.) Spreng. Arnica discoidea Benth. var. eradiata (Gray) Cronq. *Campanula petiolata DC. Cornus canadensis L. Cryptogramma acrostichoides R. Br. Dodecatheon dentatum Hook. Dryopteris linnaeana C. Chr. Habenaria unalaschensis (Spreng.) Wats. Haplopappus hallii Gray Lewisia columbiana (How.) Robins. *Lomatium angustatum (C.& R.) St. John subsp. longistylis Wherry Polemonium carneum Gray *Polypodium vulgare L. var. columbianum Gilb. Populus tremuloides Michx. *Saxifraga bronchialis L. var. vespertina (Small) Rosend. *Saxifraga caespitosa L. Saxifraga rufidula (Small) Macoun *Stenanthium occidentale Gray Suksdorfia violacea Gray Trautvetteria grandis Nutt. Vaccinium membranceum Dougl. Woodsia scopulina D.C. Eaton 1958] DETLING: COLUMBIA RIVER GORGE 169 Ten species of the foregoing list (designated by asterisks) constitute an interesting group because of their relationship to the flora of Saddle Mountain. This peak is located in Clatsop County, Oregon, about 75 miles northwest of the western portal of the Gorge. Although only slightly over 3200 feet in elevation, it is the site of an isolated boreal flora which it may be assumed was at one time continuous with that now found in the Olym- pic Mountains (Detling, 1954). At the present time this flora has been eliminated from the areas between Saddle Mountain and the Olympics, and occurs only occasionally southward in the Oregon Coast Range, e.g., on Marys Peak. The ten indicated Columbia Gorge species have this pe- culiar and interesting distribution, occurring also on Saddle Mountain and in the Olympics, but so far as we know not between these points. 6. ENDEMIC SPECIES. Of the 206 species considered here as representing the flora of the Columbia Gorge, seven are endemic to the Gorge: Bolandra oregana Wats. Erigeron oreganus Gray Douglasia laevigata Gray Hieracium longiberbe How. var. laevigata Sullivantia oregana Wats. Erigeron howellii Gray Synthyris stellata Penn. It is true that two of those included in this category, Bolandra oregana and Sullivantia oregana, have been found at one other station outside the Gorge, namely at Elk Rock on the lower Willamette River, and thus are not truly endemic to the Gorge and in the strictest sense of the term; however, they must be considered here in any discussion of endemism as it relates to the history of the flora. Elk Rock is a sheer cliff on the west bank of the Willamette River a few miles above Portland and twenty-two miles by air from the west portal of the Gorge at Crown Point. The face of the bluff has an easterly and slightly northerly exposure; in certain spots seepage water keeps the surface wet most of the time. Apparently conditions here are sufficiently similar to those in certain niches of the Gorge to have encouraged the per- sistence of a few rare plant species which are otherwise found only in the cool, misty vicinity of the waterfalls of the Gorge. This situation suggests that during some period of glaciation in the Northwest the cold was in- tense enough to cause the extension of the boreal flora from the Cascades out over the valley floor at least as far as the Willamette River. There are two possibilities always to be considered regarding the origin of narrowly endemic species. In the first place, they may be of relatively recent origin, having risen through mutation or hybridization from some pre-existing species in the area they now occupy. Theoretically, given time and proper environmental conditions, such a species would be expected to increase in number of individuals and to extend into nearby territory. On the other hand, endemic populations may be relicts of species which once were more widely distributed, but which, due to some factor such as failure to cope with a changing environment have been reduced to their present limited range. 170 MADRONO [Vol. 14 Viewed as a whole, the endemic species of the Gorge would seem to be of the latter type. This conclusion is based upon the occurrence of two of them isolated at Elk Rock, upon the fact that they seem to be morpho- logically stable and not undergoing further evolutionary development, and finally that for the most part they have no near relatives in the vicinity. These seven species are well-defined entities, and the plant taxonomist has no difficulty in placing them in the category of “endemics.’”’ However, when one studies carefully the wide-ranging species represented in the Columbia Gorge his attention is soon called to the fact that evolutionary processes have frequently taken place in the Gorge, and presumably are still going on, which have made that portion of the species slightly, but still noticeably, different morphologically from the rest of the population. Specialists in certain groups have already called attention to some of these, e.g., Cronquist (1955) for Arnica amplexicaulis and Haplopappus hallu, but they are frequently loath to give such local forms taxonomic standing. However, we cannot ignore the fact that they are there, and if we wished to follow the lead of less conservative botanists and give vari- etal or subspecific names to these local entities we could increase consider- ably the number of endemics in our list. These would of course be of the first type as to origin. GENERAL DISCUSSION Viewed from the standpoint of the major migrations of the Pacific Northwest flora, there are two outstanding phases in which the geographi- cal situation, physiographic features, and structure of the vegetation of the Columbia River Gorge are probably significant. The first of these is the rdle of the Gorge as a gateway for the passage of lowland species from one side of the Cascade axis to the other. For the rather large number of species (81, or 39.4 per cent) now confined to one side or the other of the axis, the Gorge obviously has not fulfilled this function. Even among the 70 species with a general distribution both east and west it is unlikely that many have migrated through the Gorge. Judg- ing from their present widespread occurrence they probably reached the vicinity of the Gorge by parallelling the range on either side. There is evidence, however, that a significant number of lowland species from other groups listed above have migrated through the Gorge in assum- ing their present distribution. Chief among these are the Rogue Area species found in the Columbia Area, with their isolated occurrences in xeric islands west of the Cascades. It was inferred in the earlier brief dis- cussion of this group that the migration of its members through the Co- lumbia Gorge took place from west to east. The basis for this inference is the fact that the centers of distribution of these species are to the south and west today, and any evidence of former occurrence is now found west of the Cascades. Quercus garryana may be cited as an example. This mi- gration probably took place at a time of maximum northward extension of the Rogue flora. Floral migrations from an areal center normally occur 1958] DETLING: COLUMBIA RIVER GORGE 171 while climatic conditions at the center are becoming intensified and the environment in advance of the migration is becoming progressively more like that of the center. Since essentially the Rogue Area center is relatively warm and dry, the northward extension of the Rogue flora probably coin- cided with a xerothermic phase in the climatic cycle of the Pacific North- west. Studies on post-glacial forest succession in the Northwest, based upon pollen profiles from peat bogs (Hansen, 1955), indicate that such a warm, dry phase did set in approximately 8000 years ago and lasted until about 4000 year ago, with its maximum probably about 6500 years from the present. The other distinctive feature of the Gorge flora is the fact that such a large part of it constitutes an island of boreal forms isolated at the bottom of the main canyon or its smaller tributary canyons. These are in all prob- ability relict occurrences, left from a time when the boreal flora, pushed to lower levels from above, was continuous from the middle mountain ele- vations down to the present valley level. It is hard to conceive that such a large number of species would have been established at their present levels by being transported from higher elevations across a transitional zone during a period of temperate climate. Rather, it is more likely that there was a general downward shifting of the flora during a cold phase of the climatic cycle, and subsequent isolation of many of the species in fa- vorable niches as the climate became warmer and the main populations of the boreal species retreated to higher elevations. The presence of a number of Columbia Gorge species on Saddle Mountain and in the Olym- pics is another indication of a former continuity of the boreal flora at low levels in this part of the Pacific Northwest. The narrow endemics briefly discussed above were probably reduced to their present restricted range during or following the culmination of this cold maximum. The occurrence of some of these isolated along the lower Willamette River may indicate that at one period their range at valley level was considerably more extensive. The cold maximum which caused the downward migration of the boreal species must have preceded the warm, dry phase of climate previously mentioned, as the pollen profiles indicate only moderate cooling in the last 6000 years. These same profiles, correlated with radiocarbon datings (Broecker, Kulp, and Tucek, 1956), indicate that the last major cold maximum in this part of the continent occurred about 12,000 years ago, a time probably coinciding with the maximum advance of the last conti- nental ice sheet in western North America. Museum of Natural History University of Oregon, Eugene LITERATURE CITED Broecker, W. S., J. L. Kurp, and C. S. Tucex. 1956. Lamont natural radiocarbon measurements. III. Science 124:154~-165. CRONQUIST, ARTHUR. 1955. Compositae, ix C. Leo Hitchcock, Arthur Cronquist, Marion Ownbey, and J. W. Thompson, Vascular plants of the Pacific Northwest, Part 5. Univ. of Washington Press, Seattle. 172 MADRONO [Vol. 14 Detiinc, LERoy E. 1948. Concentration of environmental extremes as the basis for vegetation areas. Madrono 9:169-185. . 1953. Relict islands of xeric flora west of the Cascade Mountains in Ore- gon. Madrono 12:39-47. . 1954. Significant features of the flora of Saddle Mountain, Clatsop County, Oregon. Northwest Science 28:52-60. Hansen, Henry P. 1955. Postglacial forests in south-central and central British Columbia. Am. Jour. Sci. 253:640-658. THE GENUS COLLINSIA. III. THE SIGNIFICANCE OF CHIAS- MATA FREQUENCIES AS A CYTOTAXONOMIC TOOL? E. D. GARBER The genus Collinsia Nutt. (Scrophulariaceae) includes twenty-one rec- ognized species, divided into two groups (Newsom, 1929; Pennell, 1951). The species in one group have “‘sessile”’ flowers congested in whorls, with pedicels shorter than to no longer than the calyces of the lower whorls, and with flat, mature seeds. The species in the other group have pedicelled flowers, either solitary or in whorls, with the pedicels of the flowers of the lower whorls from as long as to longer than the calyces, and with either flat or thick, mature seeds. There are other differences between the species in these two groups but they are not as clear as those which have been mentioned. The basic chromosome number for the genus is 7; no polyploid species have yet been found (Garber, 1956, and unpubl.). The species in each of the two groups apparently differ in their mean number of chiasmata per bivalent at metaphase I. With the exception of C. corymbosa Herder, the species with “‘sessile” flowers have mean values of 1.1—1.5 and the species with pedicelled flowers, 1.7—1.9 (Garber, 1956). Collinsia corymbosa was placed in the species group with “sessile” flowers by both Newsom (1929) and Pennell (1951), yet its combination of char- acters shows it to be somewhat intermediate between these two groups of species (Garber and Gorsic, 1956). The flowers are borne in dense, capitate whorls on pedicels 3—7 mm. long, with calyx lobes approximately 5 mm. long, and the mature seeds are thick. The mean number of chias- mata per bivalent at metaphase I in C. corymbosa, however, has been found to be 1.7—1.8, a value characteristic of the group of species with pedicelled flowers. This paper presents evidence regarding the validity of chiasma fre- quency as a cytotaxonomic tool in studying relationships among species of Collinsia by considering the chromosome associations and aberrations 1 This investigation was aided by grants from the National Science Foundation and the Dr. Wallace C. and Clara A. Abbott Memorial Fund of The University of Chicago. 1958] GARBER: COLLINSIA t73 in interspecific hybrids involving C. heterophylla, a species with “sessile” flowers. TABLE 1. A comparison of certain morphological characters of C. corymbosa, C. heterophylla, and their interspecific hybrids (hey 56632). Character Interspecific Character C. heterophylla hybrid C. corymbosa Flower, position “sessile” corymbosa-like “sessile” Flower, upper lobes, length prominent intermediate rudimentary Flower, upper lobes, markings present present absent Flower, lateral lobes, color blue-red segregating * cream white Flower, tube color pale blue-red light blue light blue Flower, upper stamen filaments basal spurs no basal spurs no basal spurs Calyx lobes glabrous hairy, glandular hairy, glandular Leaf petiole short long long * Pale or light blue. MATERIALS AND METHODS For the current studies, plants were grown from seed which was kindly supplied by Dr. R. Bacigalupi. Of the species involved, C. corymbosa is apparently restricted to the area around Fort Bragg, Mendocino County, California; C. heterophylla Buist, typical of the group with “sessile’’ flow- ers, may be found throughout the hilly portions of the western regions of California from the extreme south almost to the Oregon boundary; C. sparsiflora Fisch. and Mey., representative of the pedicelled group, occurs at low and middle elevations northward in the California coast ranges from Marin County and in the Sierra Nevada from Tuolumne County to Butte County. Clusters of buds were fixed in a solution (6:3:2) of methanol, chloro- form, and propionic acid (Pienaar, 1955), which proved superior to the familiar alcohol-acetic acid fixative, and the buds were then stored in a deep freezer until needed. Smears of pollen mother cells were stained with acetocarmine. Pollen grains were stained with basic fuchsin in lacto- phenol. HYBRIDIZATION RESULTS Interspecific hybridizations involving “‘sessile’’-flowered C. heterophylla and C. corymbosa were easily accomplished. The yield and quality of the resulting seeds were excellent and approximately 90 per cent of the seeds germinated. Interspecific hybridizations involving the pedicelled-flowered species C. sparsiflora and C. corymbosa, however, were almost completely unsuccessful. 174 MADRONO [Vol. 14 TABLE 2. Chromosome configurations at metaphase I in interspecific hybrids (hcy 56632) between C. corymbosa and C. heterophylla. Plant No. II ee ot IV 3 5 6 10 7 am 14 48 42 19 6 2 19 15 61 18 5 4 12 4 24 18 4 6 2 4 4 3 8 oust 2 5 1 1 1 2 Het 1 4 3 1 ae 1 3 ) 1 ive 1 ae 1 5 1 chain ae 1 2 1 Sip ee, Sore 1 ring eect or soe! 1 4 rs 1 chain La oe 1 2 4 oe, ae 1 ring arp 1 oon 1 3) Ae. ‘pects 1 chain a hie 1 No. of pollen mother cells 49 i 138 66 MorPHOLOGICAL STUDIES. Only the interspecific hybrids between C. heterophylla and C. corymbosa were studied. The morphological differ- ences between these species are so obvious that there is no difficulty in distinguishing these species. Certain characteristics of each species and their hybrids are summarized in Table 1. In general, the interspecific hybrids resembled C. corymbosa in their vegetative characteristics, but the flowers were more like those of C. heterophylla. The hybrids were intermediate in height but almost as tall as C. heterophylla. CytoLtocy. The chromosome associations at metaphase I in the inter- specific hybrids are summarized in Table 2. Most pollen mother cells had univalents and bivalents. A few pollen mother cells also had a trivalent or a quadrivalent. The mean number of chiasmata per bivalent at meta- phase I in pollen mother cells with only bivalents was 1.3—1.4, which were the values observed for hybrids between C. heterophylla and C. sparsi- flora (Garber and Gorsic, 1956). No pollen mother cells had two trivalents or two quadrivalents. Since many pollen mother cells had univalents, it was conceivable that a ring of six chromosomes, the result of two reciprocal translocations involving one chromosome, could have been formed. Several configurations make this interpretation unlikely. The conclusion that at least a single heterozy- gous reciprocal translocation occurred in the interspecific hybrid appears to be reasonable. Pollen mother cells at telophase I often displayed a dicentric chromatid bridge and a very small fragment; a few cells had two bridges and two very small fragments (Table 3). These observations in- dicate that the interspecific hybrids had two heterozygous paracentric inversions. 1958] GARBER: COLLINSIA 175 TaBLE 3. Number of pollen mother cells with dicentric chromatid bridges and fragments at telophase I in interspecific hybrids between C. corymbosa and C. heterophylla. No. of bridges + fragments No. of pollen Plant No. 0 1 2 mother cells 5 55 24 4 83 6 50 27 2 79 10 13 31 7 111 STERILITY. The interspecific hybrids were almost completely pollen- sterile, with less than 0.3 per cent stainable pollen grains. No seeds were set even when the flowers were hand-pollinated. DISCUSSION The group with “‘sessile’” flowers includes the following species: C. heterophylla, C. concolor, C. tinctoria, C. bartsiaefolia, C. multicolor, C. austromontana, and C. corymbosa (Pennell, 1951). It has been possible to assemble data on the crossability, the fertility, and chromosome asso- ciations in interspecific hybrids involving C. heterophylla, C. concolor, C. tinctoria, and C. bartsiaefolia as well as in interspecific hybrids between C. sparsiflora and both C. heterophylla and C. concolor (Garber, unpub. ; Garber and Gorsic, 1956; Hiorth, 1933). When hybrids can be made between species, intragroup hybrids are more difficult to accomplish than intergroup hybrids, the yield of germi- nating seeds being much greater for the intergroup hybrids. Intragroup hybrids involving species with “sessile” flowers were fertile to some de- gree; intergroup hybrids were completely sterile. The hybrids between C. corymbosa and C. heterophylla behaved as intergroup hybrids. Intragroup hybrids involving species with “sessile” flowers displayed only bivalents at metaphase I; intergroup hybrids had such chromosomal aberrations as heterozygous reciprocal translocations and heterozygous paracentric inversions. These chromosomal aberrations were responsible for the complete sterility of the intergroup hybrids. In this respect also, the hybrids involving C. corymbosa and C. heterophylla behaved as inter- group hybrids. It seems reasonable to assume that barriers to hybridization have oc- curred for the species within the group with “‘sessile”’ flowers. Since such hybrids are fertile to some degree, such barriers would minimize the pos- sibility of large scale introgression. The complete sterility of hybrids between species of different groups which easily hybridize serves as an effective barrier against introgression. It is not yet clear what significance may be attached to the observation that hybrids between C. corymbosa and C. heterophylla have a mean number of chiasmata per bivalent at metaphase I of 1.3—1.4, a value also found in the interspecific hybrids involving C. sparsiflora and both C. 176 MADRONO [Vol. 14 heterophylla and C. concolor. It is possible that this observation may be related to the length of the homologous segments in the chromosomes of the species in different groups. Although the evidence may be interpreted to indicate that C. corymbosa does not belong with the species having “‘sessile”’ flowers, it appears more reasonable at this time to consider that C. corymbosa does not belong to the group of species including C. heterophylla, C. concolor, C. tinctoria, and C. bartsiaefolia. The other two species with “‘sessile“ flowers, C. mul- ticolor and C. austromontana, may yield critical information on this point. At any rate, the differences in the mean number of chiasmata per bivalent at metaphase I appear to have cytotaxonomic value in studying relation- ships among the species of Collinsia. It must remain for future investiga- tion to determine the extent to which this tool may be used. SUMMARY Interspecific hybridizations between C. corymbosa and C. heterophylla were easily accomplished, yielding a very high percentage of germinating seeds. The hybrids were completely sterile. Different numbers of bivalents and univalents were observed at metaphase I and, occasionally, a tri- valent or quadrivalent was seen at the same stage. The multivalent was interpreted as a heterozygous reciprocal translocation. One or two dicen- tric chromatid bridges plus one or two very small fragments were observed at telophase I, indicating the presence of at least two heterozygous para- centric inversions. The combined data indicated that C. corymbosa does not belong with a number of species with “sessile” flowers but did not conclusively demonstrate that this species does not belong within the group of species having “‘sessile”’ flowers. The results indicate that differ- ences in chiasma frequency appear to have cytotaxonomic value in study- ing relationships among the species of Collinsia. The author is indebted to Mr. J. Gorsic for his technical assistance and to Dr. G. L. Stebbins for reading the manuscript. Department of Botany, The University of Chicago Chicago, Illinois LITERATURE CITED GarBER, E. D. 1956. The genus Collinsia. I. Chromosome number and chiasma fre- quency of species in the two sections. Bot. Gaz. 118:71-73. GarBeER, E. D. and J. Gorsic. 1956. The genus Collinsia. II. Interspecific hybrids involving C. heterophylla, C. concolor, and C. sparsiflora. Bot. Gaz. 118:73-77. Hiortu, G. 1933. Genetische Versuche mit Collinsia. IV. Die Analyse einer nahezu sterilen Artbastardes. Pt. 1. Zeit. Ind. Abst. Vererb. 66:106—157. Newsom, V. M. 1929. A revision of the genus Collinsia. Bot. Gaz. 87:260-301. PENNELL, F. W. im Abrams, L. 1951. Illustrated Flora of the Pacific States. Vol. 3. Stanford University Press, Stanford, California. Prenaar, R. ve V. 1955. Combinations and variations of techniques for improved chromosome studies in the Gramineae. Jour. So. African Bot. 21:1-8. INFORMATION FOR CONTRIBUTORS Manuscripts submitted for publication should not exceed an estimated 20 pages when printed unless the author agree to bear the cost of the ad- ditional pages at the rate of $15 per page. Illustrative materials (includ- ing “typographically difficult” matter) in excess of 30 per cent for papers up to 10 pages and 20 per cent for longer papers are chargeable to the author. Subject to the approval of the Editorial Board, manuscripts may be published ahead of schedule, as additional pages to an issue, provided the author assume the complete cost of publication. Shorter items, such as range extensions and other biological notes, will be published in condensed form with a suitable title under the general heading, ‘“‘Notes and News.” Institutional abbreviations in specimen citations should follow Lanjouw and Stafleu’s list (Index Herbariorum. Part 1. The Herbaria of the World. Utrecht. Second Edition, 1954). Articles may be submitted to any member of the Editorial Board. MADRONO A WEST AMERICAN JOURNAL OF BOTANY A quarterly journal devoted to the publication of botanical re- search, observation, and history. Back volumes may be obtained from the Secretary at the following rates: Volume I, 1916-1929 . . . . . $5.00 Volume II, 1930-1934. . . . . 5.00 Volume III, 1935-1936 . . . . 5.00 Volume IV, 1937-1938 . . . . 5.00 Volume V, 1939-1940. . . . . 5.00 Volume VI, 1941-1942 . . . . 5.00 Volume VII, 1943-1944 . . .. . 5.00 Volume VIII, 1945-1946 . . . . 5.00 Volume IX, 1947-1948. ae 6.00 Volume X, 1949-1950. . . . . 7.00 Volume XI, 1951-1952 . . . . 8.00 Volume XIT, 1953-1954 . . .. . 8.00 Volume XIII, 1955-1956 . . .. Sg 8.00 Single numbers, Volumes} dnd 2s. 5 mos .50 Succeeding numbers, (except Volume XI, No.2). . 1.00 Volume XESNO! 22) eke es ee 1.50 The subscription price of MapRoNo is $4.00 per year. If your institution does not now subscribe to MADRONO, we would be grateful if you would make the necessary request. Since there is no extra charge for institutional subscriptions, an individual may hold membership in the California Botanical Society on the basis of his institution’s subscription. Address all orders to: Jean H. LANGENHEIM, Corresponding Secretary Department of. Botany University of California, Berkeley 4, California 48S eee ba Stal OE RR Es | Sm Piss Sis MADRONO é oullse = rt 7 VOLUME 14, NUMBER 6 APRIL, 1958 q a < 2 ES | eo) } ma 4 : ft Contents r © Bee % z THE PHYLOGENETIC DIVISION OF THE SUBFAMILY ‘i 7 CEREOIDEAE, CACTACEAE, Franz Buxbaum NUCL Pa Me e MOsSsEs OF CALIFORNIA VI. HALL NATURAL AREA os AND Mono County, Leo Francis Koch 206 vt a ™S Reviews: The Future of Arid Lands, edited by Gilbert K 7, F. White (M. Zohary); Leonid Enari, Plants of the * < Pacific Northwest (Robert Ornduff); Richard B. O — aa isa = < = MN wel < Goldschmidt, Portraits from Memory. Recollections of a Zoologist (H. F. Copeland); Rogers McVaugh, Edward Palmer, Plant Explorer of the American West (Joseph Ewan) 212 NotTES AND NEws: VARIABILITY IN TRILLIUM OVATUM PursH, Vesta F. fosse 216 ’ Jd’ * f/f bese? Sy ff \ 4 i; ; ‘\ " \ Co AG ee \h POWAY 206 1300") \ WAL & 6 IO Wf Y\ J / A J J \ ely £ 3 Lf _— =