= omer ive, sag nme p nrereneer en ee ae ee ee ar ene on eT EIS Verney eee PT ee 2 ae oe reg telenr ime ae seen ee ee aes seme Oe een eae! wore Pope RAMEE CF tere Pe eT ae ee Pre cs eer ee ee | FO ee a ee oe piereue Wie Soe eetewe when cleg te rdemree 28 a Sc) weer a ke Cee tant ey Je ee 6 28 nell ee Poe pepe eee ee ree Co aS eee ne aye eis oo ig Seg geee - a+ “so POR ye Osetia ee ee ree Coane aap platelets oop 8 eee vale g aw o ERLE wie) eee MERTEN ee Ne eee By ode beeen ker wey eh A Ee ee, RS SPs . em Aap SFE tp me ese eTeNTON ee © ae Shoe ewes - en ee ee weer A os . ‘ Poe ow 4 wos tees TA wes sige 0 epi D-m ctninl ee 7 “ ; oe OS oe ue «4 wits or, ¢ shinee Bete ane tema ergs s toe de waeweele oo vieter ane lee Pe 7 esr acre -—ww err pe Saag nrgcays ges - emai ti ele one 7 = vue <2 - ew —_ ee Prater Te ad fier wee 7 : : ee PR ge eet emery rr ee? pes wee wee erate gta Spee ie o wn eae degre phges on orere el a wwe 8 ale ne See eg we ee RIT pee et svehrus 9 > st pee 9 ee Pee where’ * rie ee me Ne trent Mt Bx BAe By St ed ae slewece rls wrote ee ee a ee ak comeing wine oA nr tet pee we we eee rete ee she eee semen we cop ategerteese 6 gan ws: Se ree PI ie ee ae Ay “Cs sige 04 ‘SPDs (Photograph by Harris and Ewing, Washington, D.C.) THOMAS HENRY KEARNEY Frontispiece to Volume 11, Maprono MADRONO A WEST AMERICAN JOURNAL OF BOTANY VOLUME XI 1951 — 1952 Published quarterly by the California Botanical Society, Inc., 2017 Life Sciences Building, University of California. Berkeley 440 South 11th Street. Lincoln, Nebraska Board of Editors HERBERT L. Mason, University of California, Berkeley, Chairman. EDGAR ANDERSON, Missouri Botanical Garden, St. Louis. LyMAN Benson, Pomona College, Claremont, California. HERBERT F. COPELAND, Sacramento College, Sacramento, California. JoHN F. Davipson, Department of Botany, University of Nebraska, Lincoln. Ivan M. JoHNSTON, Arnold Arboretum, Jamaica Plain, Massachusetts. MitpreD EF. Matutas, Department of Botany, University of California, Los Angeles 24. MarIoN OWNBEY, State College of Washington, Pullman. IRA L. Wiacins, Natural History Museum, Stanford University, Stan- ford, California. Secretary, Editorial Board ANNETTA M. Carrer, Department of Botany, University of California, Berkeley. Business Manager and Treasurer RICHARD W. Hom, Natural History Museum, Stanford University, Stanford, California. Assistant Business Manager Joun F. Davipson, Department of Botany, University of Nebraska, Lin- coln, Nebraska. To Thomas Henry Kearney, agricultural and botanical ex- plorer, creative researcher and authority on cotton culture in arid regions, and pre-eminent student of the flora of Arizona. the California Botanical Society dedicates this eleventh volume of Madrono. His retirement from the Bureau of Plant Industry after long and eminently useful service in the field of agricul- tural research has, during the past eight years, given him opportunity to devote himself to further studies on the flora of Arizona and to investigate more fully the taxonomy of the Malvaceae. A modest and retiring scholar with a keen and friendly sense of humor, Dr. Kearney commands the affection and respect of all who have had the good fortune to make his ac- quaintance. We welcome this opportunity to express to him our appreciation of his active participation and counsel in the affairs of the California Botanical Society and we wish him continued good health and much happiness. ili i i a or ~ ~,! i> = ens» ‘ 3 f : * Fi ‘ - : NS Pe ' : f i : Au ; 2, F Wk ay - e a i 7 7 a bet . ? 7 + is . fa 7 By . ‘ : ae + 7 ; iS 7 a © ™~ A = 7 ars eo P 7 ve 7 a 7 ' . of . ) -" 7 ;, s : aal iy ; = i 7 5 oc, 3 : F a; . \ a , re 2 : ; XS 3 i 7 é ; 2 7 - s aad é a 7 5 - Pe 1 i bs i i : 7” Pye 7 # ; ‘ - : ~ "aS S ~ 7 a z . - 32 A = al * - : - A : : Pee << . i / a Poot 7 = - 7 = aK au ed Re "5 : oe : oxime : u at co , 1 + ~ ¥ : + ¥ Cp ¥ \ ‘ ; - r er, , . e - . . - 4 - ‘ : : ae 2 ; = / : : Sef, + ae ‘ “ Ba ‘ oe / ; - : + rand DS. PRS a4 ha ai to : \ F $ 5 . by t 5 - . : ‘ 2 7 - - ; ie 43 - a 2 a : : : ‘ cg 7 , : : J ‘ S 7, ie ; : ; - ’ - m= * ot ane ee oe =x tray tae uM A a of ' ‘ 7 6 ‘ ”y 4 i i - , i ks ; ; P Laine i f ee a 2 7 < o 7 . a j C 3 7 at . o = : Ps : / } ; ’ > ~ : oe an fas - A i n , , = we : 7 5 z as : ¢ i Fr halt” ' ; ‘ Mm i ae - 2 : * ' int 7 we by i 7 7 ue GY : : a oes F 7 : a Md . 3s 7 4 7 . = 2 7 U _ + E a8 a ‘i. - . 7 = os 7 : = y wre a1 ae CAP =a § i 7 on ae : : ; A - som ange ee wee a ae AT Ua vy 7 a : e ¢ VE . > ‘ om - ry . 7 y * Fea 0 = 7 3 5 . A ‘3 ft foe ed , & . 7 es a ‘ - . ; - i - va ~ . - ee « mY st 7 = ; : ; ’ . : _ * : 2 . ct - 7 : 4 an , i ; ‘ . + ; = : : * oe ‘ ba = . ‘ . . ‘ t of * 7 ao - oe 1 o . ‘ Fa nap Aa “ - \ =! ' ® — . nee i s ot - 4 i é a A CONTENTS Frontispiece: Thomas Henry Kearney Youngia americana, a N ew Species of Phyletic Significance Ernest B. Babcock Cytotaxonomic Sunes in the Genus Sorghum, II. Two new Species from Australia............ E. D. Garber and L. A. Snyder Evidence for the Hybrid Nature of x Liatris creditonensis L. O. Gaiser MOrton, WatOher CCK. wie c1.. iia cries ee Lincoln Constance TEMAS Meee is ans oda 24, 149, 187, 214, 251, 279, Studies in the Genus Diplacus (Scrophulariaceae) Howard E. McMinn The Significance of Vegetative Reproduction in Quercus Cornelius H. Muller A New Fritillaria from Oregon.............0.000000.....0.. Helen M. Gilkey A Cytotaxonomic Approach to Eschscholtzia Harlan Lewis and Richard Snow Nomenclatural Recombinations in Idaho Plants............ Ray J. Davts A New Artemisia from Wyoming..................0.000..... Arthur Cronquist A New Genus of Ecuadorean Araceae........................ Alex D. Hawkes INOGES “IGN CWS ooi cco gl ie oe pein ng cation gs en etcetera bees cubs ceed LoZe 202, Notes on the Flora of Guadalupe Island, Mexico........... Reid Moran Supplement to the Flora of Whatcom County, Washington W. C. Muenscher An Anthracnose Disease of Umbellularia californica John M. Harvey Botany of the Intermountain Region. II. Lesquerella Bassett Maguire and Arthur H. Holmgren Aquatic Plants in Ozette Lake, Washington........ W. C. Muenscher A New Haplopappus from New Mexico... Arthur Cronquist The Third Locality for Cupressus Abramsiana Wolf Calvin McMillan Bacterial Leaf Spot of Umbellularia californica ...John M. Harvey Howellanthus, a New Subgenus of Phacelia....... Lincoin Constance Genetic Variation in Andropogon John F. Davidson and Paul F. Romberg A New Bacopa from California.............0000000000.0....... Herbert L. Mason Bryophytes of Chetco River Redwood State Park, Oregon Leo Francis Koch Studies in Streptanthus. A New Streptanthus Complex in Cali- HOV OU ES aeNet age eae eta Wc nein See ie ce eat eae Freed W. Hoffman Taxonomic Interrelationships in the Quercus dumosa Complex John M. Tucker The Mexican Itineraries of T.S. Brandegee.................... Reid Moran The Status of Lophotocarpus in Western North America Herbert L. Mason Wire sPimces* Wexican Pines wc tee ae N. T. Mirov Notes on Vaucheria longicaulis Hoppaugh...Wm. Randolph Taylor Carex diversistylis, a New Species from Oregon Archibald W. Roach Notes on Malvaceae. III. Abutilon and Pseudabutilon in the Gala- DAS OSMISTATIOS! & oe acces ose cece ccna Wik eens esses Thomas H. Kearney Plankton Algae of some lakes of Whatcom County, Washington George J. Schumacher and W. C. Muenscher A cytotaxonomic study of the genus Geranium in the Wasatch resion of Idaho and, Utab...........c.6c.6g.0csct sens. Richard J. Shaw Chromosome numbers in the genus Amsinckia............ Peter Kamb Geis Fep. Cae AV Ae) Oba 0 ao, GARE RnR San Saint aa Ra ey eit ner nr Oe een ERRATA Page 10, line 33: for otherhand read other hand. Page 118, line 32: for hybirds read hybrids. Page 158, line 6: for Lavatera insularis read Lavatera insularis. Page 178, line 39: for LESQURELLA read LESQUERELLA. Page 281, line 13: for west read east. vi VOLUME XI NUMBER 1 MADRONO A WEST AMERICAN JOURNAL OF BOTANY (FEB 9 \ a {Pree +a \ L, D fi ite t ] i i A. re aN” 24 a ae f# >a i As Ig SS 4 Nati hart yse Ss se UNA WS a ia seuneaccaneaese= ee Contents YOUNGIA AMERICANA, A New Species or Puyneric Signiricance, Ernest PES CO CIC IU tea Nar Rin fe Be RRL rie ww Me Veh UNe i RARER oy CRPM ye Mages 1 Cyroraxonomic Stupies In THE Genus SorcHuM. II. Two New Sprctis From AustrauiA, EH. D. Garber and L. A. Snyder .................. 6 E\vmence ror THE Hysrip Narvure or X Liatrais crepironensis, L. O. Gaiser 10 Morton Eaton Pecx, Lincoln Constance ............0...0 00 0c eee 22 Reviews: Ernst Giumann, Principles of Plant Infection (H. N. Hansen) ; William Randolph Taylor, Plants of Bikini and other Northern Marshall Islands (George F. Papenfuss); C. R. Metcalf and L. Chalk, 24 Anatomy of the Dicotyledons (Herbert L. Mason) ................. Published at North Queen Street and McGovern Avenue, Lancaster, Pennsylvania January, 1951 MADRONO A WEST AMERICAN JOURNAL OF BOTANY Board of Editors Hersert L. Mason, University of California, Berkeley, Chairman. LeRoy Apgrams, Stanford University, California. Epncar Anpberson, Missouri Botanical Garden, St. Louis. Lyman Benson, Pomona College, Claremont, California. Hersert F. Corperann, Sacramento College, Sacramento, California. Ivan M. Jounston, Arnold Arboretum, Jamaica Plain, Massachusetts. Mivpeep KE. Maruiss, Dept. of Botany, University of California, Los Angeles 24. Bassett Macuire, New York Botanical Garden, N. Y. C. Marion Ownzey, State College of Washington, Puilman. Secretary, Editorial Board—ANNETTA CaRTER Department of Botany, University of California, Berkeley Business Manager—Ricuarp W. Horm North Queen Street and McGovern Avenue, Lancaster, Pennsylvania or Natural History Museum, Stanford University, California Entered as second-class matter October 1, 1935, at the post office at Lancaster, Pa., under the act of March 3, 1879. Established 1916. Published quarterly. Subscription price $4.00 per year. Completed volumes I to VIII, $5.00 each; volume IX, $6.00; volume X, $7.00; Single numbers $1.00. Manuscripts submitted for publication should not exceed 20 pages when printed nor contain more than 20 per cent illustrative material unless the author agrees to bear the additional costs. Range extensions and similiar notes will be published in condensed form with a suitable title under the general heading “Notes and News.” Articles may be submitted to any member of the editorial board. Manuscripts may be included in the forthcoming issue provided that the contributor pay the cost of the pages added to the issue to accommodate his article. Reprints of any article are furnished at a cost of 4 pages, 50 copies $7.67; 100 copies $8.43; additional 100’s $1.59; 8 pages, 50 copies $11.13; 100 copies $12.35; additional 100’s $2.44; 16 pages, 50 copies $15.61; 100 copies $17.48; additional 100’s $3.74. Covers 50 for $4.72; additional covers at $2.84 per hundred. Reprints should be ordered when proofs are returned. Published at North Queen Street and McGovern Avenue, Lancaster, Pennsylvania, for the CALIFORNIA BOTANICAL SOCIETY, INC. President: George F. Papenfuss, University of California, Berkeley, Cali- fornia. First Vice-President: Lyman Benson, Pomona College, Claremont, California. Second Vice-President: Annetta Carter, University of California, Berkeley, California. Secretary: Phyllis G. McMillan, Department of Botany, University of California, Berkeley, California. ‘Treasurer: Richard W. Holn, Natural History Museum. Stanford University, California. Annual membership dues of the California Botanical Society are $4.00, which includes a year’s subscription to Madrono. For two members of the same family the dues are $5.00, which includes one copy of Madrono and all other privileges for both. Dues should be remitted to the Treasurer. General correspondence and applications for membership should be addressed to the - Secretary. YOUNGIA AMERICANA, A NEW SPECIES OF PHYLETIC SIGNIFICANCE Ernest B. Bascockx A single specimen, collected last year in northwestern Alaska, is of unusual interest for several reasons. Although the plant is reminiscent of Crepis nana typica in its tufted habit and dark green, glabrous foliage, yet its most distinctive features are those of either Soroseris or Youngia, both of which genera hitherto have been restricted to Asia in their natural distribution (Y. japonica has been introduced as a weed widely in tropical regions). Since this plant combines certain characters of two Asiatic genera, it presents a problem in classification, especially in view of the absence of mature achenes. Immature achenes, however, are present, and a critical comparison of the material available with species of both genera provides a sound: basis for classification. It appears certainly to belong in Section 1, Desiphylum, of the genus Youngia, near Y. conjunctiva. Youngia americana sp. nov. Herba perennis valde humilis viridissima glabra; caulis subterraneus rectus tenuis cataphyllis triangularibus vestitus basi crassus cavus dense foliosus; folia caudicalia 2—7 cm. longa spathulata vel pinnatifida, ea caulina elongata tenuissima; caules aerei numerosi 4 cm. longi dense ramosi; capitula numerosa parva congesta 11—13-flora; involu- crum cylindricum squamis exterioribus 3—4 linearibus eis interiori- bus 7 lanceolatis; corolla 10 mm. longa tubo 4.5 mm. longo glabro; antherae 2.5 mm. longae flavae; rami styli fusci 1.5—2 mm. longi; achenia (immatura) 2—3 mm. longa 11—13-costata costis 5 quam ceteris latioribus; pappus albus 6—7 mm. longus. Perennial, tufted, ca. 8 cm. high, 12 cm. wide, dark green, glabrous; caudex ca. 3 cm. long, gradually broadened upward to 1 cm. in width including the leaf bases, hollow, terminating a slender vertical subterranean stem bearing cataphylls; longest caudical leaves, forming a rosette, ca. 7 cm. long, 1.4 cm. wide, the stout reddish petiole equal to the blade, lanceolate, pinnatifid, with narrow remote oblong-acute or triangular segments, dark green on upper face, yellowish on lower face, the margin retrorsely revolute; lowest caudical leaves 2 cm. long, spatulate, with oblanceolate blade 6 mm. long, 2 mm. wide; cauline leaves similar to the rosette leaves but reduced, with narrow elongated apical segment, the longest ca. 4 cm. long, the others 1—2 cm. long, filamentous; flower stems numerous, ca. 4 cm. long including the heads, reddish toward the base, paniculately closely branched, bearing ca. 6 heads on short slender peduncles; heads small, num- erous, congested, 11—13-flowered; involucre cylindric, ca. 9 cm. long, 3.5 cm. wide at middle; outer bracts 3 or 4, narrow, terete, Manrono, Vol. 10, No. 8, pp. vi, 225-256. November 10, 1950. FEB 1 9 1951 2 MADRONO | [Vol. 1] arcuate, 14-14 as long as the inner ones; inner bracts 7, slightly coalesced at the base, in two ranks, outer ones 1.3 mm. wide, with a low thickened median protuberance near the apex, inner ones 2—-2.5 mm. wide, broadly scarious-margined, shortly white-ciliate at the apex, glabrous on inner face; corolla 10 mm. long, the tube 4.5 mm. long, glabrous; ligule 2 mm. wide, the teeth 0.5—1 mm. long, prominently crested, dark brown in sic.; anther tube yellow, 2.5 mm. long, 1 mm. wide when opened out, the appendages 0.5 mm. long, linear, acute or acuminate, the filaments 1 mm. longer; pollen grains echinolophate, ca. 32 » in diameter, the spines ca. 3 wand similar to Y. conjunctiva; style branches dark brown, 1.5—2 mm. long, 0.1 mm. wide, the barbs not as prominent as in Y. con- junctiva; achenes (immature) pale tawny, 2-3 mm. long, 0.6—0.8 mm. wide, somewhat obcompressed, fusiform, constricted be- low the expanded pappus disk, 11—13-ribbed, the ribs unequal with 5 much stronger ones; pappus white, tinged yellowish at the base, 6—7 mm. long, 4-seriate, the setae nearly equal in length, un- equal in width, the coarsest ones 50-60 p» wide at base, stiff but pliable, persistent. Flowering July, flowers yellow, anther tube yellow, stigmas dark brown. : Northwest Alaska, between latitude 68° 80’ and 69° 30’ N., along the Kukpowruk River. June—July, 1949 Robert M. Chapman 182 (U. S. Nat. Herb. no. 1973474). According to Dr. Ira L. Wiggins (in litt.) Mr. Chapman made his collections almost en- tirely along the middle portion of the Kukpowruk River during his field stay in that part of Alaska. This new species resembles either Youngia or Soroseris more than Crepis in its hollow caudex, its elongated, filamentous upper eauline leaves, the revolute margins of the caudical leaves, and in several features of the involucres, corollas, styles and achenes. Furthermore, the pollen grains in particular are like those of Youngia conjunctiva and very different from those of Crepis nana. The pollen grains of two species of Soroseris and of Crepis sibirica (a primitive species) and C. tectorum (a more advanced species) are illustrated in the monograph on Dubyaea and Soroseris (Steb- bins, 1940, pp. 29-30). The pollen of Y. americana is similar to that of C. sibirica. Soroseris pollen is orange in color, becoming greenish blue when stained with cotton blue in lactophenol; whereas the pollen of Y. conjunctiva, Y. americana and Crepis nana is yellow and stains a bright, clear blue. The data on diameter of the grains and length of the spines for pollen of S. Hookeriana erysimoides, Y. conjunctiva and Y. americana are given in Table 1. In Crepis nana it was found that the pollen is much smaller than in the two Youngia species, being only 27 uw in diameter, and that the grains are devoid of spines, the surface being merely rugose. Table 1 is a synoptical comparison of eleven morphological features and-‘one anatomical character of Y. americana with those 1951] TABLE |}. Stem Cataphylls Rosette leaves Cauline and uppermost leaves Outer bracts Inner bracts Corollas Pollen grains Style Achenes Pappus Ovary anatomy / BABCOCK: YOUNGIA SYNOPTICAL COMPARISON OF Y. AMERICANA WITH SOROSERIS AND YOUNGIA SOROSERIS Usually short, thick and often hollow above ground, arising from a long rootstock Present on stem below rosette leaves Rarely pinnatifid, the margin not revolute Gradually reduced, uppermost elongated, linear 2 (or 3-4), narrowly linear, shorter than or exceeding inner bracts 4-15, lanceolate or ovate, not crested, somewhat coalescent at base Ligulate, yellow or white, often blackish at base of ligule, 8- 21 mm. long, the tube about 4 the total length (in one species 1/3) Subechinolophate or echinate, the spines large. In S$. Hookeri- ana grains average 40, in diam., spines 6-8 Blackish or greenish, stigmatic branches l- 3.5 mm. long, 0.2-0.3 mm. wide, the barbs on outer surface numerous, close, prominent Oblong or more atten- uate toward base, con- stricted at apex, Sli- ghtly obcompressed, many-striate 7-16 mm. long, white, stramineous, yellow, gray, dusky or rufes- cent; the setae very numerous, coarse, stiff Primitive, numerous supernumerary strands Y. AMERICANA Short above ground, gradually thickened, hollow, arising from a rootstock Present on stem below rosette leaves Pinnatifid, the margin revolute Gradually reduced, the uppermost elongated, linear 3, narrowly linear, # - 4 as long as inner ones T, lanceolate, not crested but the median nerve or keel definite- ly thickened near the apex, coalesced only at the base Ligulate, yellow, 10 mm. long, the tube nearly ¢ the total length Echinolophate, the spines similar to Y. conjunctiva. Grains average 32.5 in diam. spines ca, 3-4 Brown (in. sic.), branches 1.5-2 mm.long 0.1 mm. wide,the barbs on outer surface simi- lar to &. umbrella but not quite as prominent as in Y. conjunctiva immature achenes fusi- form or oblong, + ob- compressed, 10-13 rib- bed, the ribs unequal with 5 much stronger ones 6-7 mm. long, white; the setae 4-seriate, resembling Y. conjunce- tiva Advanced, no super- numeraries YOUNGIA Long, or short in tuft- ed species, but in the latter not notably thickened, borne on a rootstock Absent Usually pinnatifid, the margin often revolute In tufted species bract- like or resembling reduced rosette leaves Few, very short or up to 4 or 2/3 the inner, mostly lanceolate or deltoid 6/12, mostly 8, lanceo- late, sometimes crested, carinate and thickened in fruit, not coales- cent at base Ligulate, yellow, some- times reddish on ligule, 5-22 mm. long, the tube mostly 1/5 to 1/3 the total length Echinolophate, the spines variable but smaller than in Soroseris. In Y. conjunctiva grains aver. 3ly in diam., spines ca. Sy Yellow or sometimes brown branches 0.5-2.5 mm. long, 0.05-0.1 mm. wide, the barbs ranging from very fine to rather prominent Fusiform or oblong, ob- compressed (sometimes strongly), with 3-5 stronger ribs: 2-11 mm. long, white, yellow, gray or fuscous; the setae rather coarse, stiff and brittle to fine or very fine and soft Advanced, no supernumer- aries observed in two species 4 MADRONO ; [Vol. 11 c Fie. 1. Youngia americana Babcock. From the type: a, rosette leaf, lower face, a’, lowest caudical leaf, a”, cataphyll, x 214; b, flower stem, x 2; c, head in anthesis, X2; d, e, inner involucral bracts from outer and inner series, outer face, x4; f, floret lacking ovary, x4; g, anther tube opened out, <8; h, detail of appendage, X32; i, marginal achene (immature, pappus removed), 8; j, inner achene with pappus, X 8. (Drawings by the author, inking by Miss Anna Hamilton.) 1951] BABCOCK: YOUNGIA oO of both Soroseris and Youngia. It will be noted that the dis- tinctive resemblances to the two genera are about equal in number, five for Soroseris and seven for Youngia; but that the Soroseris resem- blances are found mostly in vegetative features, whereas most of the resem- blances to Youngia occur in the pollen grains, the style and - stigma, the immature Cae: -_ achenes, the pappus, = and especially the yo : ovary anatomy. The YBNEB DE ALARA, < . 2 ks Cf | _ peculiarities of the LSE a are Se. styles and_ pollen ise Soars cogent DM | grains and the pres- ence of numerous su- Fic. 2. Youngia americana Babcock. Type tvande specimen in United States National Herbarium, Pe ay: Stet a folio no. 1973474. (Photograph by V. Duran.) in the ovary are con- sidered the most dis- tinctive features of Soroseris (Stebbins, 1940). Since it is in these parts of the plant, as well as in the achenes, that this new species shows definite resemblance to Youngia, its classification in the latter genus is clearly indicated. At the same time, the hollow caudex, the cataphylls on the rootstock below the rosette leaves, the elongated linear upper cauline leaves, the few, long, linear outer involucral bracts, and the long corolla tube, are not found in most species of Youngia, whereas they are characteristic of most Soroseris species. This occurrence of several marked resemblances to Soroseris appeals to the author as indicating affinity with that relict genus of the Sino-Himalayan region (Stebbins, 1940), thus endowing this plant with considerable phylogenetic significance. As in the case of Crepis pygmaea which, in certain features, exhibits resem- blance to Youngia depressa, while the latter species in turn shows relationship with the more primitive genera, Soroseris and Dubyaea, so in Y. americana we find additional evidence that Youngia and Soroseris had a common derivation. The ancestral group from which Youngia, Crepis, and the other genera of the Crepidinae were probably derived is the genus Dubyaea (Stebbins, 1940; Babcock, 1947). 6 MADRONO {Vol. 11 These clearly indicated phylogenetic relations have a direct bearing on the location of the center of origin of the Crepidinae. Extensive evidence has already been presented (cf. Babcock, 1947, Pt. I, Ch. 6, especially pp. 96-97) that both Soroseris and Dubyaea originated in central Asia and migrated southeast and south to their present areas of distribution. The discovery of this primitive species of Youngia in Alaska suggests a similar origin and migrational history to that of Crepis, section Iaeridopsis. In addition to this section, all of the other native American species of Crepis were unquestionably derived from species or hybrids that must have migrated across Beringea in Tertiary times. This evidence, together with the well justified assumption (Babcock, 1947, pp. 108, 137-139) that some of the Crepis species of northern Europe and of Iceland (C. paludosa) or their ancestors migrated westward south of the Ural Mountains in early Miocene, points to a northern central Asiatic origin of Crepis. Now, in Youngia americana, we find additional support for the concept of a northern Asiatic origin of the Crepidinae. Division of Genetics University of California, Berkeley LyrerRATuRE CITED Bascock, E. B. 1947. The genus Crepis. Univ. Calif. Publ. Bot. 21. Stespins, G. L., Jr. 1940. Studies in the Cichorieae: Dubyaea and Soroseris, endemics of the Sino-Himalayan Region. Bull. Torrey Club 19: 5-76. CYTOTAXONOMIC STUDIES IN THE GENUS SORGHUM. II. TWO NEW SPECIES FROM AUSTRALIA E. D. Garber ann L. A. SNYDER In a recent taxonomic revision of the genus Sorghum, Garber (1950) recognized six subgenera: Eu-Sorghum, Chaetosorghum, Heterosorghum, Sorghastrum, Para-Sorghum, and Stiposorghum. Among the subgenera, Para-Sorghum, and Stiposorghum form one circle of affinity and the other subgenera constitute a second circle of affinity. The basic chromosome number of Sorghum is 5 and is known only in Para-Sorghum and Stiposorghum. The application of cyto- logical methods to the taxonomic problems of this genus and especially of the subgenera Para-Sorghum and Stiposorghum has been eminently successful. A cytological study of a collection of Australian sorghums, using criteria previously validated in defining species of these subgenera, has revealed the presence of two undescribed species. MaTERIALS AND MeEruHops Herbarium specimens of Australian sorghums furnished by 1951] GARBER AND SNYDER: SORGHUM 7 Mr. S. T. Blake, Botanic Gardens, Brisbane, Queensland, yielded a small number of seed which were planted in Berkeley, Califor- nia. For cytological study, immature panicles were fixed in a fresh solution of absolute alcohol and glacial acetic acid (3:1) for 24 honrs and then transferred to 70 per cent alcohol for stor- age. Pollen mother cell smears were stained with acetocarmine. TAXONOMY Specimens number 17526 and 175386 of Blake’s collections were identified as Para-Sorghum and Stiposorghum, respectively, but could not be identified as any known species. An exami- nation of the meiotic chromosomes gave additional evidence of the distinctiveness of these collections. Sorghum australience sp. nov. Planta annua, culmorum nodis dense barbatis, ramis primariis verticillatis plerumque simplici- bus, pulvino parvo, spicularum sessilium callo obtusissimo spicu- lis sessilibus 7.0-8.5 m. longis, aristis 35-48 mm. longis, lodiculis glabris cupuliformibus crassis, spiculis pedicellatis masculinis vel neutris 6.5-8.0 mm. longis, fructu obovato. Chromosomae 2n = 20. Annual; nodes of the culms, at least the upper, bearded; pri- mary branches of the panicle whorled, usually simple; callus of the sessile spikelets obtuse; pulvinus not prominent; sessile spikelets 7.0-8.5 mm. long, brown; awns 35-48 mm. long; lodi- cules glabrous, cup-shaped, thick; pedicelled spikelets stami- nate or neuter, lacking lemmas, 6.5—8.0 mm. long; mature caryop- sis obovoid. Chromosomes 2n = 20. Type. East of Mataranka, near Elsey Station, Northern Territorv, Australia, April 29.1947, S. T. Blake 17526 (Herbarium, Botanic Gardens, Brisbane, Queensland). Sorghum australiense is the first annual Para-Sorghum found east of India. The other annual species in the subgenus, S. pur- pureo-sericeum and S. versicolor, are restricted to western India and eastern Africa. The glabrous, cup-shaped, thick lodicules of S. australiense are unique among the species of Para-Sorghum that have been examined. Pilger (1940), however, has placed S. trichocladum, a species with glabrous lodicules occurring in western Mexico and northern Guatemala, in the subgenus Para-Sorghum. Until living material of this species is available for cytotaxonomic study, this disposition cannot be verified. Within the subgenus, only S. versicolor has longer awns than S. australiense. Of the five species of Para-Sorghum, the three annual species have relatively long awns (27-54 mm.), and the two perennial species, short awns (12.5-24.0 mm.). | More than 200 spikelets of an open pollinated plant of S. australiense were examined and yielded no seed. Until extensive populations are available for hybridization, it is not possible to 8 MADRONO [Vol. 11 decide whether sterility is due to self-incompatibility as in S. leiocladum, a tetraploid species of Para-Sorghum, or to environ- mental factors. Using stainability with cotton blue in lactophenol as a criterion of pollen viability, approximately 80 per cent of the pollen grains appear to be functional. Sorghum matarankense sp. nov. Planta annua culmorum nodis dense barbatis, ramis primariis verticillatis plerumque sim- plicibus, pulvino parvo, spicularum sessilium callo acuminato 1.0 mm. longo, spiculis sessilibus 6.0—-6.5 mm. longis, lodiculis cilio- latis membranaceis, spiculis pedicellatis masculinis 6.0—7.0 mm. longis, fructu late subulato. Chromosomae 2n = 10. Annual; nodes of the culms, at least the upper, bearded; pri- mary branches of the panicle whorled, usually simple; pulvinus not prominent; callus of the sessile spikelets pointed, 1.0 mm. long; sessile spikelets 6.0-6.5 mm. long, brown; awns 35-50 mm. long; lodicules ciliate, membranaceous; pedicelled spikelets stam- inate, 6.0-7.0 mm. long; mature caryopsis broadly subulate. Chromosomes 2n = 10. Type. East of Mataranka, near Elsey Station, Northern Ter- ritory, Australia, April 29, 1947, S. T. Blake, 17536 (Herbarium, Botanic Gardens, Brisbane, Queensland). The short, pointed callus of the sessile spikelets in S. mata- rankense is also found in S. brevicallosum and serves to distinguish these species from the others in Stiposorghum. The shorter sessile spikelets, 6.0-6.5 mm. compared with 7.5—-8.5 mm., and longer awns, 40-50 mm. compared with 38-43 mm. distinguish S. mata- rankense from S. brevicallosum. The very short hairs of the bearded nodes in the former species contrasts with the long hairs of the bearded nodes in the latter species. The primary branches, es- pecially in the lowermost whorls, are usually divided, a character not yet observed in any species of Stiposorghum. Similar to the situation in Para-Sorghum, the one known perennial species in Stiposorghum, S. plumosum, has short awns (28-50 mm.) compared with the four known annual species (38-85 mm.). From more than 200 spikelets of an open pollinated plant of S. matarankense, no seed was obtained. Whether this sterility is due to self-incompatibility or environmental factors is not known. This observation, however, is not unexpected since the species of Stiposorghum are characteristically self-incompatible. CyToLoGy Sorghum australiense with a somatic chromosome number of 20 is a tetraploid. Since two plants had different numbers of rings of four chromosomes at diakinesis and metaphase I of meiosis, it is not possible to decide whether this species is an allotetraploid or autotetraploid. At any rate, the three Australian species of Para-Sorghum are tetraploids and the two African species, diploids. Since the three Australian species are either autotetraploid or allo- 1951] GARBER AND SNYDER: SORGHUM 9 tetraploid, there is reason to believe that additional collections of Para-Sorghum from Australia may yield diploid species. The meiotic chromosomes of S. australiense are similar in mor- phology to those of the other species of the subgenus at pachy- tene, diakinesis, and metaphase I. The chiasmata frequently at diakinesis and metaphase I was determined for one plant, A50/ 3-4. It had up to four rings of four chromosomes at these stages. Consequently, the data on chiasmata frequency had to be ex- pressed as the mean number of half chiasmata per chromosome. At diakinesis, the mean number of half chiasmata per chromosome was 1.90 (26 PMC’s) compared with 1.58—2.42 for the other spe- cies of Para-Sorghum, and at metaphase I, 1.79 (25 PMC’s), com- pared with 1.50—2.04 for the other species. Approximately 50 per cent of the pollen mother cells (26 of 50 PMC’s) had more than one nucleolus at pachytene; two pollen mother cells had three nucleoli at the same stage. The nucleolus organizing region was heteropycnotic; the nucleolus chromosome in cells with only one nucleolus was associated with the nucleolus at an intercalary position, thus distinguishing this species from the other two tetraploid species with terminal nucleolus organizing regions. Sorghum matarankense with a somatic chromosome number of 10 is the fourth diploid species of Stiposorghum. The meiotic chromosomes at pachytene, diakinesis, and metaphase I are simi- lar in morphology to the chromosomes of the other species of Stiposorghum at the same stages. The mean number of chiasma per chromosome at diakinesis was 0.61 (40 PMC’s) compared with 0.55—0.60 for the other diploid species, and at metaphase I, 0.60 (27 PMC’s), compared with 0.52—0.55 for the other diploid species. The percentage of bivalents with one chiasma at dia- kinesis was 78 per cent (40 PMC’s), and at metaphase I, 80 per cent (27 PMC’s) compared with 78.9-88.9 per cent and 89.0—96.0 per cent for the other diploid species at the respective stages. Two of the five bivalents in S. matarankense were associated with one nucleolus at pachytene, an observation typical of the diploid species of Stiposorghum. The position of the nucleolus organizing region was determined for each chromosome. In one chromosome, the nucleolus organizing region was terminal, and in the other, subterminal; the former chromosome was noticeably shorter than the latter at pachytene. The presence of a terminal nucleolus organizing region suffices to distinguish S. matarankense from the other diploid species of Stiposorghum in which none of the nucleolus organizing regions is terminal. SUMMARY Cytotaxonomic methods have been indispensable tools for ar- riving at conclusions as to taxonomic relationships and evalua- 10 MADRONO | (Vol. 11 tion within the complex genus Sorghum, especially in the sub- genera Para-Sorghum and Stiposorghum. The authors wish to acknowledge with gratitude the coopera- tion of Mr. S. T. Blake, Botanic Gardens, Brisbane, Queensland, who provided herbarium specimens of the Australian sorghums. We are also indebted to Dr. G. L. Stebbins, Jr., Division of Gen- etics, University of California, for field and greenhouse facilities. Division of Genetics, University of California, Berkeley, and Department of Plant Sciences, University of Oklahoma, Norman. LITERATURE CITED Garger, EK. D. 1950. Cytotaxonomic studies in the genus Sorghum. Univ. Calif. Publ. Bot. 23: 283-362. Pitcer, R. 1940. Gramineae, III: Unterfam. Panicoideae in Engler u. Prantl, Die naturlichen Pflanzenfamilien, 22, Bd. 14e. Engelmann. Leipzig. 208 pp. EVIDENCE FOR THE HYBRID NATURE OF X LIATRIS CREDITONENSIS L. O. GAISsER This paper’ is concerned with a plant described by the author (1946) as a putative hybrid between Liatris ligulistylis and L. squarrosa var. glabrata, under the name of,x L. creditonensis. Be- cause of the practical difficulty of controlled breeding of Com- positae, it has not been possible to demonstrate the hybrid nature of x L. creditonensis by planned resynthesis. The available evi- dence that it is a hybrid is here presented. Liatris ligulistylis (Nels.) K. Sch. has a known range of the three prairie provinces of western Canada and southward along the eastern side of the Rocky Mountains through western South Dakota, Wyoming, and Colorado into northern New Mexico. It favors comparatively moist habitats. Liatris squarrosa (L.) Michx. var. glabrata (Rydb.) Gaiser on the otherhand is found on the dry open plains from Kansas to South Dakota. The ranges of these two entities do not overlap. A garden plot of thirty-two plants of L. squarrosa var. glabrata (Accession No. 9) planted in 1928 from seed collected in Neb- raska (high cliff northwest of Royal, Antelope County, 4 October. 1927, Wernicke) developed into uniform, one-stemmed plants dur- ing their second year of growth, and by 19383 when they were five years old they had become several stemmed (PI. 1, fig. 1). Growing beside this plot were two plants of L. ligulistylis (Acces- 1 The author is indebted to Dr. R. Rollins and E. Anderson for reading the manuscript and to the latter also for suggesting Fig. 2. 1951] GAISER: LIATRIS 11 sion No. 1) which had been collected in Minnesota (near Erskine, Polk County, August, 1929, Gaiser). One was atypical in being more paniculate and in having more elongated basal peduncles than is usual in the species (PI. 1, fig. 2). Although stray seedlings were normally weeded out from the plots, one which was growing at the side of the bed of L. squarrosa var. glabrata and next to the two plants of L. ligulistylis was al- lowed to persist. In 1932 this plant (No. 52) bloomed and was seen to be intermediate between those two entities (Table 1), al- though the influence of L. squarrosa was more pronounced. This plant (Pl. 1, fig. 3) proved to be the supposed hybrid, x L. credi- tonensis. Because phyllaries show a wide variation in form and were considered important in distinguishing closely related species within a series, it was hoped that from the recombinations of their characters in this plant (No. 52), some knowledge of its parentage might be gained. In the phyllaries of plant No. 52, the charac- ters of L. ligulistylis and L. squarrosa var. glabrata were combined even more that it was possible to indicate in Table 1. CyTOLOGICAL OBSERVATIONS Chromosome counts from pollen mother cells of L. squarrosa var. glabrata (No. 9) were n= 10 (fig. 1 a-d) and from seedling root tips, 2n= 20. Meiotic figures (fig. 1, e, f) of L. ligulistylis (No. 1) appeared very similar to those of L. squarrosa var. glabrata. In the putative hybrid (No. 52) first metaphase plates showed ten chromosomes (fig. 1, g, h). A study of numerous lateral and polar first metaphase figures showed that there was always regular pairing of the bivalents. The anaphase and telophase proceeded regularly; at second metaphase (fig. 1, i, j), the uniformity of the chromosome plates was striking. There were no lagging chromosomes; second telophases and _ pollen tetrads were perfectly formed. Because of the similarity of the meiotic chromosomes of the two putative parental species, how- ever, the form and number of the chromosomes was of little help in determining the origin of seedling No. 52. PROGENY OF X L. CREDITONENSIS No. 52 The putative hybrid (No. 52) was partially fertile when open-pollinated. A number of achenes were obtained the first year that the plant was observed in bloom. Since the seeds were sown late in the season, many of the seedlings did not have time to develop a sufficiently thick corm to survive the winter. Thus, the fact that only fifteen strong seedlings were available for transplanting in the spring is not necessarily indicative of the vitality of the lot. In September 1934, the seedlings were in their second year and most of them had come into bloom. Two of them with 12 MADRONO [Vol. 11 narrower, more rigid, declined and somewhat twisted rather than broader upright leaves, and with projecting phyllaries were very | much like LZ. squarrosa. Three of them more closely resembled the parent plant No. 52 in their wider, somewhat erect leaves. How- ever, these three were not alike in their phyllaries, two of them again having the more outspread phyllaries like L. squarrosa and one having phyllaries similar to those of its parent. Thus, pro- geny of the putative hybrid showed various recombinations of the characters of L. squarrosa and L. ligulistylis. With regard to seven characters, the progeny were com- pared with the seed parent and with the two original species. Most of the progeny were nearly like the seed parent. Three were quite like L. squarrosa, and eight more approached L. squar- rosa in one character or another. None closely resembled L. ligulistylis. At various times during the study, voucher specimens for deposit in the Gray Herbarium were taken from x L. creditonen- sis, from the putative parents and from the various progeny. CYTOLOGICAL EXAMINATION OF PROGENY OF PLant No. 52 Meiosis was studied in the pollen mother cells of one re- sembling L. squarrosa and of another resembling the seed parent. In both, divisions were normal with a regular arrangement of ten bivalents at first metaphase (fig. 1, k, 1). Though many clear figures of all stages of the first and second divisions were seen, no lagging units on the spindles and no micronuclei were found. One count of eleven chromosomes on a second metaphase plate suggested irregularities (fig. 1, n) but was probably due to pre- cocious splitting of one chromosome or to pressure on the mount. Slight pressure on the coverslip over first metaphase plates caused a similar elongation of a few of the bivalents as they were seen somewhat laterally (fig. 1,m). Study of a third plant dur- ing two different seasons showed only mature pollen which was very regular in form. In 1934 seven of the progeny produced many good achenes, though one produced only six. From each of the seven, fifteen achenes were germinated for cytological examination of the root tips. The percentage of germination was good, varying from 40 to 86 in all except one which was not tested at the same time as the others and for which it was believed that temperature con- ditions were unfavorable. Of the plant which produced six achenes, all except one germinated. In 1935, achenes were col- EXPLANATION OF THE Ficures, PuLatTe 1. Pirate 1. SpecIMENS OF LiaTrRis GROWN IN GarRDEN. Fig. 1, Five year old seedling of L. squarrosa var. glabrata (No. 9 plant 32).—Fig. 2, LD. ligulistylis (No. 1 plant 2).—Fig. 3, x L. creditonensis (No. 52) photographed 1933. All approximately x 1/7. 1951 | GAISER: LIATRIS 13 SPECIMENS OF LiaTRIs GROWN IN GARDEN. Pruate 1. 14 MADRONO [Vol. 11 Le squarrosa vare Le ligulistylis XL. creditonensis ~~ glabrata No. 9 No. 1 Noe — a Corolla | Color** Mathews purple, 65 Lilac, 65d. Lilac, 65d. Length 1emm,. 7-9mm, ° Shape Erect to end of pappus Erect, tubular Erect, tubular and then flattened (rotate) eed Lobes Hairy Glabrous Slightly hairy Style Length 20mm. 1mm. 18mn. Color*** Mauvette, 65f. Lilac, 65d. Mauvette, O5f. Pappus Length 7mm. 7-8mm. 8mm. Nature Plumose Barbellate Barbellate Achene Length 6mm. h-5mm. 6mm. Phyllaries Color All green Green with reddish tips All green Nature Glabrous, chaffy Glabrous, membranous Glabrous, not quite as chaffy as #9 Margin No cilia, not scarious Markedly scarious No cilia, not scarious and lacerate 5 Position Outspread Erect, very slightly Mostly slightly incurved incurved Shape of Sharply acuminate Broadly ovate and Elliptical acute, outer ones short () not sharply acute Size of 12mm. long h-5mm,. long 10-linmlong, outer ones 3-hmm. wide Shape of Linear acute Broadly Linear, spatulate inner ones spatulate Length of 15mm, long 8mm. long 15mm, long inner ones Peduncles h-8mm. long himm,-12cem. 20-25mm, long (of basal flower) * These observations were made at the time the styles were exserted, August 1933, with revision of pappus and achene lengths to those of mature conditions. ** Colors are according to Ridgway’s Color Chart, plate XXV. lected from eleven plants, of which five were other than those checked in the previous season. Though their viability was not tested, it is probable from the appearance of the well-filled achenes that their germination would have been as high. The somatic chromosome number was confirmed as twenty in six seedlings. Wel ae | In numerous figures obtained from one seedling, at least two pairs of chromosomes were found to be longer than the rest, one pair long-median, and one pair sub-median. For the rest of the chromosomes, it was difficult to draw a line between those of intermediate length and the shortest ones. FurtTHER ANALYSES OF THE PROGENY Using the hybrid index method of Anderson (1949), x L. creditonensis was compared with its putative parents and with its 1951] GAISER: LIATRIS 15 e®@ e@@ S36 2 —eSe, ste oF ag ee 3 oS . d s b e @ a @ yd bd ° Ose one oSee i ©) e° ° E g h f See ® ofce Coe one cue? ee eee a” i j k (x) sege ee cm) a ae 6 Be Co @ “ee? *3s oe m n Fic. 1. Meiotic chromosomes of X JL. creditonensis and the two putative parents as seen in polar views of first metaphase (except i, j, n which are second metaphase) in pollen-mother-cells from aceto-carmine preparations, drawn by Zeiss apochromatic lenses and camera lucida at aproximately <1900. Fics. a-d, L. squarrosa var. glabrata No. 9 (n=10): a, b, two young seedlings in their first summer of bloom; c, d, seedling 32 in its fifth year. Fas. e, f, DL. ligulistylis No.1 plant 1 (n=10). Fuias. g—j, x L. creditonensis No. 52 (n=10): g, drawn in 1932; h-j drawn in 1933. Fic. k, plant 8 of progeny of No. 52. Fies. 1, n, plant 12 of progeny of No. 52. progeny, nine diagnostic and measurable characters being studied from the herbarium material—pilosity of corolla limb; nature of the pappus (length of barbules); margin, shape, and position of phyllaries; shape, surface, and margin of leaves; and pubescence of upper stem (Table 3). As to the first character considered, pilosity of the corolla limb, L. squarrosa has hairy corolla lobes, whereas L. ligul- istylis has a completely clear corolla, lacking any pilosity within the limb, throat, or tube. This is in contrast not only to L. squar- rosa, but also to other species of the Scariosae which have hairs 16 MADRONO [Vol. 11 within the throat. (In referring to parts of the corolla, the tube proper will here be regarded as the part below the insertion of the filaments, the part above that being referred to as the throat and limb, of which the latter refers to the teeth only.) In determining the hybrid index, L. squarrosa, with profuse hairs, was scored 4, x L. creditonensis, with few hairs, 1, and L. ligu- listylis, with no hairs, 0. The progeny was scored from 1 to 4. The second character considered was nature of the pappus. L. squarrosa has a very plumose pappus, whereas in L. ligulistylis the pappus is barbellate and to the naked eye not obviously plu- mose. As can be seen in those species of Liatris having a plumose pappus, the barbules toward the base of the seta are generally shorter. Also, as was noticed early by Cassini (1827) in Suprago, even along the mid portion of the seta there are dif- ferences in length of the barbules. Inasmuch as only the longest projecting barbules are readily distinguishable when setae are flattened under a cover slip, measurements were made only of these longer barbules. The figures, therefore, do not represent the full range of length. For each species, measurements were made of ten specimens from widely distributed localities. These included the isotypes and two plants of each of the accessions studied above. For each of these plants, twenty-five barbules were measured. In making these preparations, it was found that lactic acid was a satisfactory mounting medium for the short-barbuled setae of L. ligulistylis, but for the plumose setae of L. squarrosa parlodion proved to be better. Table 2 gives the means and standard errors of these measurements. Plant No. 52 falls between the two species, just at the upper limits of L. ligulistylis. This confirms the macroscopic determination of its pappus as barbellate (Gaiser, 1946). Only one of the pro- geny had a mean lower than that of the parent plants, and the rest fell between it and L. squarrosa. In determining the hybrid index, the barbule length of L. ligulistylis was scored as 0 and that of L. squarrosa as 2. Intermediate lengths were then scored fractionally. For the other eight measurable and diagnostic characters studied, the condition of L. ligulistylis was scored 0, the inter- mediate condition 1, and the condition of L. squarrosa 2, except for pilosity of corolla limb which was scored from 0 to 4. Seor- ing and totals are shown in Table 3. Thus, with the putative parents having a total score of 0 and 20, plant No. 52 receives only a total of 10.44. The scores of the progeny fell between those of L. squarrosa and x L. creditonensis, the highest nearly ap- proaching that of L. squarrosa and the lowest that of x L. credi- tonensis, none approaching closely to L. ligulistylis. This confirms the observations on the living plants, most being like the parent plant or intermediate between the latter and L. squarrosa. In- dependent segregation is shown, however, in some of the charac- 1951] GAISER: LIATRIS 17 Bal 2.0 1.9 1.8 i eT 1.6 1.5 o 1.4 a 2.3 ‘ Pr? = Sahl 620 ae > 8 Bae 27 .6 5 .4 3 2 sit 0 0 1 2 3 4 Pilosity on Corolla Limb Fig. 2. Scatter diagram of progeny of x L. creditonensis (Plant No. 52): L. ligulistylis (black circle in lower left); L. squarrosa (black circle in upper right); L. creditonensis (stippled circle); the progeny (clear circles). From left to right the symbol arms represent the phyllary, leaf and stem characters in Table 3, long arms a score of 2, and half length, a score of 1. 18 ~ MADRONO [Vol. 11 ters. In the pictorialized diagram (fig. 2), the diversity of the progeny is at once apparent. On the basis of the characters scored, no two of them are exactly alike. Discussion With the common observation of the intermediacy of an F, between the two parents, it is sometimes less well remembered how greatly the progeny of even an F, vary from one another. However, as Anderson (1949) has recently emphasized, the number of recombinations in an F, is not numerous considering the total possibilities in the multitudinous characters of: two species. Because of the linkage of genes, as at least one limit- ing factor, fewer types of combination are achieved. Also as he has pointed out from the analysis of hybrid populations under natural conditions, there is a repeated backcrossing of the hy- brids to one or both of the parents, and with each successive generation the hybrid nature is less apparent. The result is that a few resemble the recurrent parent and a few the F,, but a larger proportion will be intermediate between these two. In Plant 52 (x Liatris creditonensis) and its progeny, there was shown a change from the distinct intermediate plant No. 52 to some of its progeny. Because of the failure of some of the insufficiently developed seedlings to over-winter, the complete population could not.be represented in the study. The fifteen progeny which were grown varied so slightly among themselves that when for lack of better characterization they were divided into seven categories, two were described as practically like the parent plant and eight were considered close to it but with other combinations of characters from the one contributing parent spe- cies L. squarrosa. While two categories were recorded at that time based on close similarity to L. squarrosa, none pointed singly to L. ligulistylis. The more extended analysis (Table 3), including measurements of two characters of a quantitative nature, con- firm this. Though the total scores of the progeny range from those very nearly as low as the parent plant No. 52 to those very nearly as high as L. squarrosa, none at all had a score approach- ing L. ligulistylis. These plants seem to meet the situation des- cribed by Anderson of a backcross to the recurrent parent with L. squarrosa as that parent.