hy Ae Nt SHALE SNR 4 ‘ . aN Nant a ue Rin AAAS itp poe a zs ey \ om c New Pork State College of Agriculture At Gornell Mniversity Bthaca, N. ¥. a Library ‘ornell University Libra Studies on the genus Aquilegia. Studies on the Genus Aguilegia by illan Cameron Fraser 1918 Agknowledgment The writer wishes to express his grateful ap- preciation to Professor A. C. Beal of the Departnent of Floriculture and to Professor W. W. Rowlee of the Depart- ment of Arts Botany for assistance rendered him in the dourse of these studies. Professor Beal offered help- ful criticivms and suggestions through the course of the work and placed at the disposal of the writer, all of the land necessary for the growing of the Aquilegia plants. Professor Rowlee also wade a number of kelpful suggestions and aided further by supplying: herbarium material of na- tive and foreign species. Studies on the genus 4quilegia The genus Aquilegia contains many interesting and valusble species ef plants. The taxonomic botaniet finds intereeting problems in the classification of its many types. The horticulturist draws from it a wealth of material for his boréer plantings and his rock gardens, and encounters many preblems in adaptation and culture. and the plant breeder finds a large number of genetic cheracters on which to study inheritance. Up to the present time a certain smovnt of work has been done on the genus, but the subject has never been exhausted in any of its phases. Furthermore, the litereture on Aguilegia is scattered and no attempt has been made to review it in one paper. The studies herewith presented are not to be ae- cepted as complete in every detail. The genus is too large and its members too scattered to permit of an exhanstive etudy within a limited space of time. in attemot hos been made, hovever, to present stch informaticn ae conld be gath= ered from the literature, and to verity it and cupplement it a6 much e& poesible by studies on the living plants. Purpose ot this study The attention of the writer was drawn to the genus Ayullegia in the spring of 1913, when he was working at the New York Botanic Garden. The large collection of growing plants and the unusually good herbarium at the Botanic Garden suggested a taxonomic study of the genus. A number of crosses were made between the most etriking types, end plants of each rpecies were bagged for pure cred. At the present time the studies are not completed. It will take ecme years of in- vectigetion before a number of the minor problems which have arisen vill te settled. ‘This report is fairly complete, howe ever, as Par as it. attempts to go. The purpose of these studies has been several-fold. In the first place it was intended that a complete collection of species descriptions shotld be made and that so far es poseible the validity of the various epecies should be de- termined. Tiere would oturully follow the working out of the Syaonomy cf Ayuilegia and the coustruction of a key for the identification of unknown species. Notes upon the cul- ture of the different species and upon the horticultural uses of the columbines huve also been added. Methods of study Wherever it was poSbible, the ctedies ot the species were made upon the living plants. Luch observations could be made, of course, on all of the epecies and varieties of hort- icultural importance. There were many rare species, however, whose seed could not be obtained. In such cases the publish- ead descriptions had to be used entirely. In still other cases no information other than the name of the species could be obtained. Herbarium specimens were studied, when they could be obtained. The collections at the New York Botanic Gar- den and at Cornell University were found very helpful. In the case of each species an effort was made to get all the information possible, first by studying the growing plants, then by examining herbarium specimens, and thirdly by look- ing over published accounts of the species. The Family Ranunculaceae The genus Aquilegia belongs to the family Ranuncu- lacese of the order Ranales. Wiegand* has given the follow- ing description of this family. "Ranunculaceae - Flowers spirally constructed, ex- cept often the perianth, regular or irregular; sepals 3 to many, usually 5, separate, often petalloid; petals 3 to many or 0, often in the form of honey glands; stamens usually very numerous, hypogynous; carpele 1 to many, usually separate; fruit an achene or follicle, rarely a capsule or berry; seeds with endosperm." There are about 27 genera in the family and about 680 species, most of which are found in the north temperate and sub-arctic regions. Clematis, Anemone and Delphinium cross the equator at the south. Ranunculus is the largest genus. The family Ranunculaceae is related to Magnoliaceae, Annonaceae,, "Wiegand, K. C. - a synopsis of the plant kingdom. In Bailey's Enc. Hort. Vol. 1, p. 32 4 Dileniaceae, Nymphaeceae,and other families with acyclic flowers and numerous carpels. fhe spiral floral structure, numerous hypogynous stamens, and the usually separate carpels are the moet dis- tinctive characteristics cf the family. The family Kanunce- laceae is probably an old one. It is thought by some to be the closest to the stock from which the dicotyledons have sprung. Within the different genera of the family, the flower structure is quite variable. The petals, when present, gre rarely normal. IH Aconitum, Delphinium and Aquilegia the petals are spurred and nectar-bearing. Thalictrum is wind- pollinated. Prenti makes 3 tribes of the family: Tribe 1 - Fruit follicular, carpels fleshy, cuter seed-coat long, eg@. Paeonia and Hydrastis, Tribe 2 - Fruit usually follicular, carpels rarely fleshy, outer seed-coat not larger than inner, eg. Caltha Helleborcus, Coptis, Actaes., Aquilegie, Delphinium, etc. Tribe 3 ~- Fruit an achene, eg. Anemone, Clematis, Ranunculus, Thalictrum, etc. There are many useful plants in this series, most of them ornamentals, snd a number drug plants. Among the general which have cultivated forms are: - Aconitum------ ween wenn --monkshood Actaea----------------- --baneberry Adonis-----<----------- ~--adonis Anemone--------------- ~--windflower Aquilegia------------- ~--columbine Caltha--------------- ----marsh marigold Clematis-------------- ~---virgin's bower Coptis--------<---------- gold thread Del phinium------------- --larkspur Kranthis---------------- -winter aconite Helleborus--------------- Christmas rose Hydras tis--~-------------- golden seal Nigella------------- -----love-in-a-mist Paeonia------------------ peony Ranunculus------------=- --~-buttercup Thalictrum----------- ~---meadow rue Trautvetteria------- ~----false bugbane Trollius--------~+--- -----globe flower Xanthorrhiza--------- ~---shrub yellow root * Note The genus Aguilegia Tournefort - (1719) is given credit for the first comprehensive description of the genus Aguilegia. Prior to the publication of his Institutiones rei herbariae, in * Family description from Wiegand, K. C. —Lac._o4t, which this description appeared, Dodeene (1619), Gerarde (1633), Bauhin (1651), and others had published fragmentary descriptions of the genus and of such types as were known to them. Tournefort'e characterization of the genus is by no means a8 complete as many published after his time, but it is remarkably accurate and scientific for a brief descrip- tion. Plate I is a photograph of Tab. 242, from Tournefort's books. The species shown is undoubtedly vulgaris, a spurless double type of which appears at M. It will be noted that fTournefort has shown six follicles instead of five, in the fruit. This variation in number quite frequently occurs, although six is not the typical number for the genus. Aside from this minor defect, the plate serves admirably to il- lustrate the escential characteristics of the genus. One of the most complete descriptions of Aquilegia is that published by Asa Gray in his flenera of the plants of the United states (1849). This description follows; "Calyx petaloid, regula%, imbricated in aestivation; sepals 5, spreading, ovate or oblong, deciduous. Petals 6, all similar in size and shape, Kypogynous, inserted by the inner margin or lip of the eblique expanded limb, alternate with the cepale, and produced backverde between them into a long infundibuliform-tubular spur. stamens indefinite, hy- pogynous, obscurely collected into 5-10 clusters, filaments filiform and elongated; some of the innermost abortive, desti- Plate 1 - Showing the parts ot the colum- bine flower and fruit. From tournetort, = Institutiones rei herbariae, = 1719. tute of anthers, and converted into membranaceous scales which collectively surround the ovaries; anthers oval, inmate, the cells openins lengthwise laterally. Pistile usually fine, sessile, erect or oppressed; ovary cylindraceous, one-called, many cvvled; style Z@iliform, elongated, stigma milaterial at the apex. Ovules indefinite, occupying the whole length of the ventral suture in two series, horizontal, anatropous, the raphes collateral. "Follicles usually 5, sessile, erect and oppressed, nearly cylindrical, veiny, tipped with the slender persistent styles, dehiscent through the ventral suture from the apex downward. ‘Seeds numerous, in two series, horizontal, small, oval, smooth and shining; the crustaceous testa conformed to the nucleus. Embryo minute, at the base of the corneous- fleshy albumen, cordate; the radicle next the hilum. "Herbs, with erect and usually paniculate-branching stems, from thickened and fibrose perennial roots; the alter- nate leaves bi~triternately compoun@; and the large and showy flowers singly terminating the stem and branches, Lower leaves long-petioled; the uppermost subsessile, or reduced to bracts. Petioles more or less dilated at the base. Leaflets roundish, incicsely lobed, glauccus underneath." fhe flower parts in Aquilegia are all cyclic. In this respect it represents an advance over the genus Ranunculus, where the stamens and carpels have a spiral arrangement. ‘The parts of the perianth are frees, and the gynoeceum is apo- esrpous. The floral formule for the genus is: K-5; C-5; a-5 x (8-12), G-5. The following flower-diagram from Goebel (1867), shows the arrangement of parte in aquilegia. 10 The name Aquilegia There seems to be s little doubt as to the deriva- tion of the word Aquilegia. ‘ome writers believe it to have come from aquila, meaning eagle, from the fancied resemblance of the spurs to the beak and talons of an eagle. Still oth- ers maintain that it comes from aqua - water, and legio - to draw, from the supposed great needs of the plant tor water. The latter explanation sounds more probable. The common name, columbine, is generally agreed to come from the resemblance of the spurs to beaks of doves, or even to the whole bird. It comes from the Latin, columba - a dove. fhe species of Aquilegia When one attempts to list the species of Aquilegia of recognized standing, one is confronted with a number of difficulties. Aquilegia, like many other genera, contains a@ number of species of doubtful validity. It is very hard to decide which species should be inclyded and which omitted from the list. Those species which are commonly used in horticulture are well established and present no difficulties. There are many types from India and the Himalayas, however, and local forms from central iurope and from this country which cannot be as readily accepted. If it were possible to get self-fertilized seed of all the doubtful types and to raise plants of each, it would be a comparatively easy matter to pick out the dis- ll tinct types. the change in environmental conditions would undoubtedly bring sbout changes in the appearance of the plant, but hardly changes so grest as to cause a blending of types. In these studies an effort was made to obtsin seed of all the different species of Aquilegia. ceedsmen in this country and in imrope were able to supply seed of the common- ly grown types. The Royal Horticultural Society of Englani furnished a mixed sample of seed, saying that it was all they could send, as the war had necessitated giving up most of their gardens. liany interesting types were obtained from this lot of seed, but the plants were hardly useful as a basis for accurate description and classification. Between 60 and 80 packets of seed, representing almost as mahy named types, ware obtained through the Bureau of Plant aud Seed Introduction at Washington, by the courtesy of Mr. Dorsett and Mr. Bissett. These secés were sent from botanic gar- dens at Madrid, spain; St. Petersburg, Russia; Geneva, Swits- erland; Dublin, Ireland: Bdinburgh, vcotland; and Stockhoin, eweden; as well as from seed houfes of various countries. Many of the plants from these secds bred true to tyze ani were described. A large number, however, hed been open-pol- linated and the resulting hybrids came far from agreeing vith the descripticns of the species which they were supposed to represent. An appesl for seed was aleo made to botanists 12 and experiment station workers in the western part of this ecuntry. Tt was hoved tn thic way to get seed of the rarer enecies. Only the seed of the commoner sorts could be sup- plied, however. The botanists who had desoribed certain of the rarer species in this country, wrote that it would be im- possible to obtain seed of these except by sending a sollect- or to the isolated canons or mountain sides where they grow. Thus while many of the species descriptions were taken from the living plants, as well as from the literature, a large number had te stend exactly as they came from boetani- oal publications. Rerarding other species little could be learned except the name. These were described in Hungarian and Russian periodicals which are not accessible in this coun- try - The question which naturally arises in connection with the less common forme is this: - 4re these types dis- tinct botanical species, or are they merely races or perhaps local variations of well recognized species’ Undoubtedly some of the named species are of the latter types. Another question which may well be asked regarding any species of uncertain rank is this: - Has this species been described before under another name, and if so, thich name has the priority? In other words, we are confronted with certain probleme in synonomy. There is hardly a type @&f colum- bine which has not been described under several species names. 13 In the case of Aquilegzia wilgaris, at least 40 other specific names have been spplied. (Index Kewensis). It is ouite evident that any lict of anecies of agq- uilesia is likely to be incomplets in some respects, and also is likely tc contain the names of some species of doudtful rank. Britton (Illus. Fl. Vol. 2, p. 92) places the number of species of columbines at about fifty, The Index Kewensis lists cver ninety svecies, exclusive cf snynonyma. Over one hundred bowsuicsl types have been cescribed. The following liet of species, which has been conm- piled after a careful etudy of the available literature, is felt to be resseorably eomplete end uccurate. Species of Acuilegia *a, alpina, Linn. A. baikalensis, Hort. A. bertoloni, it chott. A. brevietyls, Hook. “A. Duergeriana, Sibb. and duce. “A. caerulea, James. *A. canadensis, isinn. *A. ehrygantha, Gray. A. Coveinea, Siall. A. ecolumbiana, Rydb. A. @calenrata, Rastwood. A. @insenleana, Fr. Schults. 4. @Qlegartuls, Creens. he Llabellata, Sieb. and Zuec. A. flavescens, 5. Wats, *aA. formoss, Fisch. A. fregecne, Benth. *a,. glandulosa, Fisch. A. glauea, Lindi. A. haylodgensis, Hort. *a, helenae, Hort. &.- donesii, Parry 4. kitaibelii, wchott. ase: 14 lactiflora, Kar. and Kir. Laranioneic, Jelson. leptoceras, Fisch. and Mey. lengi 6slacd ‘« G aay r Miadrensis, Robe. micrantha, Has twooa. mooreroftians, all. orecphila, Rydb. oxysepala, Trautv. and ley. parvifiore, Ledeb. pubescens, Coville. enhiziere, beadl. pyrenaicea, D. C. saxitiontane, Rydb. sibirica, Lam. Skinneri, Hook. Stuerti, Hook. tovueus., Pheths wid ey. viridifilora, Pallas. vulgaric, Linn. She better kmown cpecies eve uarked with * A number of Indian and Hungarian species have been omitted from tiis list, beeause of = laek oF imicrmation regarding them. al pal fal bail 15 Aquilegia Keys One of the first keys to the genus Aquilegia was prepared by J. G. Baker and published in The Gardeners’ Chronicle, 1878, p. 203. This key is of interest because it shows certain group relationships between species. As an aid in identifying unknown species of columbines, how- ever, it fails to serve its purpose. The headings for cer- tain groups are altogether too general, and no clear dis- tinctions are made between the members of any group. "Systematic Key to the Known Species of Aquilegia" - - Baker. Group 1 - Micranthae - Small flowered columbines. Stem leaves all small and bract-like. A. @inseleana, A. viscosa A. thalictrifolia, A. parviflora. Lower stem leaves large petioled and biternete. Lamina of petals much shorter than sepals. A. lactiflora, A. pubiflora. lamina of petals nearly or quite as long as sepals. Spur short and curved. A. brevistyle. Spur moderately long and hearly straight. 4e viridiflora, A. buergeriana. Ae Ganadensis, A. flavescens. Group 2 - Mesanthae - Middle sized flowered columbines. Stem leaves all small and bract-like,. 16 Ae pyrenaica, A. bertoloni. Lower stem leaves large petioled and biternate,. Spur short A. @lauca, A. flabellata. Spur about as long as the lamina of the petal. Lilac-claret-white flowers; short stamens. Ae moorcroftiana, A. amaliae A. leptoceras, 4. vulgaris A. Sibierica Red yellow flowers; long stamens. A. formosa, A. Skinneri Spur very long. A. chrysantha. Group 3 - Macranthee - Large flowered columbines. Spur short - A. glandulosa. Spur moderately long - A. alpina, A. fragrans. Spur very long - A. caerulea. Rydberg, (1905), presents a key for the determina- tion of the Colorado species. This key is good as far as it goes, but it only includes seven Species. Conliter and Nelson, (p. 191), give a key to the Rocky Mountain species, nine in all, This key appears to be fairly complete for the territory which it covers. It is al- s0 easy to follow. 17 Coulter and Nelson's Key to Rocky Mountain Columbines Spur longer than the calyx. Flowers red in whole or in part-------l. A. elegantula Flowers never red. Flowers blue or white ----------- 2. Ae Gaerulea. Flowers yellow or bluish - Spurs long (5-7 cm.) yellow------3. A. chrysantha. Spurs shorter (4 cm, or less) bluish-----4. A. oreophiia. Spure shorter than the calyx. Stem leafy or not wholly naked-scapose; flowers several. Flowers yellow or yellowish. Plant tall (more than 2 dm.) Spur as long as lamina---------- -5. A. flavescens. Spur nearly wanting--------- -----6. A. Eastwoodiae. Plant lower (less than 2 dm. high) 7. A. laramiensis. Flowers blue--~---~---- we eee ene ~--8. A. saximontana Stems scapose---~------- weno nsnn- 9. A. Jonesii. The most inclusive key to the Aquilegias which has been prepared thus far is one which was compiled from the partial key and the descriptions by bavis, K. C. - In Bailey's Cyclopedia of Horticulture. This key has the advantage of dealing almost entirely with the flower parts and not with the lees definite leaf and plant characters. 18 A. - Sepals not more than 1/2 - 3/4 in. long; expanded flowers 1 - i-1/2 in.diam. B. - Limb of petals shorter than sepal. C. - Stem very short. D. - Pollicles glabrous - A. Jonesii. CC. -Stem over 1' high. De~ Spur knobbed and bent. - A. oxysepala. DD.- Spur not knobved and nearly straight. A. lactiflora. BB. - Limb of petal about equal to sepal. C. - Spur straight or nearly so. D. - Spur 1/2" long. Stamens protruding beyond the petal-limb - A. viridiflora. DD.- Spur 3/4" long. E. »tamens equal to petal-limb - &. buergeriana. Ba. - stamens much protruding - A. canadensis AA. - Sepals about 1" long; expanded flower about 2" diam. B. - Spurs shorter than petal-limb and incurved - Ae flabellata. BB.- Spurs at least as long as the petal-limb. C. - otamens short; net much protruding. D. - Spurs slender and only slightly curved - 4s leptoceras. DD.- Spur stout and much curved. E. - Follicles densely pubescent - A. vulgaris. 19 ER. - Follicles glabrous. Flowers pale or bright lilac-blue - A. Sibirica. CC. - Stamens long, protruding far beyond the petal-limb. D. - Spurs spreading somewhat - A. formosa. DDe- Spurs not spreading - A. Skinneri. BBB. - Spurs very long, several timee the length of the petal-limb. C. - Orifice of the spur not much contracted; early flowering - A. chrysantha. CC.- Orifice of the spur contracted; late flowering Ae longissina. AAA. ~ Sepals 1-1/4 - 1-1/2 or even 2” long; expanded flower 2-1/2 - 3" in diameter; stamens not protruding. B. - Spurs long and not incurved - A. caerulea. BB. -vpurs incurved and only a little longer than the petal-limb. C. - Plant not over 1' - A. alpina. CC.- Plant 1-1/2' or thereabouts - ue €landulosa. The following key hes been prepared by the writer from a study of the living plants, where possible, and also with the aid of herbarium material and published descriptions. an effort has been made to confine the attention to the char- acteristics of the flower parts, eince these are usually the most constant and most sharply defined parts of the plant. Fruit and leaf characters as well as stature have been brought z0 into the key where it was felt that a sharper distinction could be made between species on the basis of these parts. This key may be found useful when applied to fairly typical plants of any given species. rida, Hort. Flowers large with yellow petals and scarlet sep- als, spurs spreading, long and slender. A enpposed natural hybrid with a. chrysantha. Piant about 3 dm. high. Sstem- leaves nearly sessile, glaucous below. Flowers 5.7 cm. broad x 5.5 cm. long. Spurs 3 cm. long, scarlet, straight, slight- ly knobbed. Lamina 1 com. long, bright yellow. Sepals 2.9 cm. X 1.1 cm., scarlet to Brazil red. stamene exserted by 1.3 cm. styles by 1.4 cm. 16. - &. fragrans, Benth. Plant 1-1/2 to 2 feet; flowers few; stems more slender than in A. vulgaris, finely pubescent above. Leaves with partial petioles 1 to 2 inch long, with 3 distinctly stalked divisions, as thin as in a. alpina, slightly glev- cous below. Stem-leaves biternate, lower petioled. Flowers white or pale claret-purple, very fragrant. Sepals 1-1/4 to 1-1/2 inch long, half as broad, ovate, sub- acute, not reflexing. Petal-limbs 3/4 inch broad, very cune- 53 ate and rounded. pur slender, slightly curved, barely knob- bed, short. ‘Stamens not exserted. styles exserted by 1/3 inch. Follicles pubescent, 3/4 to 1 inch long. Range - Temperate region of western Himalayas, Kashmir and Kistwar, 8,000 to 10,000 feet. Introduced into Englend about 1840 by the Royal Horticultural Society. Refsrences: Benth, in Maund Bot. Vol. IV. $. 161. Baker, J. G@. - 1878. p. 203. 17. - As glandulosa, Fisch. Plant 1 to 1-1/2 feet high. stems glandular pub- bescent above. Partial petioles of root-leaves 1 to 2 inches long with 4 distinctly stalked divisions, ultimate segments narrower and deeper than in 4. vulgaris, slightly glaucons below. Stem-leaves few and bract-like. Flowere large and hodding. ‘Sepals 1-1/2 inches long, ovate, unguiculate, a- cute, fully half as broad as long, bright lilac-blue. Pet- als less than half as long ss sepals, cuneate, rounded at top, as broad as long, lilac-blue, tipped with creamy white. Spurs 1/4 to 1/3 inch long, stout, incurved. Stamens not exserted beyond lamina. iStyles falcate, divergent. Fol- licles 1 inch long, densely pilose. Season - Early in summer. Range - Altai mountains and other districts of Central Siberia along edges of mountain streams. 54 Allied to 4. alpina, but taller, with larger flow- ers and shorter spurs. Very handsome. A. Whitmanni,Hort. is somewhat like this species. &e glandulosa is excellent for rock gardens. The Plants often die after the first year and should therefore. be treated as biennials. They do best in a rich, mellow soil which is cool and moist. English horticulturists have found it very difficult to raise this plant with any success. It does very well in Highlands of :cotland, however. References: Fisch - Hort. Gorenk. ed. II. 1812 :48 DeCandolle - Prodr. Vol. 1. pe. 50. Davis, K. ©. = In Bailey's Cyc. Hort. Vol. 1. Douglas, J. - The Garden 16 : 264 (1879) Dod, C. Wolley - The Garden 17 : 76 (1880) Robinson, i. - 1870. p. 147. Baker, J. Ge - 1878 : 203. Anon. - The Garden 9 : 3584. lva. - As glandulosa var. juncunda. Flowere smaller than in type, with white petal- limbs which are more truncate than in the type. keference: Baker - 1878. p. 2803. Davis - In Bailey's Cyc. Hort. 55 18. - Ae glauca, Lindl. Plant 1 to 1-1/2 feet high, glaucone, glabrous be- low and finely :ubescent above. Partial petiolee of root- leaves 1-1-1/2 inches long with distinctly stalked segments, broad, shallow lobes, thin texture, hoth surfaces glaucous and glabrous. Lover stem-lsaves biternate. Plowers fragrant. white tinted with claret. Sepals l inch long, not reflexing, evate, sub-obtuse, 1/2 inch broad. Petal-liub cuneate-truneate, 3/8 to 2/4 inch long and near- ly aS broad. Spur 1/4 to 1/3 inch lons, straight or nearly straight or very Slirhtly curved. oStamene as long as the lam- ina. Follicles pubescent; styles recurved, nearly as long as follicles. Range - Temnerate region of western Himalayas. Introduced into cultivation by the “ast India Company in 1846. References: Lindl - In Bot. Reg. 1846. tab. 46. Baker - 1876 - >. 76. 19. - As haylodcensis, Fort. woo, ~The Garden XIII, 1878, p. 647, believes it i@entical with A. ecaernles var. hrtride and of the same par- entage, A. caerniea x A. chrysantha. The plants of this variety which were examined by the writer proved to be a badly mixed lot with no distinct 56 type prevailines. 194. - a, haylodgensis var. delicatineima, Hort. "A new hybrid, bearing larze, Lone-spnrred flowers, the spurs being a tender satin-rose, passing imperceptibly to a delicate licht yellow at the crown." Dreer - catalogue 1915. 20. - A.» helena, Hort. "A etrene robust grower, 15 to £0 inches high, with numerous very larre bine flowers and wide, expand- ed pure white corolla. ctinilar to caervilea, but with shorter spurs." Dreer - catalorue 1916. Flowers 3.4 om. wide x &.4 om. long, stout. Sepals 2.20 x .85 cm., dark pruae purple. Fetais 1,25 em., whitish. Spurs 1.16 om. long, mulberry purple, curr ed, scarcely if any knobbed. gl. - is doneeii, erry. Plant about 5 inches high; stems minetely soft pubescent. foot-leavee tufted; partial petioles very short, if any, biternately divided, leatiets very crowded. sepals oblong, blue. Ffetais half as lone ss sepals, blue. Spurs shorter than sepals, nearly ¢traight, gradually tapering. Styles exsertec beyond petel-linb. Foilicles turgia, smooth, nearly 1 inch long. 57 revson - June ana duly. Taupe - Phiox meuntsin, «wyoring, and in Montana. A very dwarf slant. Dedicated to Captain We A. vones, U. S. fasineer, who was commander of the expedition on which this species was collected. An alpine plant found om the high peaks of the central Rocky mountains. References: Parry, ©. C. ~ Amer. Nat. 8 : 211 - 1874. Davis, u. C. - In Bailey's Cyc. Hort. Vol. l. Rydbovr, P. .. ~ Vol. ITI ~ 1696 ~ p. 482. Coulter and Nelson ~ 1909. ,. 192, eee - ae Kitaibelii, seprott. Plant exert, 1 foot or less, slender. stems short, slightly viscose, pubescent. Partial petioles of rooteleaves 1 to 1-1/2 iachee, 3 secnents 1 inch long, lobes deeper than in A. eingeleana, both sides of leaves finely pubescent. Lten~leaves eessile, were bracts. sepals 1/2 to 5/8 inch lone, odlong, aevte. Feteal-limbs as long as sepals, rounded at top. spurs shorter than petal-limb, de- cidediy incurved, knobhet. “hole flower blue-lilae. sta- mene not exserted beyond vetal lirb. Follicles 1/2 inch long, densely pubescent. Range - Hilly tracts in the southeastof France, 58 References: Baker, 1878 - pe 19 - 40 4. viscosa. Gonan. - Pl. Kkonsp. p. 267. Tab. 19. Liz. le DeCandolle - Prodr. Vol. 1. p. £5. oynonyms: A. Viscoss, Gonan. EB = ae leactiflora, Kar and Kir. Plant similis; te 4. vulgaris in habit. Stems finely pubescent above, 1-1/2 fect high. Segments of root-leavea stort-staliked, thin, 1 inch long, glabrous on both surfaces, ultimate secmentr rsona oblong. Stem- leaves petioled, commnund. Flowers nearly white or tinged with blue. Serals 5/8 to 3/4 inch long, oblong lanceolate. Petal limbe half «2 lear ae sepals. ‘Sevre 1/3 inch long, slender, nearly gtraicht. Etamens not exserted beyond pet- al-limb. ~@ason - Jure. Range - lative of the 4itai mountaine of South Central Siberia. A Tittle mow erecier, which closely approaches 4B. pobiflora of the cimalayor. Botorsuces: Kar. and Kir - Bul. Soe. Nat. (Moscow) Baker, J. G. ~ Gard. Chron. 1878. Walpe Rep. Vol. 1, p- 50 59 Davis, 4. C, «- Bailey's Cye. Hort. Vol. 1. a4. - ae lncamtiensie, =. Nels. Manv-stemmei, trom a rather lerge semi-fleshy weot, 1-2 dm. hivh; steme and petioles more or less de- cumbent aut afffuce; leaves biternate; the leaflets large, vith erenate Llobet, slichtiy ~vheseent below; lufloresceace and fruits pubescent: flowers 10-15 am. long; sepale green- lshewhite, lanceolate, with emarsinate avex; lamina of the petals eream color, elliptieal, ottuse, loncrer than the slen- Ger hooked hrnobbed sovre, Runse - Sieded cliffs in the Laramie range, Wyoming. tiofeteacer: &. Nelson - Tirct Nep. Pl. Wyo. 78. 1896. Coulter 2223 Neiron, (1909) nr. 191 - 198. 256 - 4. leptoceras, Fisch and iiey. Plaat of the etature of 4. canadensis. Stems about 1 foot high, glabrous. eenents of rooteleaves with éeeper, narrower teeth than in A. vniearie, thin, ¢labrous, biternatels aivided. Lever stem-leaves petioled, biternate, Sepals 3/4 inch lon, half as broad, oblonz, eubacute, bright linuc-bine. Patal-Linhs half as lone as sepale, rounded at tip, yellow at tip. spurs half as long as lamina, slender, Slightly curved. Stamens exserted beyond the petal-limb a little. Follicles slender, glabrors, nearly 1 inch long; 60 style 1/4 inch long. Season - Summer. Range - Eastern Siberia. A little kmown species. Resembles A. canadensis a little in the cutting of the leaves and the less incurved petal-spur. Cultivated at the Botanical Gardens of Birming- ham, England in 1838. References: Fisch. and Mey. - Ind. cem. dort. Petrop. IV. Ledeb. - #1. 08, Vol. 1. pe 57. Mannd. in Bot. Gard. t 755. Small cel. pl. Baker ~ Gard. Chron. 13878. pe. 76. Davis, K. G. In Bailey's Cyc. Hort, Vol. l. Sysoayins : A. Drachyceras - Tures - Fl. Baic. Vol. 1. p. 68. 266 - Ae longissima, Gray. Stem somewhat pubescent with silky hairs, 3 feet high. ‘egments of leaves deeply lobed, narrow, glaucous be- low. Flowers "lake white and straw yellow" (Gray). Sepals lanceolate, 1 to 1-1/4 inches long. Petal-limbs 3/4 inch long, spatulate. Spurs hanging, slender, 4 inches or more long, spur opening very narrow. Stamens not exserted. Season ~ Late July to October ist, Range - Caracol mountains, sotth of Monclovwa, Coa- huila. Ravines of southwestern Texas into Mexico. 61 Allied to A. caerulea and A. chrysanthe, but distinguished by narrower petals and constricted mouth of long spur. References: Gray, A. - Ex. S. Wate. in Proc. Amer. Acad, VII - 1881-82. p. 317. wanieen. J. Me -~ Ue. ». Nat. Herb. Cont. VYol. II, pe 9. Davis, KF. v. - Bailey's Cyc. Hort. Vol. 1. Bia A. Madrensis, rose. Stems about 3 feet high, much braached above and pubescent glabrate below. Basal leaves long-peticled, tri- ternate: lesflete usually on slender vetioles, sometimes sessile, cuneate at base, irregularly cut or lobed, pale green above, paler below, pubescent. Flowers nodding. sSep- als broadly ovate, acuminate, 15 to 18 mm. long, puberulent. Petals with a greenish rounded limb. ipur 4 om. long, very mich contracted below middle, pale red in color. Follicles 5, strongly nerved. Collected on the Sierra Madre west of Balanos in September, 1897. Plant taller than A. Skinneri, with broader sepals, paler flowers, and pubescent. Rose believes A. Skinneri to be a Guatamalan species and thinks that most of the Mexican forms belong to A. madrensis. 62 References: Rose, J. N. - U. S. Nat. Herb. Cont. 1909. p. 265. a8. - Ae micrantha, Eastwood. Lower part of stem and leaves unknown; upper part villous with white hairs, viscid. Petioles dilated at base, stipule-like margine, which are very viscid and villcus. Upper leaves almost sessile; leaflets cuneate, 3-cleft or irregularly divided with rounded obtuse or acute lobes, under side more viseid and hairy than upper. Flower about 2 cm. across, eream-white. ‘Sepals 10 mm. long, 4 ma. broad, ovate, acute, semewhat hairy and viscid on back. cpurs straight or curved, varying in length. Petul-limb truncate, 7 am. long by 5 mn. broad. wtamens and styles exserted beyond petal- limbe. Follicles about 15 mu. long; styles about same lenrth; viseid and villous. Range - abundant in canons of SanJuan River in southeastern Utah, “under shady cliffs where rocks and soil are almost always moist from oozing of alkali waters." Cimi- lar in growth and habit to A. ecalcarata. Described from a fragment of a plant. References: Eastwood, Miss Alice - Two species of aquilegia from the upper conoran Zone of Colorado and Utah. Cal. Acad. Sei. Proc. 2nd Ser. Vol. IV - 1893-94. pp. 559-662. Pls. XVIII- xi. 63 29. = A. moororoftians, Wall. Plant slender, about 15 inches high. ‘Stems fine- ly pubescent. Segments of root-leaves few, glaucous, with short glandular hairs, ultimate lobes about as broad and shallow as in A. vulgaris. Lower stem-leaves petioled and biternate. Flowers white with a lilac tint. Sepale 3/4 ineh long, half as broad, reflexing, oblong or lanceolate. Petal- limbs cuneate, nearly as long as sepal, 1/2 inch broad. Spur slender, scarcely knobbed, erect, rigid. stamens not exsert- ed beyond petal-limb. Follicles pubescent, contiguous, 5/8 to 3/4 inch long. styles 3/8 inch long. Range - Temperate region of Western Himalayas, at an elevation of 10,000 to 14,000 feet. Introduced into Eng- lish gardens in 1852. Reeently found at Kanaora, Tibet, and hence sometimes called A. kanaoriensis. References: Wall - Cat. N. 4713 Jacquem - ex. Camb. in Jacquem Voy. Bot. 7. t 5. as A. Kanaoriensis. Anon. - The Garden, IX, p. 385. Synonyms ¢ A. kanaoriensis. 30. - As oreophila, Rydb. Plant resembling A. chrysantha, A. caerulea, and A. flavescene; 2 to 4 dm. high. Flowers few, 3 to 4 em. long. 64 Sepals oblong lanceolate, acute,pale blue, longer thm the oval, obtuse, ochroleucous limb of the petals. Spurs atraight, widening regnlarly at the throat, ochroleucous or bluish, about twice as long as the limb. Middle elevations of northwestern Wyoming to Idaho and Montana. References: Rydberg, P. ae - Tor. Bot. Club Bul. 29 : 146. 1902. Coulter and Neleon, 1909, p. 191. 31. - ie goxysepala, Traut and Mey. Plant 2-1/2 feet high, slightly pubescent above. Root-leaves long petioled, secondary divisions sessile. Sep- pals blue, ovate-lanceolate, much exceeding in length the pet- al-limbs. Petal-limbs 1/2 inch long, white, round truneate. Stamene not exserted beyond petal-linbe. ‘Spurs mobbed, bent inward, shorter than petal-limb. Follicles pubescent. | > IW A. formosa || A. canadensis ee A. Skinnevt Ne A. pt } OT { Bey “| — A. ecalcavatoa, z io A hry santha. econ. tg UL Uh : % Lt, ie ida i, va) ‘ a ay ie try [-« Bee ee ie, — eee flavescens, A. Jonesi. . ae ele gantiule Neale Map 6b. } A. pubeseens. A. lonaqissima. “TA. ( x ase + Lee 7 mn 1 q se . Sr Lee AN sae see NS Sy ae ae fe pees ene ad A. Micrantha os a A. Saximontana, a A. madvensis. UN V9|aSNI9 "17 oNIdio “y p 12. Ma iy Y 129 vy -pao|ttpP \ no] °Y "oso" P robably ret ecme cf the rarer corte. Onee rlante of a éesirable type have been obtained, they shonld be propagated by self-fertilized seed. Anyone who aims to have a good ccl- lection of Aquilegias should keep pure strains of the more important species growing. Many of the species types are more pleasing than their hybride and they should be given a 90 place in the garden. In 1914 a survey was made of the seed catalogues in order to determine which varieties were offered for sale most widely. The resulte of such a tabulation cught to show which species are most popular in the garden. The summary of results is given in the following table: Species Bot. var. Hort. Var. Catalogues listing Chrysantha 25 Caernles 2e Canadensis 16 wane nee = Double mixed 14 pistes imaces Single " 13 California hybrida 12 Skinneri 12 Vulgaris Nivea grandiflora alba iil Chrysan tha ¥ ® Li enn en ne nme special hybrids 10 Caerulea hybrida 7 " " (Rose queen) 6 Glandulosa 6 Vulgaris 6 ¥ flore pleno 6 fTruncata 5 Stuarti 5 Helense 4 Alpina 4 Flabellata nana-alba 4 Jaetskani 3 Vulgaris Nivea 3 Chry can tha flore pleno 2 Haylodgensis - Hort. - é Olympica 2 Caerulea alba 2 Hay lodgensisa delicatissima 2 Vervaeneana 1 alae ria 1 sepa 1 yinbetiate 1 Sibirica 1 {9is-1914 - Forms listed - 30 Catalogues surveyed - 28 Mabel Cabot Sedgwick (p. 491) reeommends the following sorts for cultivation in the garden: Alpina Caerulea Glandulosa Stuar ti Oxysepala Vulgaris vervaeneana " alva . flore pleno Canadensis " flaviflora Trancata Flabellata = nana-alba Formosa hybrida okinneri Chrysantha ge Hybrid Columbines fome cf the prettiest forms of garden columbines are hybrids. The lone~epurred hybrids af A. caervilea cross- ed with other species areamong the best. A. caerulea by A. chrysantha gives beautiful, long-spurred flowerc of a deli- cate lavendar and yellow. The writer has made this c ross several times and obtained very pretty firet generation types in each case. A. Skinneri is said to give excellent resolts in hybridization. A. chrysantha, slgo, hybridizes well with certain other species. First generation hybrids are often very useful and it is frequently unnecessary to carry the plants into a sec- ond generation to secure beautiful types. It must be remember- ed, however, that s much freater variety of types is obtained in the second generation. These sevregates contain recombim- tions of genetic factors and are frequently quite diverse in form and color. Certain of the new forms are very unlike ei- ther parent. Some of these are valnable and others worthless. It would be well to raise second generation hybrids ina nure- ery, in order that they may be sorted according to colors be- fore being transplanted to a permanent position. It must be remembered that not all hybrids are de- eirable. Many are quite displeasing in color. It is not e- nough to raise hybrids, merely. One should rsise hybrids from known varieties which have given good results in previ- 93 ous crosses, For this purpose it is well to have growing a number of pure strains of the best types of columbines. The crosses are easy to make. The flower should be emasculated before the sepals unfold, thus avoiding any chance of natural crose-pollination. The sepals can be Sep- arated easily by means of forceps and the stamens pulled ont with the came instruments. Great care must be taken to re- move all the stamens, lest one be left in to burst and con- taminate the cross. Often an anther sac will fall into a spur unnoticed. There is & chance that it would mature and perhaps be able to scatter pollen upon the stigmas if the flowers were inverted by the weight of the bag. for most ac- curate work it is well to remove the spurs as well as the sta- mens. Such treatment does not seem to impair the setting of seed, After emasculation the bud should be bagged carefal- ly and left until the stigmas are receptive. The flowers are usually ready for pollination two days after emasculation, thourh the tims may be longer than this when the crosses are made indoors. After pollination the flowers should be re- begged until fertilization has taken place. When the styles have withered ard the ovary begins to enlarge, the bage may be removed without danger of contamination. Each successful cross will yield from a few to one hundred and fifty seeds or even more. 94 aguilegia ic a gouus within which it is pessible to make many species crosses. «4 nwaber of crosses have been Made incidental to these wtudies aud in tome cases the hybrids have been carried to the second generution. The work has nos progresse@ far enough, however, to conetitute a good genetic anaiysis of the species used. ine report on these hybrids will be reserved until a iuter tiie. Yhe following specics crosses have been cuccessful- ly made: A. flabellata x 1. vulgaris var. nivea A. Wolgaris x 4. beikalenszis &. flabellata = .. bSikalensis a. formosa var. hybridea x 4A. Ghryeucthe Ae Wilgaris flore pleno x A. chrysantha “A. chrysantha x A. alpina var. superba 2 Ganadensis x 4. Vulgaris vac. stellate ae m Z Ay " ae u KX ae Gaerulea A. Ganadensi6 x 4. chrysantha Lew o x A. Valearis var. nivea A. vulgaris Var. stellata x a. Ghrysantha A. Gaerulea X A. Chrysantha Ae " x A. vulgaris a. Vulgaris x 4. chrysantha *ob Fy plants all sterile In many cases reciprocal crosses were made. 95 Double flowers There are two types of double columbines: the short- spurred doubles amt long-spurred doubles. The short spurred doublee show an increase in the number of whorls of the corolla and calyx. ‘The sepals are usually inc¥eased to ten in number, instead of five, by the appearance of a second whorl whose parts alternate with those of the first. Similarly another whorl of petals gives the simplest type of doubling, the ten-spurred condition. While the second whorl of parts is slightly above the first, it is close enough to it to appear part of the first whorl. Each succeeding whorl of petals is added at the expeuse of a whorl of stamens. Its parts are, therefore, alternate with those of the preceding and succeeding whorls, and opposite to those of the second whorl from it. (fee diagram 1). The spurs of the third whorl] fit inside those of the first whorl in what is spoken of as a hose-in-hose manner. With three whorls of petals the vertical tiers would shew the following srrange~ ment of spurs: 1+ 2+ 1+ 2+ 1+ 2, ete. see plate 13). with four whorls the vertical tiers would bes 2+ 2+ 2+ 2, etc. With five tiers the spurs would alternate in this way: 2+ 3+ 2+ 6, ete. So with four whorls of spurs the flower would have ten sets of 2-tier spurs. With six whorls there would be ten sets of 3-tier spurs; with eight whorls, ten sets of 4-tier spurs; and with ten whorls, ten sete of 5-tier spurs. Pilate 16 = A. vilearis var. Elore pleno, showing the 1+ 2+ 1+ 2 type of doubling. 96 a4 19) ’ the very donble forms of A. vulraris var. flore plene ar. .. the Iast sort. (fee plate 9, plent 211). In some of the extremely dcnble types there has been such an increase in the number of petal-whorls that the stamenc are alinost en- tirely lacking. It ir usvally coesible, however, to find enough pollen to self-fertilize the blossoms and produce a few seeds. In the cage of the long-srurre’? doubles, the manner of doubling is the came as that described above, but the e2- fect is.different. Instecd of 2 hose-inehose dcubling through- out the flower, only oeescional hose-in-hose flowsre are found. The tips of the cpure cf the uncer whorle get caught on the parts below and the «ours become greatly bent and twisted. The flower appears mis-shapen and irregular. vor this reason the long-spurred doubles are not desirable for the border. Plate 14 - Floral abnormality in A. caer- ulea, James. Flower developing in center ot seed capsule. O7 BIBLIOGRAPHY Albee, Helen R. - Hardy plants for cottage gardene.