HERBERT CALYDOREA Peter Goldblatt^ a James Henrich E, (IRIDACEAE-TIGRIDIEAE): NEW WORLD ON NOTES THIS GENUS AND REDUCTION TO SYNONYMY OF SALPINGOSTYLIS, CARDIOSTIGMA, ITYSA, AND CATILA' Abstract Subequal tepals spreading from the base, free stamens, and short slender style branches with simple, obtuse to emarginate apices, the defining generic criteria for the southern South American Calydorea (Iridaceae-Tigridieae) Mexican (Herbert, 1843a) also characterize the Florida (U.S.A.) monotypic Salpingostylis (Small, 1931), the Cardiostigma (Baker, 1876), and the Venezuelan Itysa Ravenna (1986). Salpingostylis and Cardiostigma hardly differ from one another and are congeneric; they can be distinguished from Calydorea only by their secund flower, weakly eccentric style that divides near the apex of the anthers, and short style branches. Itysa also has secund having flowers with spreading subequal tepals, free stamens and slender, simple style branches. It is unusual only in the anthers coherent at their bases and a style that divides above the level of the anthers. Typical Calydorea has upright flowers and a central style that divides near the bases of the anthers into relatively longer style branches that We twist weakly so that they come to lie between, rather than opposite, the style branches. consider the differences between Calydorea, Cardiostigma, and unite Salpingostylis, and Itysa too insignificant to merit generic segregation them An all under Calydorea, unusual species of Calydorea, C. pallens, which has unequal inner and outer tepals and style branches that lie opposite the anthers, is strikingly similar to Catila amabilis (Ravenna, 1983), the latter having somewhat larger and more intensely colored flowers, the anthers of which clasp the style branches after we anthesis. These two species belong in the same genus, and we recommend treating both as Calydorea in which them consider to be relatively unspecialized. Other members of Tigridieae, notably Onira, Eleutherine, and Gelasine, that have free stamens and simple style branches are considered Calydorea. The phylogenetic position in relation to of the species or species clusters of Calydorea are analyzed cladistically and are discussed in relation to other taxa of Tigridieae and the basic character states of the tribe and its sister taxon Mariceae. ™ W fll Described in 1843 by the and more British botanist Wil- morphology had already been described, liam Herbert, Calydorea was added, erected been to include a than 15 ones have since additional species of bulbous and plicate-leafed Iridaceae from Except some synonyms, each differed obvious for Sisyrinchium Chile, speciosurn, which was mis- largely morphology and was charac- in floral its Two placed Sisjrmc/iium. in features special- distinguished terized by unusual stamen and/or stylar Calydorea from other genera of Iridaceae then and sometimes by differences in the rel- izations, known had that also a bulbous rootstock and and outer tepal plicate ative size and shape of the inner and same leaves, the basic vegetative morphology: The genera of whorls. bulbous and plicate-leafed free stamens; and a style with long, slender, un- now members of Tigri- Iridaceae are treated as divided branches loosely alternating with 1990) of Iri- the sta- dieae, one of four (Goldblatt, tribes mens. Irideae, doideae, the others being Sisyrinchieae, By 843, genera the with one of 1 1 1 similar vegetative and Mariceae. Iridoideae the only is -kA Support DEB for thi. study by grants 81-19292 BSR Foundation and 85-00148 ' from the U.S. National Science We IS gratefully acknowledged. ako thank confirming C. Humphries and J. for his critical review of the manuscript for our cladogram HENNIG86 using the program. 63166- ^"'^^''' ^^"''''" ^ox^ny, Missouri Missouri ""^ Botanical Garden, P.O. Box 299, Louis, nolo' St. ^t'c^["''"^ ujyy, u.b.A. ^ Missouri Botanical Garden, P.O. Box 29Q- St. Ty^hU Mice.,,.; ^-^i^a noon tt c a Missouri 51L 199L Volume Number 78, 2 & Goldblatt Henrich 505 1991 Notes on Calydorea «^ four subfamilies of Iridaceae that occurs in the Cardiostigma sensu Ravenna comprises three New World A (Goldblatt, 1982, 1990). notable species, C. longispathum (the type), C. mexican- feature of Tigridieae um C (Figs. the unusual 1, 2) is (R. FosterfRavenna, and /u«<o«u (r' Fos- degree of variability in the style branches, which Ravenna. Ravenna ter) never explained concept his form frequently a highly specialized and complex of Cardiostigma, but only notable seems its feature structure, intimately associated with the stamens, be to flower. This secund and its is consists of a which are sometimes also specialized. This con- blue perianth of subequal spreading tepals, not markedly trasts with the vegetative uniformity in clearly divided into a limb and claw and without the tribe. perigonal nectaries; free stamens; and a slender. More than 25 genera of plicate-leafed Iridaceae eccentric style with three short, simple branches. with bulbs have been described, and but a few This all at least applies to C. longispathum and C. of these appear to be very weakly founded. They mexicanum. In C. hintonii the three well-defined are largely based on trivial differences in style or style branches are each deeply forked about for stamen Compared structure. with Iridaceae-Iri- two-thirds of their length. This style corresponds doideae and in Africa Eurasia, the generic limits exactly to that of Ainca (Ravenna, 1979) and Alophia finely (Herbert, 1840), and we assume that it seems timely to attempt reduce some to of the is related to these genera rather than to the other We genera that differ so little from each other. Not species o{ Cardiostigma. not deal further will some only degree of consistency is in the criteria with C. hintonii here. for recognition of genera desirable at least within New the family, but there a need to define the SalPINGOSTVLIS is World This genus monotypic, containing only the is descriptions can be Cardiostigma Baker written. Florida species Salpingostylis coelestinum (basio- (1876), Salpingostylis Small (1931), Catila Ra- nym Ixia coelestina Bart.), which was, as recently venna and Ravenna (1983), Itysa (1986) are four as 1974, regarded as a not particularly unusual such examples of weakly defined genera, and they member Sphenostigma oi (Goldblatt, 1974). Small are reduced synonymy Calydorea to in here. Also ^11 1- J- (1931) described Salpins'ostyhs and regarded r>..w;^^n -J it critically considered Calydorea in relation to are _ 11 as monotypic, but h. i-s opi. ni• on was ilareel, y l- enored. r'^/^ rk i^elasi• ne, Onir• a, andt ET^lieuth1 enn• e, which are at ^ / ,^ /rT.r ,^^-t r. i,. tlo.^-r7.. - (Foster, 1945; Goldbla. tt, 1974). Onli y after 1977, 1^,,, J r J L \ least part.l1 y defined b1 y, or contai. n speci.es wi.th, ' . II when Ka venna reduced ophenosti2;ma synon- „i^ to 1 J- • slender, undivid1 ed1 1 ^ branches. style o ymy . Gel,asi.ne, .has bai,pi-ngostyu?s- ^ more or i in less by been recognized. default, Cardiostigma A comparison careful of the available collections ^ D„ of Salptnerostrlis and Cardiosti£(ma sensu Ra- XM tiased1 on t.1he sout1hern Mexi•can C. l1 ongi'spa- f l u i vcuna Icavcs US lu no dioubt tihat. With the cxceptiou /A,,^ tnum /i^i ..\ r> /-t I- . /r» ri-Tz: (Klatt) BakI er, Cardiostii^ma (Bak1 er, 11 876: ^ ... ' , . . . i . i ino\ of 6. hintonii (mentioned above), these are con- o J iuz; was reducedT synonymy • Sph/ enost.igma to in o l - » i. » • Salp.ngo.tyhscoetesUnunt has iihe iden- & Baker (1876: 124) by Bentham Hooker (1883) ^^"f"*^- 11 flower described above tor L. longispathum tical csro^o^n^ r* •. -1 r,,, . r alter was f 1 The 1 founded ^ described. latter it is n and, mexi.canum, secundi - orientation wi-tih subi_- n« o L. in tu on r> /T^i the BrazU-ii. an selIflowianum \ Bak1 er. S. (Klatt) RaU^ /ton^\ o 11 equal tepals, free stamens, and a long style. Ex- oaker (1892) expand1ed1 Sph/enostigma include ^ to ' j ,,11 amm. ati.on of ampl,e h, er,bar-ium materiali and some ^^ l^ ]tl1 speci.es, and. Foster (,,1945,-)v added three more. .„ j r t . - made . accompanymg trom pli ants illustrations live ,i^-„, Rqx.^ navenna , , made . . (1977) has, however, , a convincing j- j j \ ro, al, so maki es - cli ear .tihat. t.ihe st. yli e i• s divided mt*o it r argument . , . tor transferrme the type of ^jp/zenos/^^e-- u j ,. . , , i • some- ^^ distmct branch, es, not obscurely ilobed as / is , !.,.. ^'^a, out not ot,her memb,ers of t1he eenus, to Ge- /t^- a\ Aa j* r •u r. ]„• t.imes t,hough, t (Fig. ilA). reading of the literature m r 1- f^sine, which now 1known by an earlier n- it is .- j j and, exami.nation of • and preserved specimens living With erma. / c^, \ ^ n have not indicated any reason to separate Salpin- ' j^i-,- ^pnenost.igma . , , . Gel,asm. e, Kavenna thus included in r *u a nQ7^ ,. and Cardiostigma Irom one another, a^ nd gostvlis rSir^ r^r.^^ r r (1977, 11979, 11986) transferredi most of t1he erst. Li/.. im .u.J^f..^ r./,-^.... ^F;„ ^:..:.,...:.u them from Calydorea distinguish (Fig. IB). to little ^hile Ennealophus N.E. species genus of that to new ^J"-, Phalocallis, Cardiostigma, and the gen- iTySA era Ainea Ravenna and One Lethia, species that has not dealt with to date Sphenostigma coe- Described by Ravenna in 1986, Itysa comprises is lestinum (= and two species of northern South America. Both are Salpingostylis coelestinum), this Calydorea and Cardiostigma IS discussed broadly similar to in in detail below. 506 Annals of the Garden Missouri Botanical — Figure 1. Flowers, style, and stamen detail and enlargement of style apex in species of Calydorea. A. C, — — — ^B. coclestina (Salpingostylis) (with habit). Calydorea sp. C. C. gardneri (Itysa). D. C. pallens. E. C. amabilis (Catila). Habit and flowers full size; stamens and style x 3; style branch apices much enlarged. vegetative morphology, and they have secund flow- the style. Except for this unusual feature, both ers with subequal inner and outer tepals lac^king species of Itysa seem to correspond to the species and The Cardiostigma diff^erentiation into a limb claw. filaments of in particular. are free, and the long style divides above the level of the anthers into three short, slender branches „ (Fig. IC), the structure of which closely resembles One that in Calydorea. peculiarity of the genus Flowers of Cardiostigma, Salpingostylis, and is that the bases of the anthers are coherent around Itysa correspond closely to the type and most other Volume Number & 78, 2 Goldblatt Henrich 507 1991 Notes on Calydorea species of Calydorea (Fig. IB) with the following of the limb folded back on and bearing a itself The median differences. flower upright not secund, and zone of nectariferous There no tissue. is is the style shorter, dividing opposite the base of basis for the generic separation of Calydorea pal- is the anthers. The style branches are thus compar- lens and Catila. atively longer and more slender, and by a weak between twisting loosely the anthers. (This lie is Phylogenetic Considerations not a fundamental difference in the position of the ovary branches The or style in relation to the stamens.) basic floral morphology in Tigridieae ap- In Salpingostylis and Cardiostigma, the pears be complex style di- to similar to the structure en- vides just above the apex of the anthers and the countered Cypella Ravenna, 1981a, in (e.g., b), exact position of the style branches relative to the which very similar to the basic flower of Mariceae is anthers obscured because weakly is the style is (Fig. 2A), sister taxon of Tigridieae (Goldblatt, eccentric(whetherby designer Here the effect of gravity 1990). the tepals are clawed with the claws in the secund flower is not known) (Fig. lA). How- forming an open cup, and the inner tepals are ever, the structure of the style and style branches smaller than the outer, each folded back on itself of Salpingostylis, Cardiostigma sensu Ravenna, at the base of the limb, in which area there a is Itysa^ and Calydorea very similar (compare zone of nectaries surrounded by an area of con- is details of style arms in Fig. lA-D). trasting coloration. The style branches are thick- As far as we can tell from live specimens of a ened and compressed, and terminate in a pair of Calydorea sp. from Argentina, and an illustration flat appendages, called crests because of their re- of the closely related C. xiphioides (Poeppig) Es- semblance to similar structures in The Iris. fila- (synonym pin. C. speciosa)^ there are no other ments are free, usually threadlike, and too weak taxonomically between significant differences Caly- to support the anthers, which are closely applied dorea, Cardiostigma, Salpingostylis, and Itysa. (not united) to the abaxial side of the style branch. The secund The flower, longer style, short style branch- anthers usually reach to just below the stigma, es, and basally connate anthers are inadequate a transverse band of tissue (Fig. 2A) at the base criteriaon which to base genera in Iridoideae. Thus, of the crests on the abaxial surface of the style may as presently understood, Cardiostigma^ Salpin- branch. (The stigma bear a pair of small gostylis, and Itysa cannot be upheld and are here appendages similar to the crests.) reduced synonymy. The minor complex to differences that This structure thought to be basal is we have much outlined, in conjunction with the correlated (plesiomorphic) for a tribe in which simpler geographic disjunction, appear to merit sectional stylar structures also occur (as in Calydorea) be- The separation within one genus at best. necessary cause the flowers of the central genera of Mariceae, nomenclatural changes are presented below, to- Neomarica and Trimezia, have exactly the same gether with short notes about each species. structure. The genera of Mariceae, however, have a rhizomatous rootstock and plane leaves, and these features are less specialized than the bulb and pli- Calydorea fallens and New cate leaves of Tigridieae. In the other World Ravenna Catila tribe of Iridoideae, Sisyrinchieae, which also has a When other species currently assigned to Caly- rhizomatous rootstock and plane leaves, the inner dorea are taken into account, the circumscription and outer tepals are similar, not normally unguic- C of the genus is somewhat expanded. In pollens ulate, lack a zone of nectaries, and have style Griseb., the style is shorter than in C. xiphioides branches that are always simple, filiform structures and Calydorea and branches sp., the lie opposite sometimes reduced to short lobes. the anthers, though not appressed to them (Fig. Thus, the pattern of evolution floral in Tigri- ID); the tepals are weakly unguiculate; and the dieae, as presented in the cladogram (Fig. 3), is inner tepals are smaller than the outer. Calydorea best seen as a progressive reduction from the com- pallens hears a remarkable resemblance to Catila plex to the simple (Figs. Goldblatt (1990) 1, 2). Ravenna amabilis and To (Fig. IE), the type only presented this argument in detail. consider the species of Catila (Ravenna, 1983), which differs phylogenetic relationships as having a different pat- significantly from only in having a larger flower tern, from the simple to the complex, tempting, It is more with strongly unguiculate tepals of a darker It is, however, less parsimonious as this means that and purple color in the anthers being closely applied either the exact type of flower structure evolved The to the style branches. inner tepals of Catila independently in Mariceae and Tigridieae, or that many resemble other Tigridieae in having the base Mariceae are derived from supposedly specialized Annals the 508 of Garden Missouri Botanical A C Figure 2. Flowers, style, and stamen detail and enlargement of style branch apex in selected Iridaceae.—^A. — steyermnrkii.~B. Onira unguiculata.—C, Gelasine elongata {= G. azurea). D. Eleutherine bulbosa. Trimczia much stamens and x branch apices enlarged. Habit and flowers full size; style 3; style Tigridieae and acquired a rhizome and plane leaves this process, leading to species with flowers like C, and such Northern Hemi- seem speciosa, ultimately to secondarily. Both scenarios milikely. In Iri- C from complex type sphere species as C. coelestina and longispa- progression the deae, a similar of flower to the simple seems likely (Goldblatt, tha, with their secund flowers and elongate, ec- 1981, 1986, 1990) and the same pattern seems centric style (Fig. 3). have occurred Cipiirn, a small genus of Ti- Other genera of Tigridieae with simple style in to & Caljdorea branches resembling those of are Henrich, 1987). gridieae (Goldblatt above, Caljdorea ap- £'/ea^^erme Herbert (1843b), which comprises two In the scenario outlined pears to be a moderately specialized member of species; one species of Gelasine Herbert (1840), Ravenna and Onira most evolved from a Cypella- G. elongata (= G. azurea); Tigridi likely It . we ancestor which the style branches lost their (1983). Regarding Eleutherine (Fig. 2D), feel like in and unrelated Calydorea. unusual red- thickening and apical appendages, differences that to Its it is between the inner and outer tepals were reduced dish bulb, large subterminal cauline leaf, and short seem from (Fig. 3). The flower of Catila amabilis is inter- inflorescence spathes to isolate it pos- The genus should reduction of preted here as being close to this ancestral type, sible relatives. this when immediate ancestry can be and Calydorea pallens represents another step in be considered its Volume Number & 509 78, 2 Goldblatt Henrich 1991 Notes on Calydorea # ftf » #<>^ e .<f ^" iP' .o'^ ^^^ .v^.* ,x^ i»- o- XxS- *& O^ 11 anthers basally 14 style eccenlric coherent > MARICEAE .^ .e^^ 12 anthers 6 flower secund alt. arms style 15 style branches distant from anthers 10 stamens uniied unresolved branching involving Cypella, 7 tepals subequal Cipura. and perhaps olher genera 5 flowers blue 9 lacking nectaries bulb 1 pseudomidrib 2 leaf with 8 claws laxly spread 3 leaf plicate 4a = 7 13 style branches slender ilaments weak branches thickened stylg inner lepals wilh neclaries glands oil I Figure 3. Phyiogeny of the genera and species closely allied to Calydorea suggested by the cladogram based on the character and data matrix presented in Table 1. The major branching patterns in Tigridieae and sister list some Onira-Calydorea tribe Mariceae (from Goldblatt, 1990) are indicated to give perspective to the clade. Polarity of unnumbered characters from Goldblatt (1990). Table 1. Data matrix and character list for the cladogram (Fig. 3). Presence of the specialized condition is denoted by absence by state inapplicable by ? 1; 0; number Character Genus 2 3 4 5 6 7 8 9 10 11 12 13 14 15 1 Mariceae 1 Calydorea 1 1 1 1 pnllens C. I 1 1 Sa Ipingostylis 1 1 1 1 1 1 Cardiostigma 1 1 1 1 1 1 Itysa 1 1 1 1 1 1 Catila 1 1 1 Onira 1 1 Cypella 1 1 other genera, e.g., Gelasine, Herhertia 1 1 1 Characters; the derived (apomorphic) states ed followed by the presumed ancestral (plesiomorphic) condition, lis first — — 1. rootstock a bulb rootstock a rhizome; 2. eaf with pseudomidrib —leaf without pseudomidrib; 3. leaf plicate — = = leaf plane; 4. basic chromosome number x 1 -base number not x 7 (but unknown); 5. flowers shades of blue — — flowers shades of yellow; 6. flower secund flower upright; 7. tepals subequal and not clawed tepals unequal and — — clawed; 8. tepal claws laxly spread claws forming an open cup; 9. tepals lacking nectaries zone of nectaries present — — on inner tepals; 10. stamens united stamens free; anthers coherent at base anthers separate from one another; 1 1. — and 12. anthers alternating with the style branches anthers opposite the style arms; 13. style branches slender not — — forked style arms thickened and apically forked; 14. style eccentric style central; 15. style branches distant from — same the anthers style branches at level as anthers. J 510 Annals of the Garden Missouri Botanical much identified. The subterminal cauline leaf, rather fleshy whorls are well differentiated, the inner being bulb scales, and basic chromosome number, x smaller than the outer, and strongly folded back 6, suggest that possibly a derivative of Ge- at the juncture of the limb and claw. This diflFer- it is = much lasine. Base number in Tigridieae is a: 7 (Gold- entiation is more pronounced than in Catila. = We and from x 6 Onira and genera should be blalt, 1982), apart Eleutherine, feel that related known made in the tribe only in G. elongata (reported studied further before a decision regarding is is & as G. azurea) (Kenton Rudall, 1987). The sim- their status and disposition, ilarity in the karyotypes in these two striking is may Treatment and Systematic indicate a relationship. Gelasine elongata 2C) has united filaments (Fig. Calydorea Herbert, Bot. Reg. 29, Misc. Matter.: and resembling coerulea and a large plant, G. is 85. 1843. TYPE: C. speciosa {Hook,) Herbert G. uragualeasis vegetatively. These two species (= xiphioides (Poeppig) G. Espin.). have more also united filaments but elaborate style branches that have paired crests and a broad trans- Cardiostigma Baker, Bot. London 14: 188. 1876. J. TYPE: longispathum (Herbert) Baker (= Caly- verse, abaxial stigma (Ravenna, 1984). Gelasine C. dorea longispathum (Herbert) Baker). elongata an autogamous, complex heterozygote is New York Salpingostylis Small, Bot. Card. 32: 161. J. & (Kenton 1987) reduced Rudall, with a basic 1931. TYPE: S. coelestinum (Bartr.) Small (= Caly- = chromosome number, most 6, unlike Tigri- dorea ri coelestina (Bartr.) Goldbl.). = C which have seems Catila Ravenna, Nordic Bot. 1983. TYPE: amabilis dieae, 7. likely that the It J. jc C Ravenna (= amabilis Ravenna) Goldbl. apparently simple flower and style of G. elongata hysa Ravenna, Nord. 582. 1986, TYPE: Bot. 6: J. /. from more complex are the result of reduction the gardneri (Baker) Ravenna (= Calydorea gardneri structures present other species of Gelasine. in Baker). Chromosome number known only one other in is n—1 Gelasine^ G. uruguaiensis, (Ravenna, 1984). NOMENCLATURE OF SPECIES OF Presumably, Gelasine evolved from a Cypella-\)ke Calydorea discussed this paper, in and major between ancestor, the distinction the NEW INCLUDING COMBINATIONS two genera and in the filaments, free long in is & Calydorea We 1. coelestina (Bartr.) Goldbl. Hen- Cypella, short and united Gelasine. assume in rich, comb. nov. basionym: Ixia coelestina that similar patterns of evolution occurred floral in 155 Bartr., Trav. ed. PhUadelphia, (1791) 1, Gelasine and Calydorea. In both genera species & London, 153 [ed. 3 (1792)]; Sal- 1, pi. have evolved with flowers with subequal unclawed New pingostylis coelestina (Bartr.) Small, J. tepals and simple cylindric style branches lacking York Bot. Card. 32: 161 (1931). [See Gold- crests. 974) homotypic blatt for additional syn- 1 ( onyms.] TYPE: Bartram, Travels, Illustration in Onira 3. pi. Described 1983 by Ravenna, Onira includes 2. Calydorea longispatha (Herbert) Baker, in J. only O, unguiculata (Baker) Ravenna, based on Bot. London 14: 189 (1876). BASIONYM: Ge- Ilerhertia unguiculata. a small plant with lasine longispatha Herbert Benth., It in PI. is large, pale blue-green flowers (Fig. 2B). Apart from Hartweg. 53 (1840). Cardiostigma longispa- the color and size of the tepals, and the well- tha (Herbert) Baker, Linn. Soc. Bot. 16: J. developed tepal claws, the flower unusual only 102 (1878); Sphenostigma longispathum is & the short which divides below the middle (Herbert) Benth. Hook., Gen. 695 in style, PI. 3(2): of the filaments, and the comparatively long, slen- (1883). TYPE: Mexico. Hartweg 403 (BM, G) 2B The der style branches (Fig. detail). filaments [See Foster (1945: 12) for additional homo- are free and the anthers are appressed to the style typic synonyms. & branches and are closely held to them by the re- 3. Calydorea mexicana (R. Foster) Goldbl. The curving of the anther sacs after anthesis. style Henrich, comb. nov. BASIONYM: Sphenostig- ma mexicanum branch apices resemble fairly closely those of Caly- R. Foster, Contr. Gray Herb. dorea pallens and Catila except for the small, 155: 13 (1945); Cardiostigma mexicana (R. irregular outgrowths of nonstigmatic tissue on the Foster) Ravenna, Notic. Mens. (1978). etc. 2B may stigma lobes (Fig. detail). These represent [Additional homotypic synonyms in Foster the vestigial remains of the style crests that are so (1945).] TYPE: Mexico, Mexico: Temascalte- 8010 prominent in, for example, Cypella, and hypoth- pec, Carboneras, Hinton et al. (holo- The GH; MO, NY, esized as plesiomorphic for the tribe. tepal type, isotypes, F, US). & Number 511 Volume 2 Goldblatt Henrich 78, Notes on Calydorea 1991 Homer Calydorea gardneri Baker, London 1986. Convergent evolution of the ia 4. Bot. J. . new Moraea flower type in six species of (Iridaceae- 14: 188. 1876. Itysa gardneri (Baker) Ra- Irideae) in southern Africa. Ann. Missouri Bot. Card. venna, Nord. Bot. 6: 582. 1986. TYPE: J. 102-116. 73: 2322 Boa Esperanca, Gardner Brazil. Piaui: 1990. Phylogeny and classification of Irida- . — BM [Known 607-627. (holotype, not seen). to us from ceae. Ann. Missouri Bot. Gard. 77: & Henrich. 1987. Notes on Cipura E. Ravenna J. (1986).] illustrations in (Iridaceae) in South and Central America, and a new Calydorea venezolensis (Ravenna) GoldbL 5. species from Venezuela. Ann. Missouri Bot. Card. & Henrich, comb. nov. BASIONYM: Itysa ve- 333-340. 74: W. nezolensis Ravenna, Nord. Bot. 6: 584. Herbert, 1840. Gelasine a zurea; Calydorea. Bot. J. Ravenna Mag. 66: 3779. 1986. TYPE: Venezuela. Guarico: t. 1843a. 138. Calydorea, Edwards's Bot. Reg., avenna . Misc. Matter 29: 85. [Known from us only the illustrations in to 1843b. Eleutherine anomala. Edwards's Bot. . the protologue.] Reg. 29: 57. t. & & An 6. Calydorea amabilis (Ravenna) GoldbL Kenton, A. P. Rudall. 1987. unusual case of ama- complex heterozygosity in Gelasine azurea (Irida- Henrich, comb. nov. BASIONYM: Catilla ceae), and implications for reproductive biology. its Ravenna, Nord. 197. 1983. Bot, 3: bilis J. 95-103. Trends Evol. PI. 1: Ravenna 3 TYPE: Argentina, Entre-Rios, (ho- Molseed, 1971. The genus Tigridia (Iridaceae) of E. [Known Ravenna lotype. Herb. not seen). to Mexico and Central America. Univ. California Publ. 1-113. and Bot. 54: us from the illustrations in the protologue Ravenna, P. 1977. Notas sobre Iridaceae V. Not. Mens. from hving material.] Mus. Nac. Hist. Nat. (Santiago) 21(249): 7-9. 1979. Ainea, new genus of Iridaceae from a . 467-469. Literature Cited Mexico. Not. 132: Bot. A new 1981a. submerged species of Cypella . Baker, G. 1876. Systema Iridacearum. Linn. Soc. J. J. new and genus (Iridaceae), a section for the (s. sir.). 61-180. Bot. 16: 489492. Nord. Bot. 1: J. Handbook George 1892. of the Irideae. Bell . 1981b. Eight new species and two new sub- & . London. Sons, 13-22. Cypella Wrightia species of (Iridaceae). 7: & Bentham, G. D. Hooker. 1883. Genera Plan- J. new 1983. Catila and Onira, two genera of . tarum 3 London. (2). — 197- South American Iridaceae. Nord. Bot. 3: J. 1945. IIL Contr. Foster, R. C. Studies in Iridaceae, 205. 3-54. Gray Herb. 155: The 1984. Gelasine delimitation of (Irida- . Goldblatt, 1974. Revision of the bulbous Iridaceae P. and uruguaiensis nov. from Uruguay. ceae), G. sp. 373-385. of North America. Brittonia 27: 347-350. Nord. Bot. 4: J. 1977. Herbertia (Iridaceae) reinstated as a . new 1986. Itysa and Lethia, two genera of . generic name. Ann. Missouri Bot. Gard. 64: valid 581-588. neotropical Iridaceae. Nord. Bot. 6: J. 378-379. Small, E. 1931. Bartram's Ixia coelestina redis- Homeria J. and 1981. Systeraatics biology of . New York 155-161. covered. Bot. Gard. 32: 413-503. J. (Iridaceae). Ann. Missouri Bot. Gard. 68: Chromosome 1982. cytology relation to in suprageneric systematics of neotropical Iridaceae. 186198. Syst. Bot. 7: