A Systematic Review of the Genus Crinum

(Amaryllidaceae)

L. S. Hannibal

Keywords: Amaryllidaceae, Crineae, Genus Crinum, Crinum, Amaryllis, Ammocharis, Cybistetes, Brunsvigia, x Amarcrinum, Subgenus Crinum (ex Platyaster), Subgenus Stenaster, Subgenus Codonocrinum, Crinum series, Genus Crinum Revision.

This is the most recent documentation available on the Genus Crinum. Previous better known publications are William Herbert's Amaryllidaceae of 1837, J. G. Baker's Handbook of the Amaryllidaceae, 1888, and the writer's Crinum Bulletin of 1970-71 with its master list. As a whole, the genus is not too well known and numerous errors have existed from the time of Linnaeus's long abandoned trinomial nomen specificum nomenclature on up into the current decade. It has taken nearly 40 years to track down some of these errors. One must work with living material to appreciate the obvious distinctions.

The genus Crinum was established in 1737 by Linnaeus while cataloguing and naming both live and dried plant specimens in the Hortus Cliffortianus folio, --all as found in George Clifford's garden and herbarium at Hartecamp, Holland. The mature long-trailing white petals to the Crinum americanum blossoms seen there suggested 'comet tails';----thus Crinum is derived from the Greek 'Krinos', meaning trailing hair or comet tail. One must see a Crinum americanum with its long trailing white petals to appreciate the significance of the name Crinum. Linnaeus made an outstanding choice.

L. S. H.

4008 Villa Court

Fair Oaks, CA 95628

A Systematic Revision of Genus Crinum

By L. S. Hannibal

4008 Villa Court

Fair Oaks, California 95628

Synopsis

Past attempts to classify and subdivide the various Crinum species according to related types have been relatively piecemeal in scope. Both William Herbert and J. D. Hooker recognized the Crinum asiaticum as well as the C. zeylanicum groups as two distinctive polymorphic taxa. J. G. Baker established the subgenera Stenaster, Platyaster, and Codonocrinum groupings. Now according to the 1988 Code the subgenus Platyaster should be Crinum subgenus Crinum, at least the American portion, not the Asian. Ten years back Ingar Nordal recognized the C. jagus and several other confusing African polymorphic groups as distinctive species. In 1988 the Australians recognized that C. flaccidum and allied Australian forms were another geographical polymorphic interbreeding group. The following report, based in part on these findings, is a detailed study which attempts to segregate the majority of all known Crinum species into their recognizable polymorphic or cladial groups, a near score in all, wherein these respective groups are botanically known as 'series', ---where some series are practically a lumped species..

In effect, each series is composed basically of sister-species which have evolved from a distinctive prototype. Members of each series are genetically and morphologically interrelated and many cross interspecifically yielding quite fertile hybrids. In contrast, outcrosses between the various series which occur in numerous instances are less inclined to be pollen fertile. Partial or complete seed sterility is quite common. To date, no known fertile intra-series forms appear to exist under natural conditions, except in Australia. Practical names are proposed for the major series in accordance with William T. Stern's definition of a series. A preliminary clarification is presented illustrating Crinum flowering patterns where these occur in several distinctive umbellate cymose arrangements amongst the various series. Next, it appears that Linnaeus's Crinum americanum supra, his major form, had been confused with a 12-15 flowered C. asiaticum. Then, it has been found that Amaryllis ornata Aiton has an existing 1785 basionym on file which identifies it with C. broussonetii, thus possibly validating the latter as C. ornatum (Dryander ex Aiton) Bury. Lastly, several long ignored early-day 'type specimens' cited by J. G. Baker have been given their overdue recognition, while various fragments of near forgotten botanical history have been unearthed.

---Contents---

Synopsis..........................................................................................................................................2

Background History---Confused Nomenclature.............................................................................2

Code-required genus Crinum and subgenus updates....................................................................16

Crinum Evolution, the Helicoid Cyme and Cytogenetics.............................................................18

Identification Key for Crinum Series (Table I).............................................................................21

Organization of the various Crinum Series (Table II)...................................................................23

Individual Crinum Series (I to XII)..............................................................................................24

Non Crinum Species (Ex Crinum)................................................................................................55

Compendium of Amaryllis ornata Linnaeus fil. key publications ...............................................57

Plates.............................................................................................................................................61

Reference Publications..................................................................................................................65

Abstract of D. Lehmiller's So. African Crinum Revision,1992...................................................69

Information Sources......................................................................................................................70

BACKGROUND HISTORY----CONFUSED NOMENCLATURE!

A modern systematic revision of the Genus Crinum is long overdue. A number of mix-ups in nomenclature have occurred from the time of Linnaeus down to the present. Other nomenclatural problems can be attributed to inadequate research, failing to comply with past Botanical Codes, insufficient information, or a general misinterpretation or recognition of the material at hand. Current taxonomy is concerned in part with a review of past identifications (and misidentifications) with attempts to establish a better understanding of the various interrelationships involved, whether merely morphological or more subtle polygenetic, and secondly, to reevaluate early day descriptions, type-specimens, and tabula (botanical illustrations) as to validity, particularly where cited as lectotypes or holotypes. Here an adherence to current codes one finds some conflicts with past nomenclatural practices ranging from pre-Linnaean and Linnaean to some names established under the now forgotten Kew Code. Many early day species were commonly introduced without adequate descriptions, herbarium specimens, reliable illustrations, or their native source properly cited. The Bank's Herbarium and the subsequent British Museum lack many known species or original type-specimens. Reference publications like William Herbert's Amaryllidaceae of 1837 and J. G. Baker's Handbook of the Amaryllidaceae, 1888, have obvious limitations. Herbert cited a number of Asian species and introduced a score of new ones without providing adequate details, specimens or reference illustrations. -Some,we find, were described by Wm. Roxburgh. J. G. Baker was prone to consolidate remotely related species on limited morphological evidence gleaned from herbarium specimens, while ignoring some obvious morphological and ecological relationships which existed amongst the newly discovered specimens which he described and named. In several instances we have found it necessary to go back to Linnaeus's Hortus Cliffortianus of 1737 to unravel some long overlooked details. In general it appears that some 90 to 95% of most Crinum species are readily associated with one of some nine or ten distinctive taxonomic or cladistic groups, wherein each of these groups has evolved from a well isolated geographical protoform. Knowledge of these taxonomic groups, their general characteristics, geographical distribution, and environmental requirements can aid materially in a plant's initial as well as specific identification, even if it is found under exotic conditions,which is not unusual.

We could use the terms 'Superspecies' or 'Polymorphic Species' (Huxley 1942), or even 'Homologous Species', but 'cladial' (Greek for feather) appears to be the best term applica­ble to delimit a taxonomic group of interrelated Crinum sister and subspecies which are obviously derived from an early primitive form. We note particularly that W. T. Stearn in his Botanical Latin, page 103, 1966, suggests the term 'Series' for such groups, with the name of the series being in the Latin nominative plural. Thus the term 'series' will be employed. It avoids diverse opinions as to what constitutes a biological species. (See D. J. Crawford, 1985).

As an example, there is little question but that the C. asiaticum group is a distinctive clade or series, as are the numerous variants of C. jagus, or the extensive C. americanum alliance with its distinctive rhizomatic rooting characteristics. Historically, around the 1800-1830 period a number of Crinum were being collected about India, the Dutch Indies, Tropical West Africa and the Caribbean. These were introduced into English and European greenhouses where some flowered, were often illustrated and promptly named. The existing competition in naming new species suggests that there was little knowledge at hand regarding the various morphological (and cytological) delimitations. Even today it is still quite unclear what basically constitutes a valid biological species. Opinions vary widely. Back in 1800 the forms with narrow petals like the C. asiaticum or C. americanum were readily recognized as Crinum, but during this period the broad petaled Codonocrinum forms with their trumpet-shaped blossoms were commonly classed as belonging to the genus Amaryllis. Then, in 1806, Ker-Gawler, on reviewing the available published plates and descriptions, decided that all of the then known Codonocrinum forms were basically members of a single diverse polyglot Amaryllis species. He proceeded to identify this entire group as 'Amaryllis ornata', selecting the younger Linnaeus's proposed name for a Tropical African 'Cape Coast' form from Ghana as this group's name in lieu of the senior Linnaeus's A. zeylanica from Ceylon (Now Sri Lanka). He regarded Linnaeus's A. zeylanica name as too provincial for such a widely distributed polymorphic 'species', but he did recognize the senior Linnaeus' plant as his supra Amaryllis ornata K.-G. form.

In 1808, when England assumed administration of Sierra Leone, several large shipments of West Tropical African Crinum arrived on the London flower market where the bulbs sold like Narcissus for 8 pence each. Soon thereafter, Wm. Roxburgh sent back a botanical collection from India which included a number of Asian Crinum, so attempts were made to identify both past and current collections with Ker-Gawler taking an aggressive lead. His lumped grouping of his 'A. ornata' K.-G. went through several confusing changes: 1806, 1809, 1810, and 1817. By noting the species and synonyms which he progressively reorganized one will find that his 'A. ornata K. -G.' evolved from all known Codonocrinum down to those forms with sessile blossoms only. Concurrently, the Asian Radix toxicaria Rumph was recognized as one of the half score plants which Linnaeus had identified as a member of C. asiaticum in Species Plantarum II. Thus the latter erroneously became C. asiaticum Linn. wherein the name C. asiaticum sensu Ker-Gawler is more appropriate. At the same time it was realized that a Crinum identified as American in Commelin's Rariores 14, t.14 plate was not American but from Java and allied to the above C. asiaticum sensu Ker-Gawler. However, the biggest botanical headache was and still is Ker-Gawler's above lumped 'A. ornata' assemblage of 1806 to 1817, and later.

(Note: In order to differentiate between A. ornata Aiton and Ker-Gawler's polyglot 'A. ornata' assemblage we enclose the latter henceforth by single quotation marks. In Botanical Magazine t.913, 1806, Ker-Gawler listed (1) Amaryllis ornata in Hort. Kew, (2) A. zeylanica L. Sp. Pl. II: 418 (Syn. A. ornata supra fide B. M. t.1171), (3) A. broussonetii Red., (4) A. yuccaeoides Thomp., (5) A. spectabilis Andrew's Botanical Repository., (6) A. latifolia Burm., (7) C. zeylanicum Reich., (8) Lilio narcissus africanus Ehret, (9) L. n. zeylanicus Comm., (10) Tulipa javana Rump., and (11) C. giganteum Bot. Rep. all as his 'A. ornata' K.-G.).

For the next ten years nomenclature was chaotic until Salisbury, Hooker, Curtiss, Lindley, and Herbert all agreed that several Crinum species, not Amaryllis, were actually involved. A memo appeared in the Botanical Magazine 45: t.1943, 1819, suggesting several revisions which resulted in Herbert's Enumerato of 1820 (Appended to B. M. t.2121 on C. Broussonetii). Here 'A. ornata' K.-G. was recognized by Herbert as consisting of six or more distinct Crinum species. Since all but C. jagus (A. gigantea) showed a common interrelationship, Herbert listed these as members of his subdivision Ornatae, the near equivalent of a series. Additional identifications and revisions occurred in Herbert's Preliminary Treatise, often called his 'Appendix' of 1821. Here, on page 8 Herbert pointed out that the original A. ornata described in the younger Linnaeus's 1782 manuscript was represented by Aiton's 1785 mounted specimen from Kew Gardens in the Banks' Herbarium as well as by his own C. broussonetii. He added that these examples were the proper A. ornata Aiton form as listed [by Dryander] in Aiton's Hortus Kewensis Ed. 1 of 1789. Unfortunately, botanists and horticulturists of that period often overlooked Herbert's Enumerato, 1820, and his Appendix, 1821. Ker-Gawler's broad views and various erroneous writings were better known to the public than Herbert's findings, so 'ornata' often remained appended to C. distichum, and more particularly to Ehret's Lilio narcissus africanus, (syn. A. zeylanica beta Linn.). The crux of the confusion was Ker-Gawler's above mentioned 1806 attempt to identify C. zeylanicum Linn. along with ten other African and Asian species as his 'A ornata' K.-G. This confusion remained in British horticulture until 1821,--and in part until 1837 and later. However, some British and Continental botanists had conservative views. When Herbert undertook the much needed revision of the Amaryllidaceae we are not surprised to find some 49 distinctive Crinum species cited in his 1821 Preliminary Treatise and Appendix. Here, he listed some 32 species as having slender actinomorphic tepals like C. americanum or C. asiaticum, while 17 additional species had bilateral trumpet-shaped blossoms, where nine of these were distinctly sessile flowered. Several had been listed by Ker-Gawler as members of his all inclusive 'A. ornata' K.-G.

In 1830, when J. A. Schultes published his Systems Vegetabilium, Vol. 7, he depended largely upon Herbert's classifications. He consolidated several obvious variants, but his species descriptions were quite complete, and all in Latin. In most instances he ignored Ker-Gawler's composite nomenclature, particularly the 'A. ornata' K.-G. lumped grouping. Under the Crinum series Ornatae we note Schultes listed C. broussonetii correctly as the supra form with the Kew 1785 A. ornata specimen as the type form.

Following, in 1834, Mrs. Bury published her illustrated Hexandria Plantae. Several African Crinum were illustrated including a six-flowered C. broussonetii which she recognized as C. ornatum. Wm. Herbert, 1837, Kunth in 1843, and Roemer in 1847, all ignored this publication and its fine plates. But we deem that Mrs. Bury had technical assistance since her C. ornatum was recognized as a multiflowered A. ornata Aiton, as well as C. broussonetii.

During this period (1820-1847) there was a continual conflict over nomenclature between Herbert and associates with their conservative views versus Ker-Gawler, publisher of the Botanical Register, who was always argumentative and attempting to retain his expertise by claiming after 1830 that C. zeylanicum L. should still be recognized as the C. ornatum var. zeylanicum. Rather than have an open fight, Herbert adapted Ker-Gawler's C. ornatum var. zeylanicum terminology for the Asian members when publishing his Amaryllidaceae in 1837,- while dropping all mention of his subdivision Ornatae or the 1785 Kew specimen of A. ornata (Syn. C. broussonetii) which Aiton had filed in the Banks' Herbarium. It appears now Ker-Gawler was the one who instigated the now forgotten 'Kew Nomenclatural Code'. This code remained in effect until near 1900. (The de Candolle 1868 Paris Code was long ignored by Kew).

In Herbert's Amaryllidaceae, 1837, a general grouping of Crinum species was attempted, based largely upon the blossom form and foliage. Despite Herbert's keen knowledge of the genus, this classification was far from complete. Botanists following Herbert like J. G. Baker, J. D. Hooker, and I. Nordal have not been in complete agreement with Herbert's classifications and nomenclature, largely due to the numerous interrelated polymorphic taxa. Thus nomenclature varies according to the field experiences or herbarium views which various botanists have accepted concerning their interpretation of a biological species. Herbert's above subdivision Ornatae, rec­ognized here as the C. series Ornatae, has encountered numerous updatings. In 1881 Baker, in his Gardener's Chronicle 'Crinum Synopsis' improved upon Herbert's overall classification by dividing the genus Crinum into three subgenera: Stenaster, Platyaster and Codonocrinum which involves a simple inspection of the tepals (petals). Further identification was based upon the geographical location of the native habitat. In 1888 Baker saw fit to return to several older classical names to avoid the confusion instigated by both Ker-Gawler and Herbert in their unjustified misuses of 'A. ornata K.-G.' and Asian C. ornatum Herb. Baker lumped C. broussonetii under C. yuccaeflorum with the 1785 A. ornata Aiton as the type-species, while separating C. zeylanicum from C. latifolium. In 1894 J. D. Hooker recombined these last two Crinum as a single form of C. latifolium. However, his revision has been largely ignored. In 1968 F. N. Hepper, (in his Flora of West Tropical Africa which covers the 'Guinea Gulf' area from Sierra Leone to the Cameroons, a stretch of 2000 miles), combined C. yuccaeflorum, C. broussonetii, Mrs. Bury's C. ornatum, and C. sanderianum together into a lumped species of 'C. ornatum (Aiton) Bury'. And in 1977, and 1980 Ingar Nordal suggested calling this same basic African group the C. zeylanicum (L.) L.. Her proposed name based on Linnaeus' C. zeylanicum Beta (Syn. Ehret's L. n. africanus) was soon found in conflict with the Asian species whose valid lectotype was identified in the 1689 Moninckx Atlas, plate 34, as Lilio narcissus ceylanicus.

Both F. N. Hepper, 1968, and I. Nordal, 1977, had based their initial consolidations in the series Ornatae on extensive morphological studies of Tropical African Crinum specimens in various Herbariums. In fact, Nordal subjected her findings to numerical analysis, so both had concluded independently that a number of accepted species could be consolidated. However, with field studies, where environmental habitat is a major factor, it was soon found that environmental adaptations were often linked with cytological distinctions.--And that living specimens in the wild showed numerous diversities of species significance that are not obvious in herbaria specimens. A score of past American Crinum growers could have advised Hepper and Nordal that innumerable hybridization attempts conducted largely in Florida over the past 90 years have shown that the C. series Ornatae consist of a stubborn score of like species where outcrosses onto other series are easier to attain that intra-series. Also, many hybrids attained from either combination are invariably seed sterile. Crinum jagus gives similar difficulties. It is not a matter of simple polyploidy, as experienced with some C. bulbispermum or C. macowanii desert-adapted forms, but of distinct cytological incompatibilities.

Ingar Nordal also consolidated numerous forms of C. jagus, (Ex. C. giganteum) and allied types into a common species, C. jagus, in 1977 and 1980. This C. jagus group is quite distinctive. Very few of the named forms are properly described, nor can many of these be readily recognized as distinctive species, thus the consolidation has been well received. Most of the variants in horticultural use appear seed sterile, which may be attributed to B-chromosomes. Like some C. ornatum, most of these C. jagus are quite aquatic, and in the native state increase rapidly by vegetative propagation and flood distribution. As a consequence, exotics introduced into South America and Asia have spread rapidly. The original introductions by African slaves were for medical purposes, but the fragrant white blossoms have favored their garden use. Many present day South American botanists are not aware that these bulbs are Gold Coast exotics.

Nordal's herbarium studies also lead to the consolidation of several tropical forms of C. macowanii. Innumerable variants for the latter are scattered from the Orange Free State northward into Tropical Africa. A portion of the desert forms are tetraploids. In times past C. macowanii has been confused with C. bulbispermum but interspecific hybrids are usually quite sterile which suggests no common interrelationship. In general, the South African Crinum species exhibit more morphological diversity than the tropical. Their interrelationships as well as identities have not completely stabilized. Dr. David Lehmiller cites several corrections where South Africans relate to several tropical species. He is most exacting in his studies. (See International Bulb Soc. Herbertia 48: 32,70,75,86,1992, and appended abstract, herein, also volume 52, 1997.)

A similar consolidation applies to the C. asiatica series. Hooker, 1894, regarded the C. pedunculatum of Australia and C. japonicum as mere subspecies of C. asiaticum. Parey's Blumengarteneri, page 363, 1931, published a more comprehensive consolidation of the Crinum asiaticum group, listing some ten former species as variants of the type form. On visiting the Waimea Botanical Gardens in Hawaii, it is obvious that there is far more diversity within this Asiatica clade than has ever been noted in the literature. All intercross readily, a number of distinctive morphological features exist like the pedicelled blossoms and sequential flowering habit.

Another fairly distinct clade is the Australian Platyaster group,--- the Australasica Series which is best represented by C. flaccidum and C. angustiflorum. Members of this Australian series intercross readily and their introgressively mixed populations are found through much of the country, even in the desert areas. The interblending populations make it difficult to determine which colonies are the basic forms, and which are intraspecific. Crinum flaccidum suggests a major evolutionary drift in that the floral parts resemble those typical of the Codonocrinum, particularly the broad tepals. Yet the tepaltubes are only seminutant during flowering.

There are several lesser known cladial groups of Crinum about Asia, Indonesia and northern Australia whose exact interrelationships are difficult to place due to Herbert's and Baker's rather scant descriptions. Herbert filed no specimens. Kunth's Enumerato is of some aid. The writer summarized these Asian series initially in the June l985 Notes of the Waimea Arboretum and Botanical Gardens. Here in the land areas about Asia and Australia we find the subgeneric groups of Stenaster and Platyaster are rather morphologically distinct with little evidence of introgressive intrusions. In fact, each series in itself is rather distinctive as to foliage and blossom form as well as geographical location, but since the respective species are so fugitive and little known we find considerable confusion exists concerning their respective identifications.

Crinum moorei's cladial relationships are a challenge. This distinctive species with pedicelled flowers outcrosses readily with a large number of Crinum but most of these hybrids rarely set seed. However, those hybrids with the sessile flowered forms like C. yemense, C. abyssinicum, the African forms like C. scabrum, and the Asian C. latifolium variants show some unexpected levels of pollen fertility in B-1 and F-1 crosses than are experienced with hybrids from other cladial groups with C. moorei. Foliage wise, C. moorei resembles that of C. album Herb. or C. abyssinicum when the latter are grown in full shade (Note: there are some morphological differences in the foliage of these two tropical Crinum when shade grown. The foliage becomes more moorei-like.) This, plus the deciduous habits of C. moorei foliage, the bulb's shape and seed similarities seemingly relate C. moorei to the above cladial group. Although, C. moorei may be derived from a primitive Ornata prototype, it shows a close alliance to other South African pedicelled floral forms. Cooler latitudes and climatic changes have encouraged greater speciation and morphological adjustments..

Finally, we have the C. americana series, which is a geographically isolated and morphologically closely interrelated, but a distinctive group and best identified by the rhizomatic means of offsetting, or the blunt apices to the foliage. In this cladial group the floral distinctions of Platyaster and Stenaster are of no genetic significance as tepal width may vary from population to population, --- as do leaf width, number and bulb shape. Unknown to many botanists, Tropical West African 'Gold Coast' Crinum were often imported as medicinal charms by Africans when taken to the New World to work in the sugar cane plantations. Other exotics were brought in from India, the Dutch Indies and China to Rio de Janeiro and Surinam. Several exotics escaped and have spread throughout Central America, and are often considered native species. Some, like C. scabrum cross by moth pollination onto the C. americanum forms giving vigorous but normally sterile hybrids. It is possible that Philip Miller's Crinum in plate 110 was a natural cross like C. submersum, (C. scabrum x C. americanum).

As for botanical errors and confusion, such started with Linnaeus. While at Clifford's estate in Haarlem, Holland, he identified and named two forms of Crinum americanum, a large 'Supra' and a minor 'Beta'. Linnaeus's 'character essentialis' description of C. americanum supra and beta is: "Crinum corollarum apicibus introrfum ungviculatis." meaning "Crinum with apices of the corolla (petals) clawed toward the axis (tepaltube.)" In other words, the last 5 to 8 mm. of C. americanum's tepal apices normally form inturned longitudinally rolled-up claws or drip-tips common to tropicals. And according to Linnaeus's Philosophia botanica, page 202, Aphorism 257, this phrase name is his primary nomen specificum legitimum name, based upon the Clifford specimen (Hortus Siccus Cliffortianus 127.1), while americanum is his trivia name. His subsequent references were secondary to his above primary, namely Commelins's rar. 14,t.14 and Dillen's t.161 must comply morphologically with the above phrase name to be fully valid. These actually lack the claws and are Indonesian, (presumably C. macrantherum Engler sensu Lehmiller), --- so are nonvalid, which Linnaeus well realized, but the plates were accepted even if erroneous since Commelin had cited them as American. In turn, Commelin's rar. 15, t.15 has the claws, so is a valid secondary form of the above legitimate specific type as variant Beta.

(Specifically the name Crinum was adapted by Linnaeus from the Greek 'crinus' which briefly defines the genus name (Critica botanica 246), while americanum (Crit. bot. 258) defines the species source. The 'clawed tepals' are a specific diagnostic feature. Thus, C. americanum L. Supra in Species Plantarum I is amply described by its specific tepal features and cannot be rejected as a nomen nudum despite the appended Commelin misreference. In fact, the Clifford HSC 127.1 genus-specimen authenticates the 'species proper' since it shows these the unique apical claws (drip tips) which Linnaeus noted above as typical of the species. Unfortunately, the late D. O. Wijnands, 1984, like many botanists, disregarded Linnaeus' 'character essentialis', and erred in accepting Atlas 6: t.29 & 30 with pedicelled blossoms as the type C. americanum L., with Commelin's t.14 and Dillen's t.161 plates of Asian Crinum as erroneous paratypes. (Note. Linnaeus cites t.160 in lieu of t.161.)

For those botanists who question the above identification of t.14 and t.161 as Asiatic, note that the blossoms shown are Stenaster with linear tepals, and that all blossoms are pedicelled and thus cited by Dillen as flowering sequentially, individually at random over an apparent flowering period of several weeks. In contrast, the American Crinum are Platyaster, normally with lorate tepals, where all blossoms being sessile usually mature within a period of three or five days, and normally flower in sets of three or four blossoms each. There are no Platyaster features evident in t.14 and t.161, and neither bears a significant resemblance to the Hortus Siccus Cliffortianus 127.1 herbarium sample in the British Museum. This HSC 127.1 specimen appears to be a Floridan form common to the Jacksonville area. Further investigation is under way.

Linnaeus listed Amaryllis zeylanica as common to both Ceylon and Tropical West Africa, while his Crinum africanum was an Agapanthus africanus. Other than seeing C. americanum in flower at Clifford's, as listed in his Viridarium Cliffortianum, I frankly doubt that he saw another living species in flower, only illustrations, although others had brought in bulbs from Ceylon or Sierra Leone. Linnaeus may have seen individually pressed blossoms, but mounting and preserving Crinum blossoms complete with their umbel is particularly difficult as such require special drying techniques to prevent continued growth or rapid decay----a near impossibility in the tropics. Linnaeus apparently had to contend with poorly drawn illustrations and vague, uncertain verbal descriptions. (See W. T. Stearn, 1988). He accepted several broad petaled Crinum as amaryllis, but he identified a Hippeastrum barbatum as a Crinum.

Linnaeus' Crinum asiaticum's true identity has been the most confusing since he simply described this species as having keeled foliage. In the Species Plantarum first edition of 1753 Linnaeus listed four references involving two distinct plants, namely Paul Hermann's Lilium zeylanicus bulbiferum and Rheed's Tanghekolli----An obvious error now as the Rheed reference page cites Belutta pola taly, the native name for Hermann's above bulb. And in edition II of Species Plantarum of 1763 Linnaeus added references listing several more distinctive plants: Radix toxicaria Rumpf, a C. erubescens from America, and another, a Chinese Crinum noted by Osbeck. Then in his Mantissa he cited the Amaryllis bulbisperma (now a Crinum), and finally the correct native name for the above Rheed reference.----Obviously, with seven references listing four or five distinct species Linnaeus definitely had both a broad polyphyletic interpretation and no clear concept concerning a correct identity for C. asiaticum. He seemingly was lumping or accumulating references, geographically, for possible future revision.

Subsequent botanists have obviously encountered difficulty in interpreting Linnaeus's numerous C. asiaticum references. Codewise, his first reference, in Flora Zeylanicus, item 127, should have been his C. asiaticum basonym as noted by Roxburgh's manuscript. But Ker-Gawler selected the last reference from Species Plantarum edition II, the Rumpf Radix toxicaria as the Linnaean C. asiaticum, and then in 1817 identified the Linnaeus original primary reference in Flora Zeylanicus 127 as the C. defixum. Actually, in keeping with Linnaeus's Philosophia botanica, aphorism 256 and 257, page 202, the 'CRINUM---Foliis carinata, Flora Zeylanica 127' is Linnaeus's primary nomen legitimum specificum name. While the supplementary listed forms in his Species Plantarums I and II, including Radix toxicaria Rumpf, were questionably uncertain Asian geographical associates pending further study by Linnaeus. So did Ker-Gawler err or purposely ignor Linnaeus's primary Fl. Zeyl. 127 form in Bot. Mag. t.1073 while citing Radix toxicara as a C. asiaticum, where toxicaria was in a separate subgenus?

Linnaeus the younger ran into problems too. In naming C. erubescens in his manuscript and the Hortus Kewensis, 1789, he and Aiton used the Philip Miller Icon t-110 reference which Linnaeus had erroneously cited to be C. asiaticum. The plate is apparently an Americanum Crinum but so drawn that it is difficult to recognize. As a consequence this invalidates the plate as a reliable lectotype for his C. erubescens Linn. f. At the same time his reference to his father's C. americanum Beta is clouded. Baker noted in 1888 that a 'type-specimen' of C. erubescens which was filed in the Banks' Herbarium in 1784 came from Banks' garden at Spring Grove. Baker describes this specimen in his Handbook of the Amaryllidaceae as the type-specimen. Dryander, Banks' secretary, obviously saw it before editing C. erubescens into the Hortus Kewensis, 1789, which validates it as a lectotype. The Redouté plate 27 in Les Liliacees is considered to have type-precedence by the 1802 date of publication over Ker-Gawler's 1809 illustration. Redouté's plate is an excellent illustration showing deeply pigmented foliage, which was the primary cause for Linnaeus the younger selecting this name. Thus, compared to Banks' specimen, Redouté's and Ker-Gawler's plates are secondary variants, as are Herbert's.

But the greatest confusion exists over the true identity of the Amaryllis ornata Linn. f. As already described above, until 1820, all trumpet-shaped Crinum in the subgenus Codonocrinum had been accepted in England as being in the genus Amaryllis, namely as members of a single composite or group species. While the younger Linnaeus was in England in 1781-82, he undertook to list and describe all bulbous plants and mounted specimens which he encountered. He named a number, but excluded published material where living or pressed specimens were unavailable. He often used the latter published material as supporting polyphyletic references, and here he made a number of gross botanical errors: Like many, he was completely misled by Commelin's rariorum 14, t.14 illustration of C. asiaticum thinking it C. americanum, and with C. zeylanicum by using Ehret's illustration of Lilio narcissus africanus (a C. scabrum form) from Capecoast, Guinea (now Ghana) as the type-reference for his father's well known Ceylonese species. However, his major confusion occurrs now with his Amaryllis ornata, a Capecoast Crinum which Linnaeus f. saw and identified for the Marchioness of Rockingham, the Prime Minister's wife. She was particularly interested with the importation and growing of exotic plants, having them identified and named, and in many cases illustrated. So on seeing and identifying this Crinum as an Amaryllis, Linnaeus f. described it with a typical Linnaean diagnostic phrase name and correctly cited Gulfcoast as the native source. He then listed the Rumphius Tulipa javana as a back-up or synonym reference, not knowing his father had previously identified this T. javana correctly as a synoym form for C. zeylanicum in his Species Plantarum ii of 1762. He was obviously ignorant of the fact that Gulfcoast, Africa, and Ceylon or Java were worlds apart.---And he obviously did not discuss his manuscript's contents with Solander or Dryander, else these and other errors could have been noted. He named several other plants for the Marchioness and apparently left a copy of his manuscript with Aiton, not Dryander as commonly assumed, since a few of his listings were possibly furnished to 1'Heriter during the latter's extended stay in London in 1786-88 when I'Heriter's Sertum Anglicum was in preparation. Meanwhile in 1785, a specimen of a two-flowered C. broussonetii from Kew Gardens was filed in the Banks' Herbarium and identified as a specimen of A. ornata. This still exists and has been accepted as a valid type-specimen, as noted by Herbert. 1821, and Baker, 1888. It could be the Marchioness' A. ornata specimen since both were Gulfport imports.

In 1789, Dryander, under Aiton, in editing Hortus Kewensis used Linnaeus fil's above description of A. ornata to describe Aiton's 1785 Kew basionym, but dropped mention of the Rockingham source, while listing a Lord Petre's bulb and Ehret's Crinum illustration as uncertain first-cultivated A. ornata specimens, assuming they were synonyms. No mention was made of the Rumph's Tulipa javana, or A. zeylanicus Linn. Then, in 1799 Willdenow listed Aiton's A. ornata as a distinctive species while re-establishing Ehret's African C. scabrum illustration under the Asian A. zeylanicus L. along with C. latifolium L. and Rumph's tulipa javana. Otherwise Willdenow's group classification followed that of l'Heriter's Sertem Anglicum.

In 1793, Aiton senior passed away and his son was appointed to manage Kew Gardens, while Jonas Dryander continued as librarian to file Kew specimens in the Banks' Herbarium. It was about 10 years later when Ker-Gawler became aggressively interested in identifying Amaryllidaceae while publishing numerous Latin descriptions for Aiton in the Curtiss Botanical magazine (Henceforth identified as B. M.). In 1805 Salisbury published a plate of C. yuccaeflorum, a near variant of C. broussonetii (Syn. A. ornata), thus recognizing the genus Crinum in lieu of Amaryllis. The next year, 1806, Ker-Gawler published B. M. t.923 showing C. giganteum and C. jagus, listing both as his 'A. ornata beta'.---Concurrently, he lumped in all known Crinum in the subgenus Codonocrinum as mere variants of 'A. ornata', including Salisbury's above specimen. He (and Dryander) were following the original Linnaean lumped or polyphyletic means of group classification. In 1809 Ker-Gawler published a plate of C. zeylanicum, the Coromandel Coast variety, as the Supra form with Rumphius Tulipa javana, featuring both as his 'A. ornata'. Meanwhile, Dr. Afzelius had brought some Crinum back from Sierra Leone which were flowered in Kew Gardens with specimens filed in the Banks' Herbarium. Despite the obvious distichous foliage and one or two slender tulip-like blossoms, this botanically distinct plant was lumped into Ker-Gawler's 'A. ornata' collective group while Dryander labeled the herbarium specimen erroneously:'A. ornata Linn. f., Capecoast Lily', similar to the existing 1785 Aiton type-specimen. Ker-Gawler promptly published a plate of this bulb (B. M. t.1253), this time reducing his 'A. ornata' group to those Codonocrinum with sessile blossoms only, an unannounced modification made obvious by one checking his revised synonym lists. He finally recognized the pedicelled forms as Crinum, but not the sessile.

It is apparent Ker-Gawler's opinions as editor of the Botanical Register carried considerable weight, since, with the issuing of the second edition of Hortus Kewensis, 1811, Dryander edited the Bulbous section using part of Ker-Gawler's above lumped nomenclature. (Apparently Banks, who had disliked Linnaeus f., was also involved in the editing of Hortus Kewensis II, first in dropping Linnaeus fil's name as author from any of his re-included 1789 descriptions of Kew cultivars, and secondly, giving credit for the publication of Hortus Kewensis II to "the late Wm. Aiton", whom had died 18 years previously). Thus, the 'A. ornata' K.-G. now cited in the 1811 Hortus Kewensis consisted of several specimens from the group listed by Ker-Gawler in B. M. t.1253, above, plus Willdenow's 1799 A. ornata description. Willdenow's recognition of A. zeylanica Linn. with Ehret's illustration as a non A. ornata species was disregarded by Dryander and Ker-Gawler.

In 1817 Ker-Gawler published a Review of the Genus Amaryllis in the Journal of Sciences and Arts 2: 342-371 where he still retained his 'A. ornata' K.-G. as a lumped species with the Coromandel C. zeylanicum form as supra. As stated, Salisbury, Hooker, Curtiss, Herbert, and several Continental botanists were not in agreement (B. M. t.1943). Thus Herbert undertook a herbarium and literature search. His preliminary findings were published as an addenda to B. M. t.2121 where he transferred A. broussonetii Redouté to the genus Crinum. In 1821 in his Appendix and Preliminary Treatise, pages 8 and 22, Herbert, as stated, seemingly clarified the status of the A. ornata Aiton by comparing it to the Kew 1785 herbarium specimen and the Linnaeus f. manuscript description. He clearly states that both were the same as his C. broussonetii. He also noted that most of Ker-Gawler's 'A. ornata' forms had some obvious common morphological features, so he recognized this group as a 'subdivision Ornatae' taxonomic entity. He considered transferring the A. ornata Aiton to the genus Crinum, citing C. broussonetii as a synonym, but the involved confusion and Ker-Gawler's possible temperamental rebuff was a deterrent. He finally accepted Ker-Gawler's supra, the Coromandel C. zeylanicum as 'C. ornata var. zeylanicum' as cited in his Amaryllidaceae of 1837. Both recognized Rumph's Tulipa javana as a C. zeylanicum variant.

What is the present status of the two A. ornata herbarium specimens? Both are in the British Museum. Aiton's original 1785 A. ornata mounted specimen was reidentified as C. broussonetii by Herbert, in 1820, while the second, a C. distichum (British Museum #1) was first identified erroneously as C. broussonetii, -- then around 1821 or thereabouts as C. distichum Herbert by Afzelius, also by F. N. Hepper. The Rockingham plate mentioned by Linnaeus f. is either lost or unidentified. It is most unfortunate that no log of these plants or plates was kept by Banks or Dryander. The filing date of the C. distichum, still the #1 specimen, would be of interest, as would be Rockingham's single flowered illustration as a holotype. Some recently discovered Marchioness of Rockingham's correspondence shows that she was actively engaged in plant importation and propagation.

(Since the younger Linnaeus 1782 manuscript is near unavailable to most botanists, as is Wm. Herbert's 1821 Appendix and Preliminary Treatise, we reprint the key texts relating to A. ornata in an appended compendium attached to this report, likewise the actual floral dimensions of the 1785 Kew specimen of the A. ornata Aiton, (syn, C. broussonetii)).

Apparently there was little concern during the 1830 period over the identity of A. ornata Aiton. It was acepted as C. broussonetii. Thus, when Mrs. Bury published her illustration t.18 labeled C. ornatum in her 1834 Hexandria Plantae little comment occurred. Herbert and the Continental botanists disregarded the publication. We now believe that Mrs. Bury and R. Harrison, the grower, were advised by Lindley of the plant's identity, that it was a six-flowered A. ornata Aiton (syn, C. broussonetii). She did state that it belonged in Herbert's Crinum subdivision Ornatae and resembled a small slender C. scabrum with narrower petals and foliage. Herbert's 1821 views on Aiton's A. ornata would have been forgotten excepting that J. A. Schultes in 1830 had likewise recognized Herbert's 'subdivision Ornatae' as a distinctive series, which he identified by the sessile or near sessile blossoms and lorate or ensiform foliage, either scabrous or entire, Here C. broussonetii was the recognized representative of the series (as with Herbert in 1821), with the Redouté Liliacees plate 62, 1802, as the holotype and the Banks' Herbarium A. ornata specimen the basionym. Crinum distichum, C. scabrum, C. submersum, C. zeylanicum and C. latifolium were members of Herbert's series, while C. giganteum (now C. jagus) and C. petiolatum being sessille, were included in error. Both K. S. Kunth, 1843 and M.J. Roemer, 1847, adapted Schultes' Ornatae format in lieu of Herbert's 1837 rearrangement in their subsequent publications.

By 1837, Herbert dropped all mention of his subdivision Ornatae classification in his Amaryllidaceae and identified C. zeylanicum as C. ornatum var. zeylanicum. As a result Herbert's 1821 subdivision Ornatae only remained in a state of occasional use until 1881 when J.G. Baker issued a preliminary synopsis on the Genus Crinum in the Gardeners' Chronicle. Here, Baker promptly returned C. zeylanicum and C. latifolium to their original Linnaean status. He made no mention or use of Herbert's subdivision Ornatae, seemingly in order to clear British botany from the long prevailing 'Ornatae' confusion. He had difficulties in distinguishing the herbarium specimens of C. broussonetii from other related Tropical West African species. He was not able to differentiate between C. broussonetii and C. yuccaeflorum, so he combined the two.--And using the Kew botanical Code gave preference of names to C. yuccaeflorum despite A. broussonetii being first published by Redouté. Here, he noted Aiton's 1785 Kew Herbarium specimen and cited it as type-specimen for his C. yuccaeflorum--C. broussonetii combination. We question this combination, since both Schultes, 1830, vol. 7; page 854 and A. Chevalier, 1950, state that C. yuccaeflorum has distinct foliage and seldom flowers more than two blossoms. We, like Herbert, consider it a separate species, both morphologically and cytologically distinct from the above C. broussonetii. Lastly, Baker encountered a Crinum at Sanders Nursery which was cospecific with Mrs. Bury's C. ornatum, and again, under the Kew Code, named it C. sanderianum, with Mrs. Bury's plate t.18 as lectotype, thus eliminating C. ornatum (Aiton) Bury. What he and subsequent botanists have failed to note is that C. sanderianum is merely a multiflowered C. broussonetii. So Aiton's type-specimen is not only the type for C. broussonetii, but the cospecific multiflowered variants of C. sanderianum (syn. C. ornatum Bury).

In 1950 A. Chevalier also consolidated C. broussonetii with C. sanderianum which he found widely scattered from present day Guiana and Sierra Leone, across the southern Sahara and into southern Sudan. In 1961 J.K. Morton, to be followed in 1968, by F.N. Hepper, both rerecognized C. sanderianum as C. ornatum (Aiton) Bury, not knowing the latter name invalid since cospecific with C. broussonetii. Their consolidation of cospecific forms seemingly included C. scabrum, which is a larger plant with distinctly differing cernuous floral habits. Their type basionym is not stated, and it appears that they related C. sanderianum to A. ornata Aiton via a 'questionable' reference to Ehret's L. n. africanus in Hortus Kewensis I as a lectotype. In this instance a valid herbarium specimen such as Aiton's 1785 basionym of C. broussonetii takes preference over a 'questionable' reference like Ehret's or Lord Petre's uncertain specimens.

Finally, in 1977-82 Ingar Nordal and associates attempted to reidentify the African portions of the series Ornatae as a single lumped species of C. zeylanicum (L) L. But this has been rejected by Dassanayake, 1981, and Wijnands, 1984, who pointed out that Jan Moninckx's 1690 illustration of Lilio narcissus zeylanicus, which was republished by the Commelins in 1697, was Linnaeus's original isotype. This evidence disqualifies Nordal's suggestion that Ehret's Lilio narcissus africanus was a valid lectotype for A. zeylanica L. of 1753. Meanwhile Aiton's 1785 Kew specimen had been relocated and noted as a holotype for C. broussenetii, while cytological studies disclosed that several Tropical East African series Ornatae involve three distinctive forms: namely a C. kirkii with the normal 2n = 22 chromosome count, a C. politifolium, newly named, with a 2n = 30 count and adapted to damp forest conditions, and lastly, a multifloral Crinum resembling C. scabrum of 2n = 22 and 33, and adapted to moist shaded areas.

Previous cytological studies by Jones and Smith at Kew, 1967, have shown some 27 semi and non tropical species of Crinum to be diploids with a normal chromosome count of 2n = 22. In turn, Dr. Abilio Fernandes found that the Orange River, South African C. bulbispermum had a count of 2n = 66, and that some desert C. macowanii are 2n = 44. These polyploid species produce heavy seed sets, but many seed are often parthenogenetic triploid or diploids which lack vigor and seldom result in adult plants. But like Nordal, Jones and Smith had found extensive polyploid diversity amongst the African portion of the series Ornatae which consisted of a number of somatic races. Their findings verified that some C. scabrum are tetraploids, while counts of 22, 24, 44, 66 and higher were encountered for their C. ornata, with some forms having B chromosomes. The study is far from complete, but it indicates great diversity in the various Ornatae forms scattered along the Guinea Gulf and other tropical West African areas where rainfalls are most prevalent. The evidence of such cytological species is best confirmed with attempted crossings of differing ecological or intra-series Ornatae material where numerous incompatibilities occur, or in noting that the resulting F-1 hybrids are often sterile. Numerous dedicated gardeners and plant breeders have encountered innumerable unexpected failures and unexplained results over the past century with these Ornatae, myself included. Sterile C. scabrum outcrosses are common, but a few fully fertile intra-series forms have been finally obtained, mostly in the past ten years with tetraploids, or by backcrossing with F-1 hybrids.

Excluding the C. jagus group, no other Crinum series exhibits such cytological diversity as exists in the series Ornatae. In conclusion, the lumping of the species amongst classical African forms in the series Ornatae is ill advised unless supported by nuclear sequencing, cytological, protein and hybridization evidences. Mere morphological similarities are quite insufficient. In all probability the mono and bifloral forms may have cymose structures which differ from the polyfloral umbels. (Is C. distichum and allied bifloral Ornatae including several from Chad, the Congo basin, and through south central Africa, a distinctive long established taxa when compared to the multi-flowered forms? Further study is desired).

As for the Asian branch of the series Ornatae, D. O. Wijnand, 1984, has pointed out that the C. zeylanicum L. basionym is logically the Commelin Hort. Amst. I: 73, t.73, 1697 illustration. The original drawing in the Moninckx Atlas I: t:34, 1689, is also cited as an isotype.

There has been some debate whether C. latifolium and C. zeylanicum are mere variants of a common species, or not.--And if a valid lectotype exists for the former, namely the Sjovanna-pola-tali Rheed, which Wm. Roxburgh describes in some detail in his manuscript. This plant's plate and description was reprinted in the Botanical Register t.1297, 1830, and his Flora Indica, 1832. The species was re-collected by Wm. Carey, who knew the Crinum and was well versed in the numerous native languages in India and Burma. Carey assisted Roxburgh in plant collections, maintained the Calcutta Botanical Gardens and printed various botanical papers on the Mission Press. Most classical species are now considered mere variants including the Tulipa javana Rumph.

Crinum latifolium has never been in the trade. The C. zeylanicum found in some Floridan gardens is the Coromandel coastal form. This latter variant is nearer C. latifolium than the narrower petaled Moninckx lectotype form. No hardy C. latifolium or C. zeylanicum hybrids are in circulation in American gardens, but freely intercrossing intraspecific hybrids have escaped along the warmer Mexican Gulf below Veracruz where quite a diverse population is taking over coastal marshes. Crinum scabrum x C. zeylanicum often yield runt seed which verifies Jones and Smith's above findings. A fairly close intraspecific relationship exists.

As far as known, the native C. careyanum has been wiped out in the Mauritius Islands, but its hybrid C. virginieum is found in a few Crinum collectors' gardens.

When Linnaeus was establishing his various plant genera he adapted the Greek name 'Krinon', as noted, for the genus Crinum, which he initially applied to the narrow petaled subgenera forms of Stenaster and Platyaster.---while his genus Amaryllis dates back to an early copper etching prepared and published by Jaquin le Moyne, a French artist in 1586 who etched a Hippeastrum vittata, which he called 'Amaryllis'. This etching was republished in 1614-17 by Crispain van de Passe in his Hortus florida, pars altera addition. Linnaeus adapted this epithet as his new genus name for the entire pre-Linnaean 'Lilio narcissus' group which included near all broad petaled Amaryllidaceae from Hymenocallis to Zephyranthes, as well as the Cape Amaryllis belladonna and the trumpet petaled Crinum. In 1744 Georg Ehret illustrated a Cape A. belladonna which flowered in Philip Miller's greenhouses. He later sold the painting to C.W. Trew to publish in his 1750 Natidissimis. The drawing, t.18, was labeled Lilio narcissus americanus belladonna, but is definitely the Cape bulb. Ehret was Miller's brother-in-law. He also illustrated Linnaeus's Hortus Cliffortianus, so it is obvious that all three individuals knew the Cape A. belladonna by this early name, assumed that it was from America, and that it was cospecific with the Hippeastrum which le Moyne obviously knew or named 'Amaryllis'. This explains Linnaeus's lumping of Hippeastrum references under his Amaryllis belladonna in Species Plantarum I and II. The Cape source was not completely verified until 1788, by l'Heriter, while Herbert established the genus Hippeastrum (Horse-star) in 1821. Ironically, when Linnaeus finally saw a mounted Hippeastrum he identified the herbarium specimen as a Crinum barbatum. Meanwhile, in his 1771 Mantissa he wrote: "Amaryllis zeylanicus, excludatur, Crini species est". Apparently his son was quite unaware of this genus change, as were Dryander and Ker-Gawler. Herbert, 1821, dismembered much of Linnaeus's original genus Amaryllis and established a near score of new genera leaving only the Cape bulb A. belladonna unaffected.

Also noted above, although Ker-Gawler initiated considerable nomenclatural confusion he also made numerous improvements in plant descriptions. The majority of botanists preceding him had very little knowledge concerning the genus Crinum, and their brief Linnaean style methods employed in describing species were often all too brief to permit those following to recognize the specific plant forms actually involved. Such was the case with Linnaeus, the younger Linnaeus, L'Heriter, Willdenow, Banks and Dryander (who edited Aiton's Hortus Kewensis I and II). Better descriptions were furnished by Philip Miller, Redouté, Hooker, Salisbury and Roxburgh, or Wm. Carey. Ker-Gawler's knowledge was essentially what he could learn from London gardening associates and the Kew Gardens management. His proficiency in handling Latin descriptions of plants became his vocation.

It is not known who concluded correctly that the original C. americanum supra Linn. misleading synonym references to Commelin's Rar. 14, t.14 in both Species Plantarum and Hortus Kewensis I were nearly cospecific with the Radix toxicaria Rumph from India. Philip Miller was growing C. asiaticum, under the name of C. americanum erroneously. The younger Linnaeus's 1781-82 manuscript indicates that Banks and Solander had accepted the Javanese C. asiaticum to be a C. americanum, probably due to the erroneous identifications by the Commelin and Dillen illustrations which Linnaeus had cited. Linnaeus fil. and Aiton in the Hortus Kewensis of 1789 also accepted C. asiaticum as a C. americanum with pedicelled blossoms! Then upon receipt of William Roxburgh's manuscript and herbarium specimens from India it was realized that Linnaeus had included Radix toxicaria Rumph as a form of C. asiaticum, while seemingly involving it with C. americanum supra. Thus we have Ker-Gawler publishing an illustration of R. toxicaria, identifying it as C. asiaticum L. sensu Ker-Gawler and citing the Commelin's rar. 14, t.14 and Dillen's t.161 references as synonym C. asiaticum forms (Bot. Mag. t.1073). Why Ker-Gawler disregarded the identifications given in Roxburgh's manuscript is unknown, but the plant now known as C. defixum Ker-Gawler was Linnaeus's first choice for C. asiaticum, as cited by Roxburgh. In turn, Ker-Gawler reidentified C. americanum beta as the typical C. americanum form (B. M. t.1034). During this period Linnaeus's C. americanum supra silently vanished from botanical use. In 1888 Baker finally recognized the Clifford HSC 127.1 C. americanum specimen as the Linnaeus original supra form. In turn, he reidentified Linnaeus's Beta form (Commelin's rar. 15, t.15) as the C. commelynii Jacquin Hort. Schon. 2: p.40, t.202, 1802. Jacquin's action in 1802 takes precedence over Ker-Gawler's above attempt to reidentify the Linnaean beta form as the type C. americanum in B. M. t.1034, 1807.

In general, Herbert's identification, classification and naming of the many Amaryllidaceae which he handled is a remarkably study. This became evident in his 1820 Specierum Enumerato, his l821 Appendix, and l837 Amaryllidaceae. Lindley and Herbert did not always agree over Crinum identities. Herbert obviously had the advantage of growing many of the plants which gave him a better insight into subtle differences. However, one prevailing error still exists: In l826 Herbert identified a form of C. americanum as C. strictum. Unfortunately, the name had been used in l807 by Hornemann for a C. strictum which Hornemann briefly described in Latin, but not sufficiently to be readily identifiable. By the Botanical Code, established by Linnaeus this Latin description validated the Hornemann name. Herbert erred in reusing it. In 1830 J. A. Schultes attempted to correct Herbert's oversight and renamed Herbert's plant C. herbertianum. About the same date Wallich named a C. latifolium variant C. herbertianum. Herbert, quite irked, recognized the Wallich bulb as a variant of his Asian C. zeylanicum and rejected the Schultes correction, suggesting that the Hornemann plant could have been one of his hybrids,--a claim which is questionable. Even if a hybrid, the C. strictum Hornemann is still a valid name. In 1847 M. Roemer rerecognized the validity of the name C. herbertianum Schultes for Herbert's plant as illustrated in the Botanical Magazine t-2635. For some reason Baker ignored this correction in his Handbook of the Amaryllidaceae, 1888, probably due to the Kew Code. But L. H. Bailey noted it in 1912. Unfortunately, Traub and others have used the Herbert invalid name on several occasions, applying it to several Louisiana Gulf and Mexican Pacific coastal plants in obvious disregard of the Code's provisions. We should abide by the Code and employ C. herbertianum in lieu of C. strictum Herb.

In the period following Kunth's 1843 and Roemer's 1847 publications, as well as Herbert's death in 1847, interest in the Genus Crinum entered a slack period until J. G. Baker published his Handbook of the Amaryllidaceae in 1888. Most botanists working with the Amaryllidaceae since then are prone to refer to this Handbook. Baker was a herbarium botanist, not a field man, and depended largely upon published illustrations, herbarium specimens and a few Kew grown plants. Herbarium specimens are never as easy to identify as living material. Salient details often vanish or are less obvious. As noted, he dropped a number of obscure species cited by Herbert, Kunth and Roemer, particularly American species where he lacked herbarium material. However, the opening up of the African interior gave him access to new material which he promptly described. Unfortunately, several of these bulbs have turned out to be mere variants of well known species, and nine others were finally recognized as Ammocharis. A restudy of his Asian species may result in several consolidations as his expression "Allied to" or "near C. defixum" distinctly implies that a variant or subspecies is involved.

J. D. Hooker's knowledgeable work on consolidating Asian Crinum in his Flora of British India, 1894, has been largely ignored. Starting about this period Krelage, van Tubergen, Jan de Graff and other Holland bulb growers began introducing and growing Crinum from Africa and India, depending upon Baker's Handbook or his expertise for identification. Hooker's work was overlooked. Specimens soon arrived in Florida where Henry Nehrling attempted their reidentification for L. H. Bailey's 1912 Cyclopedia of Horticulture. Nehrling was an outstanding horticulturist but his reference material was limited,----Thus several apparently sterile hybrids, possibly dating back to some of Herbert's bulbs which were sent to Jamaica, were accepted as species. Other local misidentifications have occurred and some of these bulbs are still in circulation under erroneous names. At no time did the Holland growers, Nehrling or L. H. Bailey indicate any knowledge of Hooker's nomenclatural revisions.

From 1934 on, Dr. Traub has attempted to introduce several new species. Unfortunately the plants involved were inadequately investigated, and those listed in Plant Life as new species have been described previously or are variants. Others were escaped exotics from Tropical West Africa which are widely scattered about South America. Several were noted in the writer's 1970-7l Crinum Bulletin, or in the more recent June 1987 Notes of the Waimea Arboretum and Botanical Garden. Then in the June 1985 Notes the writer presented a preliminary listing of the known Asian-South Pacific Crinum species arranged according to their respective cladial or 'series' relationships. At that time it became obvious that Herbert's descriptions as well as Baker's were often far from adequate and that subsequent botanists have often been confused over identities, particularly where plates or herbarium specimens were lacking, but also whether the plants were in subgenus Stenaster or Platyaster, or if escaped exotics like C. jagus, or several other difficulties. But the main difficulty is with authors not having a comprehensive acquaintance with growing plants and not being able to correlate early written descriptions with the material at hand. Descriptions of Kew's hothouse grown plants are seldom quite typical of plants found in the wild, as floral colors and flare of the perianths differ significantly. This even is reflected in many of the published plates. A second major difficulty is that the specific source of the early Crinum species being described are seldom recorded and thus on reintroductions one is encountering variants. This is particularly true where migratory aquatic- borne bulbs are involved like the C. asiatica series, and particularly the C. americana. Subse­quent collections seldom match up with the holotype or lectotype if such exists, and descriptions are seldom accurate enough to permit more than a general comparison.

Despite some questionable conclusions regarding nomenclature, the work of A. Chevalier, 1950, Andrews, 1956, Hepper, 1968, and I. Nordal, 1977, 1980, and 1983 is of importance. All saw fit to collect and reidentify a number of named Crinum species through inland Tropical Africa. Their work has given us a better understanding of the distribution of this flora, although general inaccessibility makes many forms unavailable for research or garden use. However, identifying C. scabrum as a multiflowered C. ornatum is open to question, as is the consolidation of differing cytological forms, particularly the India C. zeylanicum with C. sca­brum. As mentioned, live garden specimens show contrasts not evident in herbarium specimens, as do cross breeding results.

The question has been raised as to the exact identity of Herbert's original C. scabrum, since a number of polymorphic forms exist. Herbert's bulb came from Rio de Janeiro, having been taken there from Guinea or vicinity around 1620. In Brazil Herbert's C. scabrum is called 'Lilios rajados' in garden usage. Four variants appear to be in circulation in American Gulf gardens, but in Africa the scabrum variants are scattered across the Southern Sahara from Sierra Leone to the Southern Sudan and even Ethiopia. Obviously, over this 5,000 mile span many variants exist as several forms have been introduced into Kew for identification from time to time. The forms in garden use are of unknown origin, but none matches up completely with the Herbert plate as far as pigmentation is indicated. One is distinctly the Ehret version with 4 to 5 mm. wide sharply defined red midribs along its widely expanded tepals.

In several instances intraspecific hybrid Crinum colonies have been noted in the wild. I recall one involving C. flaccidum near Quirindi, N.S.W., Australia, where thousands of bulbs were in flower. No two were alike as to tepal shape and pigmentation. This was in the wild, so no chance of intraspecific hybridization was possible by man, unless by the aborigines. On the other hand intraspecific forms of red pigmented C. asiaticum are common through Asia. These pigmented bulbs have been transported about for medicinal and ornamental purposes for many centuries. As a result few red forms breed true, and most seedlings revert to semi or non pigmented dominant stock.

A Code-required Genus Crinum and subgenus update

It is to be noted that J.G. Baker, 1881, subdivided the genus Crinum into three subgenera: Stenaster, Platyaster, and Codonocrinum, based upon tepal shapes. This is a convenient means of identification. But, according to the 1988 Botanical Code, the subgenus which contains the original 'type-species' of Crinum should be identified as Crinum subgenus Crinum. This creates an interesting problem in nomenclature and specific identity, which requires a detailed clarification:

While Linnaeus was at Clifford's in Holland, 1737, a Crinum americanum flowered. He named the specimen Genus Crinum, for Greek 'Crinon'. He indicated that he had two forms, a larger which flowered and had apical claws on the tepals which he noted accordingly, but his references, based upon Commelin's Horti medicini rariores illustrations of 14, t.14 and 15, t.15, 1706, indicated no claws in 14, t.14. Nor were they evident in a subsequent Dillen Elthamensis plate, also cited. Actually Commelin had erred and had identified plate 14, t.14 as American when the bulb was from Java. (It and Dillen's display pedicelled blossoms which flower sequentially, typical of the C. asiaticum). Linnaeus was not experienced enough to recognize the error. (See W. T. Stearn, 1988). On the other hand Commelin's plate 15, t.15 was a tropical American Crinum and the tepals show the typical claws or drip tips. Thus this latter plate was properly identified as of American origin.

When Linnaeus prepared his Species Plantarum I, 1753, he added species names to all of his Crinum. Thus, he identified C. americanum by the character essentialis: "Crinum corollarum apicibus introrfum ungviculatis"---which is quite correct, but according to Linnaeus's Philosophia botanica aphorism 257, Commelin's rariorum 14, t.14 and Dillen's t.161 lack the apical claws so are invalidly listed, more so, since they are actually Indonesian. Commelin had erred. However, Commelin's 15, t.15 specimen has claws so it is correctly idintified as a valid American minor Beta form. [Note. Linnaeus cites Dillen's t.160. It should be t.161].

In 1782 Linnaeus fil identified his father's beta form as allied to a newly identified C. erubescens cited in his manuscript. In 1784 Banks flowered a C. erubescens in his garden at Spring Grove and filed a specimen in the Banks Herbarium at Kew. This was listed by Dryander in Aiton's Hortus Kewensis I, 1789, as 'C erubescens Linn. f.' with both the above white flowered beta form and Miller's C. foliis carinatis cited as synonyms. In 1797 N. J. Jaquin rejected Dryander and Aiton's associating C. erubescens with Linnaeus's beta form (syn. Commelin's Planta. rar. 15, t.15 specimen) and renamed the white beta form C. commelynii. In 1807 Ker-Gawler indicated surreptitiously that Linnaeus had erred in 1737, 1753, and 1762 in that the Commelin's Planta. rar. 14, t.14 description and illustration was of an Asian Crinum, not American,---that he considered Commelin's t.14 cospecific with the Radix toxicaria Rumph which he recognized as a C. asiaticum L. sensu Ker-Gawler. This left the Linnaeus supra form seemingly nomen nudum without reference. At the same time Ker-Gawler disregarded Jacquin's C. commelynii identification and reidentified Linnaeus's beta form as allied to a West Indies Crinum which he introduced as C. americanum (Beta) sensu Commelin's Planta. rar. 15, t.15 form. (See Bot. Mag. t.1034 and t.1073).

Up to 1807 no one attempted to decipher Linnaeus's 'character essentialis' or examine the Hortus Siccus Cliffortianus 127.1 specimen in the British Museum,-----a specimen which Linnaeus had apparently prepared since it bears his written identification 'Genus Crinum'. Its tepals display the small apical claws (drip tips) which Linnaeus's character essentialis cites. So basically Linnaeus's supra form is validity identified by these claws, and supported by the HSC 127.1 specimen which has these claws. Finally, in the 1881 Gardeners Chronicle, page 388, J.G. Baker reinstated the C americanum L. (supra) as based upon the HSC 127.1 type specimen, employing the latter's leaf and floral dimensions as typical for a C. americanum from the southern U. S. Gulf area. He also noted the Kew Herbarium specimen of C. erubescens from Banks' garden and cited it as the type-specimen, and validly recognized the Linnaeus beta form as C. commelynii Jacquin. ---- But he still ignored the character essentialis, as have all known botanists since Linnaeus. Apparently none have realized that these claws are typical drip-tips common to many tropical plants,----tips which roll up and are most apparent during inclement weather and are characteristic to many native American Crinum. This morphological feature specifically validates the identity of Linnaeus's C. americanum supra form, as first mentioned in his Hortus Cliffortianus folio of 1738. The Commelin Planta. rar. 14, t.14 illustration lacks this feature, as do all Asian Crinum. This latter Crinum is the C. macrantherum Engler as suggested by the number of blossoms, their form, and Dillen's described sequential flowering pattern.

In conclusion, based on the 'character essentialis', the above C. americanum L. supra of 1737 is validly the type-species for the Linnaean genus Crinum, as well as the Crinum subgenus Crinum (ex subgenus Platyaster), while the HSC 127.1 specimen is the Linnaean genus Crinum type-specimen, as well as that for the subgenus, series, and species.

Baker's C. subgenus Stenaster remains unaltered, but is now represented solely by C. defixum Ker-Gawler (syn. Crinum foliis carinatis Linn. in Flora Zeylanicus 127, 1747). While Baker's subgenus Codonocrinum is represented by C. zeylanicum (L.) L.

CRINUM EVOLUTION, THE HELICOID CYME AND CYTOGENETICS

The tribe Amarylleae Baker, 1881, emendatus in parte, non Traub, consists of the Genera Amaryllis, Ammocharis, Boophone, Brunsvigia, Crossyne, Crinum, Cybistetes, and Nerine. This tribe is a very distinctive natural taxonomic group and had its origin as a semitropical marsh plant in East Africa. All of the above genera have solid scapes and produce fleshy albuminous seed which readily float and respond by immediate growth. In contrast, the New World Hippeastrum and allied genera have hollow scapes and produce flat, black, winged seed which are normally wind scattered. This suggests that the genus Hippeastrum and tribal associates developed under semi-open savanna conditions where mild weather favored spring growth and flowering. The hollow scapes retain less moisture typically required for maturing the non-fleshy winged seed. Briefly, when environmental conditions introduce unfavorable stresses, plant and animal cells, including sex, have a tendency to mutate in a great number of molecular sites. In contrast, when conditions are more favorable the tendency to mutate is reduced. Thus those mutations which provide a better fit to the environment often contribute to the plant's evolution and are accentuated in periods of stress.

Basically those members of the above Amarylleae which remained in their original aquatic or semiaquatic habitats have seldom experienced many of the drastic climatic stresses initiated by long dry periods when little or no rain has occurred during the last 20 million years or more. By contrast, those forms which migrated progressively from the aquatic to a savanna habitat, and particularly into desert areas, have encountered repeated climatic changes and the associated dry-land environmental stresses. As a result, numerous modifications, adaptations and evolutionary diversifications have evolved. The most significant being a shift from pauciflowered sessile umbels to multiflowered pedicelled umbels. Secondly, a conversion from primitive symmetrical actinomorphic blossoms to a land-type bilateral zygomorphic arrangement and thus better shield the stigmas and long styles from excessive exposure. Many Amarylleae species have become near homozygous, Crinum included. Thus, when comparing the various Asian, Australian, American and African Crinum the results are self evident, The series Defixa members with their actinomorphic, linear-tepaled blossoms are undoubtedly still near their primitive tribal form, whereas the semiaquatic C. series Asiatica, Amoena, Americana, and Australasica are obvious diversions. In turn, the African (and Asian) land-adapted codonocrinum have experienced numerous climatic changes; much as noted by Jonathan Kingdon in his excellent book Island Africa, 1989. The most overlooked feature is that the above sessile flowered Codonocrinum adapt to the tropics while the various pedicelled to central-southern and South Africa proper. Concurrently, several distinctive foliage patterns have evolved for the dry land Crinum, ranging from the long, near linear foliage of C. bulbispermum and C. macowanii to the broad maize-like foliage of C. moorei and the petioled, reticularly veined leaves of C. jagus. Foliage patterns are usually indicative of the specific series involved, as do the umbel types.

Recent investigation of the sessile and pedicelled umbels and their bostryx structures suggest that the pedicelled have a determinate sequentially flowering structure, while the sessile have an indeterminate group-flowering arrangement. In effect, the flowering cycle is determined by the cyme arrangement. The Muller-Doblies' 1977 report shows that C. moorei has a single helicoid cyme of two turns containing a maximum of ten buds, wherein one bud opens progressively each day spiraling from the perimeter toward the center. In contrast, some major forms of C. asiaticum contains two helicoid cymes of 22 blossoms each, in twin spirals of 3 1/2 turns each. In these latter plants the buds open in pairs progressively from the perimeter to each center over a period of several weeks. Environmental conditions control these cymose structures (See Plate II).

In the case of C. americanum we have several cymes containing one to four buds each., Thus all blossoms open simultaneously on one day, or on two separate days, 48 to 72 hours apart. Apparently the cymose structures for C. acule, C. distichum, and some C. broussonetii (syn. A. ornata Aiton) have been reduced to a vestige of merely one or two floral buds, a marked evolutionary development, associated with a return to a near aquatic habitat. The C. asiaticum and C. moorei manner of flowering is referred to as sequential flowering, whereas C. americanum is semi-simultaneous flowering. The latter commonly occurs with the sessile blossomed forms. Sessile umbellated forms rarely bear more than 10 to 12 blossoms.

As a general observation, excluding C. asiaticum, it appears most Crinum species adapted to moist or boggy locations, usually increase vegetatively, and normally have umbels bearing few blossoms which seed freely. In contrast savanna and desert variants or species such as found in the Kalahari Desert of South Africa seldom produce offsets, but compensate by bearing umbels which may have 40 or more individual blossoms. These latter tend to flower completely within a period of a week or ten days. Crinum macowanii is a prime example of such diversity as the more moist savanna forms may bear 4 to 20 blossoms, depending upon the geographical location. In contrast, the desert forms which only flower at rare intervals (when rains do occur) often produce umbels of 44 or more blossoms. Here flowering progresses centripetally from the entire perimeter to the center of the bostryx in a few days. Further investigation concerning the diversity in umbellate structure and flowering habits is desirable. Herbert noted similar diversity amongst the Nerines (See Amaryllidaceae p.284, 1837).

In numerous instances isolated Crinum colonies (particularly inland forms) have inbred for so many centuries that the entire colony may be completely homozygous. Such plants often self with difficulty, but cross readily with related clones from neighboring colonies which have a slightly differing genetic history. In some instances pollens of foreign species, even Amaryllis, may induce heavy yields of maternal parthenogenetic seed which are fully fertile, or B-chromosomes may induce genetic incompatibilities, particularly amongst the C. jagus forms. Tetraploid and hexaploid Crinum often produce numerous parthenogenetic diploid and triploid seedlings which rarely grow to maturity since most are quite deficient in vigor.

The Jones and Smith 1967 chromosome studies on Crinum indicate a number of genetic abnormalities, particularly in the Crinum series Ornatae where 2n counts of 22, 24, 30, 33, 44, 66, and 87 have been found along with B chromosomes. Hybridization attempts, as mentioned, indicate numerous unsuspected incompatibilities within this series where morphological features suggest mere minor distinctions. By definition, the 1940 Biological Species Concept (B. S. C.) for all living material states a species consists of a group of potentially interbreeding populations which are reproductively isolated from other such groups. (See O'Brien and Mayer, 1991). We attempt to follow this B. S. C. definition in this report, in recognizing species both morphologically and genetically, but realize that cytological incompatibilities as well as interspecific compatibilities present an abnormal B. S. C. problem. It is these abnormal genetic conditions which makes credible species identification such a confusing enigma.

Ever since Herbert attempted crossbreeding Crinum starting about 1820, other horticulturists and plant breeders have attempted to duplicate or improve upon his efforts, often with little success due to the baffling genetic diversities within the genus. In general, the intercrossing of aquatic and semiaquatic species presents no great problem, but the land forms (Codonocrinum) have gone through a number of cytogenetic modifications. It is only by trial and error that some acceptable garden hybrids have been developed in this group. Failures are legion.

Generally, numerous Codonocrinum like C. moorei will take on C. americanum. But such widely crossed hybrids are often sterile, and like this particular cross, will exhibit 95% or more dominant C. moorei features. The conclusion is that in such an outcross the bulk of the dominate chromosomes catenate in lieu of fusing and act as a single dominant entity. Next, we encounter a near seed-sterile, semi-fertile pollen group of hybrids. Usually features from both parents are evident in the F-1 hybrids. Back crosses are best effected by pollinating either parent, but the B-1 hybrids, if obtained, often turn out near identical to the seed parent rather than the classical 75%-25% back-cross. Apparently during meiosis the chromosomes in the pollens from interspecific hybrids tend to catenate and segregate out according to parentages giving near pure line gametes (genomes). Thus the B-1 hybrids are semi pure-line maternal 'recurrent' throwbacks, usually called recurrent parental forms. However, a few genes from the other parent are bridged into these B-1 back crosses to C. moorei, enough to give hybrid vigor (heterosis) and free seeding progeny, whereas the original parents are less so, having come from long established homozygous colonies.

The next level of hybrid crosses are where the F-1 pollen gametes are only partially catenated and the B-1 back crosses often appear as intermediate structural hybrids. Usually the most fertile amongst such hybrids are intraspecific or closely related, where pairing of common chromosomes occur. Here the B-1 and F-2 hybrids usually will exhibit the expected Mendalian diversity, often showing hybrid vigor and normal free seeding characteristics. Structural hybrids having distinctive catenated groups often occur in these instances.

All of the above problems apply to diploids. Polyploids and B-chromosomes add further complications. For a breeder, the classical names are an aid, particularly amongst the C. series ornatae, since each form has different breeding characteristics. Lumping of names only adds to the confusion for the breeder whose results are based upon specific types.

Of late, some intensive intraspecific breeding has occurred, particularly between several forms of C. scabrum, C. yemense, several C. latifolium variants including the Asian zeylanicum, C. flaccidum and several pollen-fertile, seed--sterile Floridan hybrids dating back to Henry Nehrling. In some cases free seeding progeny have been obtained. In part, this supports the broader species concept held by Herbert, Hooker and Nordal. The marked fertility of a flaccidum--scabrum hybrid comes with some surprise as C. flaccidum is Australian, and the other parent is Tropical West African. In effect, this is a fertile subgeneric Platyaster x Codonocrinum cross.----a unique condition wherein little chromosome alteration has taken place over the ages, whereas significant morphological change and geographical migration has occurred. Such marked fertility and heavy seed production has only been duplicated by the writer where he effected intraspecific hybridizations between various forms of C. bulbispermum, or amongst C. macowanii variants. The F-2 generations revealed a number of latent transgressive features. Such intraspecific crosses differ from wider interspecific outcrosses between species where F-2 viable seed are less likely to develop. Those that do develop seldom indicate typical segregation features. Actually, most named Crinum hybrids in circulation are interspecific hybrid 'mules' and for that reason, diverse, freely intercrossing garden populations have not evolved in a manner as exists amongst the Hippeastrum hybrids. Yet, in several instances, viable pollens from wide outcrosses have been used in backcrossing wherein the resulting B-1 seedlings suggest that the male gametes have catenated, basically splitting into the respective parental forms such that the B-1 seedlings appear entirely maternal. However, as stated above, hybrid vigor and free seeding habits suggest that some hybrid genes are present.

Numerous inter-generic hybrids are possible, the x Amarcrinum howardii being introduced in 1911. The writer has produced a number of crosses, some using the Bidwell A. x multiflora (syn. A. belladonna x an unidentified multiflora form) with Crinum subgenus Codonocrinum types. Thus some relatively diverse x Amarcrinum and x Amarcrinaflora comb. nov. hybrids have been obtained. All are sterile. The red flowered x Amarcrinum Dorothy Hannibal is a popular cut flower on the Tokyo flower market. Some unique triploid crosses can be obtained by use fo tetraploid C. moorei pollens. These crosses are 67% crinum and quite attractive.

Key To Crinum Series

1. Blossoms actinomorphic, Ray-shaped, tube erect, linear ...........................3, 7

2. Blossoms zygomorphic, bilaterally symmetrical, tube curved:......................11

3. Floral segments linear, spreading: Subgenus Stenaster:

4. Plants large, semiaquatic, 8-15 leaves 1 to 2 meters long

x 8 to 20 cm. wide, 12-50 blossoms per umbel pedicelled. Asian

and W. Pacific..........................Crinum series Asiatica.............................I

5. Plants small, semiaquatic, long neck, 4-8 slender

linear leaves 40 to 80 cm. long, scape 30-80 cm. long,

5-15 blossoms on short pedicels, tube and segments

5-8 cm. long. Southern Asia and W. Pacific.

.........................................Crinum series Defixa..........................................II

6. Plants small, segments linear to lanceolate, specific

subgenera & series undetermined.....South Africa.......................................XI

7. Segments lanceolate, spreading patently, filaments spreading:

Subgenus Platyaster: (Now according to the 1988 Code

Crinum subgenus Crinum.)

8. Bulb semi-aquatic, globose 5-9 cm., dia., short neck, 6-12

linear-lorate leaves 30-45 cm. long x 2 1/2-5 cm. with, scape

30 cm., 6-12 blossoms on short pedicels, tubes 7-12 cm.

long, tepals 6-8 cm. long. Southern Asia.

..............Crinum series Amoena...and allied forms....................................III

9. Bulb 5-7 cm. dia., short neck, 6-10 lanceolate to linear leaves

30-60 cm. long x 2-5 cm. wide, deciduous; scape 30 cm., 5-9 blossoms

on short pedicels, tube 8-10 cm. long, tepals 6-8 cm. long.

No. Australia.......Cr. series Australasica...................................................IV

9B. Bulb 8-10 cm. dia., 4-8 linear leaves 80 cm., deciduous; scape 60 cm.,

6-8 blossoms on 4 cm. pedicels, tube 7-10 cm., tepals

7-10 cm. x 2-2 1/2 cm. wide, filaments spreading, blossoms

pseudo-Codonocrinum, C. flaccidum.

South Australia.......Crinum series Australasica .......................................IV

10. Bulb slender 2-10 cm. dia., neck long, semi-aquatic, 4-8 leaves,

50 cm. long, spreading. apices blunt, scape 30-50

cm., 2-10 flowers, sessile, fragrant, tubes 7-17 cm., tepals

linear to lanceolate 8-12 cm. Tepal apices form small

longitudinal claws noted by Linnaeus. Plants spread by long

rhizomes. North and So. trop. America.

......................Cr. series Americana..........................................................XII

11. Blossoms trumpet shaped, with curved tube, bilaterally matched and

segments distinctly oblanceolate:

Subgenus Codonocrinum....................................................................12 to 18.

12. Blossoms sessile, fragrant, segment keels pigmented. Bulbs

7-20 cm. dia, short neck, 7-10 leaves often scabrous, margins

undulating, apex not overly acute, deciduous; scape 30-50 cm., 1-20

sessile fragrant blossoms, tubes 7-10 cm., segments 7 cm. x

2 1/2 cm. filaments declinate. Tropical Asia and Africa.

Often semiaquatic in Africa....Crinum series Ornatae................................V

12B. Blossoms normally single or biflowered, bilateral with curved tube. Plants

small, slender foliage. Subtropical and So. Africa.

................Crinum subseries Ornatae pauciflora.........................................V

13. Blossoms pedicelled, foliage spreading umbellately from a

sheathed neck. Bulbs 7-12 cm. dia., 5-6 leaves lorate

70-100 cm. x 6 to 12 cm. wide, deciduous;. 2-10 large

white to pale pink blossoms. desires shade. Natal, South Africa.

................Crinum series Umbratica.........................................................VI

14. Blossoms sessile, fragrant, segments broad, white, bulbs

3-10 cm. dia., globose, semiaquatic, neck short, 7-10 leaves,

petioled, uniquely reticularly cross veined 50-75 cm. long,

x 7-10 broad, deciduous; scape 50-75 cm., 4-6 blossoms,

fragrant, tepaltube 10-15 cm., segments oblong, obtuse 7-12 cm.

Central and west Tropical Africa.

...........Crinum series Jagi.......................................................................VII

15. Blossoms usually 6 to 40, pedicelled, flowering sequentially;

bulbs 6-14 cm. dia., long neck, foliage long and linear;

acute apices: takes three years to mature, 5-8 leaves often

glaucous 60 to 200 cm. long; scape 45-60 cm., tepal-segments

6-10 cm. x 2-4 cm. width, pungent odors, many with pigmented red

blossoms. South Africa.

.................Crinum series Capense...and allied forms............................VIII

16. Blossoms few (1 to 8), sessile or subsessile; foliage slender 0.5 to 2.5 cm.

wide by 20 to 60 cm. long, arching, v-channeled, varies with the species.

Plants often growing in summer flooded pans or moist river embankments.

Found in Botswana north into Zambesi River headwaters and west into

Namibia and Angola.

...................Crinum pseudoseries Tenuifolia.........................................IX-A

17. Codonocrinum. Unplaced African species.

....................Proper series not determined................................................X

18. Native African subgenera Stenaster and Crinum (ex Platyaster) species,

................Crina antiqua pro parte..................................................................XI

19. Plants inadequately described. Blossom type not determinable.....................XIII

20. Non Crinum (initially misidentified, mostly Ammocharis) ..........................XIV

Note! The above foliage and floral dimensions are general for each particular series under study. There are always exceptions in plant sizes and blossom number due to environmental adaptations or genetic conditions. But for one experienced with Crinum the distinctions common to each series are quite obvious.

Organization of the various Crinum Series

Crinum

Above Crinum series comb. nov. mihi.

Table 1.

Individual Crinum Series

I. The Crinum series Asiatica (Ker-Gawler non Linn.)

- subgenus Stenaster

The Crinum series Asiatica, or clade, was recognized in part as a single species by both William Herbert in 1837, and then by J. Dalton Hooker in his Flora of British India, 1894. Finally in 1931 Parey consolidated some ten classical species as a common species. Geerinck, 1993, has similar views. In usage, this cladial group is normally treated as such, but some like C. xanthophyllum are regarded as distinct. Morphologically, there are so many transitional forms in existence that well defined variants are only maintained through geographical isolation. All intercross intraspecifically all too readily. However, the clade is unique as it has gone through several distinctive stages of evolutionary development which sets the larger multiflowered forms apart from the lesser, and particularly from all other Crinum forms. The multiple bostryx is unique. (See Muller-Doblies reference on helicoid cymes). It is a primitive environmental adaptation to permit fertilization of some blossoms during long periods of moist or damp, inclement weather.

Basically, it appears the more primitive forms of C. asiaticum bear only 10 to 20 blossoms in an umbel and have rather limber foliage seldom over 8 to 10 cm. in width. The basal growing plates are normally subsurface. Then an increase in plant size occurred when the multiple bostryx evolved, wherein the umbel would bear from 40 to l50 or more individual blossoms. Secondly, probably due to aquatic adaptation, some basal plates have resisted deterioration and evolved into peduncular trunks not unlike some Pandanus, thus elevating the growing crowns well above ground level. One semiaquatic group represented by the Australian variant pedunculatum evolved a peduncular stump 8 to l6 cm. in diameter by 50 to 80 cm. or more in height, while the huge variant procerum may have a keg-like stump 25 to 50 cm. in diameter by two or three times that height. Foliage of these elevated plants is invariably quite rigid, whereas that of the non-elevated is prone to fold over with the apices drooping or hanging down. Another evolutionary development, or redevelopment, is the appearance of red pigmentation in the foliage and blossoms of some of the procerum and allied variants. Man has scattered these red forms widely.

It is normally accepted that Linnaeus had at least four or more distinctive Crinum listed as C. asiaticum in his Species Plantarum I and II, and in latter publications like his Mantissa, as noted below, --- all in addition to his long ignored 'Crinum foliis carinatis, Flora Zeylanica #127'. Actually, if one examines his Philosophia botanica, page 202 citing aphorisms 256 and 257 it is obvious that his Fl. Zeyl. 127 (of 1747) was his original nomen specificum legitimum name for his accepted C. asiaticum species,--- while its holotype was Paul Hermann's original 1677 Musaeum zeylanicum 5: t.124 sketch, and syntype Hermann's 1687 Lugduno batavus 682, t.683, an etched copy of t.124; (Syn. Belutta pola tali Rheed). Since Linnaeus' diagnosis of Asian Crinum was 'having channeled foliage' including Radix toxicaria Rumph, as listed below, those at Kew in 1807 saw no irregularity in recognizing R. toxicaria as a valid form of C. asiaticum in lieu of the near unknown Musaeum zeylanicum t.124 sketch, which Banks then owned.

1. Lilium zeylanicum bulbiferum & umbelliferum, Herm. lugbd. 682, t.683, 1687.

2. Tanghekolli, Rheed mal. 11: 155, f.69. (In Spec. Pt. I only).

3. Radix toxicaria Rumph. Amb. II: 155. f.69. (In Sp. pl. II, subg. Stenaster).

4. Amaryllis bulbisperma Burm. Prod. 9 (In Mantissa, C. bulbispermum).

5. Belutta pola talis, Rheed mal. 11: p.75, t..38; Rudb. elys. 2: t.90. (Syn. Fl. zeyl. 127 Linn.; C. defixum Ker-Gawler).

Historically, due to the few subadquate descriptions and crude drawings Linnaeus's early knowledge of Asian Crinums was quite limited. Thus in 1747 when he catalogued Paul Hermann's Musaeum zeylanicum Asian collection of specimens and drawings he briefly identified a bulbous Crinum drawing in his above Flora Zeylanica as '127 Crinum foliis carinatis' (i.e.: specimen 127 with channeled foliage). No trivia name was assigned, but he added the four above cited references, (2) to (5), which he assumed were probable synonyms: In his 1753 Species Plantarum I he reinstated this Crinum as 'Crinum follis carinatis Fl. Zeyl. 127 asiaticum' making no .improvements to his descriptive name, but adding the trivia name asiaticum, --- while relisting Lilium zeylanicum bulbiferum & umbelliferum Herm. lugbd. 582, t.683, p.77 and Belutta-pola-talia Rheed. Mal. 11: t.38, p.75. as synonyms. Then in the 1763 Species Plantarum II he repeated the above Crinum foliis carinatis name with Lilium zeylanicum bulbiferum & umbelliferum, Herm. lugbd. 682, t.683 and Radix toxicaria Rumph. amb. 11: p.155, t.69.---And in his 1771 Mantissa 362 he relisted Belutta-pola-talia and added Amaryllis bulbisperma (syn. C. bulbispermum). Synonyms (1) and (5) are now considered valid while (3) and (4) are complete misfits, being other subgenera. Unfortunately (3) was finally selected as C. asiaticum, ie:

About 1806 Aiton, Dryander or Ker-Gawler finally realized that the so-called C. americanum forms in Kew and in the 1789 Hortus Kewensis with pedicelled blossoms were near cospecific with the Asian Radix toxicaria Rumph. and Commelin's rar. 14, t.14. Likewise, on examination, that Linneus's original C. foliis carinatis, Asiaticum's short nomen specificum's description, was completely inadequate for positive identification, though possibly allied to Belutta-pola-talia. So Linnaeus's C. asiaticum was discarded as non verifiable. However, since Linnaeus had cited R. toxicaria as a C. asiaticum synonym in Species Plantarum II above, and since Kew had such a specimen, Ker-Gawler published a plate of R. toxicaria Rumph in the 1807 Botanical Magazine 27: t.1073, identifying it as C. asiaticum. He went to some effort to list all known publications citing Asian Crinum, but subtly omitted those citing the original 'C. asiaticum L.', as in Species Plantarum I and II. Thus Hermann's bulb and name was discreetly banished, but in France Redouté was not aware of this action and identified an Asian Crinum much like Commelin's 14, t.14 in his Liliaceees 7: 332, 1810 as C. americanum. It is now C. redoutéanum Roemer. Not all botanists agreed with Ker-Gawler's nomenclatural changes. William Roxburgh identified R. toxicaria as Crinum toxicarium (Rumph) in his 1814 Hortus Bengalensis. According to the botanical code he is correct in his interpretation, but by usage the name C. asiaticum Ker-Gawler non Linn. is well established for a wide range of Asian forms.

The forms of the variant japonicum appeared few-flowered when I flowered this plant, but plants in Hawaii appear multiflowered, so more than one type undoubtedly exists, or the plants there are hybrids. However, I consider the few flowered and the multiple cyme C. asiaticum sufficiently distinct to retain species distinction as indicated below in SeriesI-A.

Next, we have several Crinum native to Madagascar and vicinity which appear to have some debatable affinity to the C. asiaticum by their Stenaster flower form as well as foliage. Descriptions are too limited to adequately evaluate these plants, thus their placement is tentative.

Finally we have C. amabile and C. augustum which Herbert considered hybrids since they are seed and pollen sterile. They are noted for their large showy wine-red blossoms. Crinum amabile is the lesser in size and its 30 odd blossoms have linear Stenaster-type tepals while C. augustum has a larger conical bulb and semi-lorate Platyaster tepals. Raina and Khoshoo report the latter sterile since it is a triploid. Several breeders, including the writer have semiduplicates of C. amabile obtained by crossing C. asiaticum var. pedunculatum with C. latifolium variants. Such wide bicladial outcrosses are normally sterile.

Around 1830 Wm. Herbert sent offsets of C. amabile and C. augustum with other Crinum to Jamaica. These bulbs are now widely scattered through tropical gardens. However, a Gilbert Island Crinum recently cited as a possible C. augustum is more likely the red flowered C. procerum variant listed below. The latter are widely scattered through many Pacific Islands.

It is suggested that the entire C. asiatica series be examined in greater detail as none of the above morphological distinctions cited have received attention in past literature. The African C. buphanoides should be included in the study.

The large umbels of C. asiaticum forms makes pressed specimens near impossible. In some instances single blossoms and a leaf have been pressed. This explains the early-day lack of herbarium specimens. If specimens are warmed in a microwave oven to terminate cell growth, then drying the specimen is less difficult.

I. Crinum series Asiatica (Ker-Gawler non Linn.) and Associates

Subgenus Stenaster

Blossom pedicelled, flowering sequentially.

I-A Multihelicoid Blossoms, (20-150 or more blossoms per umbel)

1. Crinum asiaticum sensu Ker-Gawler non Linn. [non Fl. Zeyl. 127 (1747); Spec. Pl. ii: 419, 1763 ex part]; Bot Mag. t-1073. Asia. Weathers Bulb Book 155. 1911, Kew.

   1. var. toxicarium (Rumph. amb. 11: p. 155, t. 69.); Herbert Amaryl.. 243; Roxb. Hort. Bengal. 23; Bot Mag. 27: t-1073. China.

   2. var. declinatum Herb. Bot. Mag. 48: t-2231. Silhet, China.

   3. var. sinicum Roxb. Bot. Mag. 47: t-2121, p. 7. China.

   4. var. anomalum Herb. App. & Treatise 22, 1821, by Bot. Reg. 1821; Bot. Mag. 47: t-2121; 57: t-2908. China.

   5. var. procerum (Carey) Herb. Bot. Mag. 53: t-2684; Bengal.

      1. Syn. C. gigas Nakai Bot. Mag. 44; 34, 1930. Mariana Is.

      2. Subvar. kaaawanum Hann. La. Soc. Hort. Research iii (5): 259, 306, 1970-71. (Red pigment form). Kaaawa, Hawaii.

      3. Subvar. splendens Hann. La. Soc. Hort. Research iii (5): 259, 306, 1970-71 (Red pigment form). Hawaii.

      4. var. pedunculatum (R. Br.) Hooker Prod. Nov. Holland: 297, 1810; J. D. Hooker Fl. British India 6: 284, 1894; Red. Lil. t.408, 1813.

      5. var. japonicum Baker. Handbk. Amaryllidaceae 75, 1888. Japan.

3. Monohelicoid forms (10-20 blossoms per umbel)

5. C. redoutéanum Roemer Regni Vegetabilis IV Ensatae, pt. 1:64, 1847:

   1. = Syn. C. americanum Redouté Les Liliacees 6: t-332, 1811; Commelin's Medici 14, t.14, 1706(?);

   2. syn. C. americanum Linn. f. 1782 MS; Hort. Kew. I:413, 1789; Willd. Spec. Pl. 2: 46, 1799.

   3. C. concinnum Roemer 1847. (an Indonesian spp.)

6. C. xanthophyllum Hannibal La. Soc. Hort. Research. Jr. iii (5): 266, 1970-71. Fiji, New Caledonia. Holotype filed at Calif. Academy of Sciences, Golden Gate Park, San Francisco. (yellow foliage).

7. C. brachyandrum Herb. Bot. Mag. 47: 2121 p. 8, 1820; 48: 2231, p. 2; Hannibal in Australian Plants. vi (45): 20, 1970. Fitzroy R. Queensland.

8. C. sumatranum Roxb. Hort. Bengal. 23; Lindl. Bot. Reg. t-1049. Sumatra. (C. rigidum Herb.)

9. C. douglasii Bailey, Queensld. Dept. Agr. Bot. Bull. X: 27, 1895. Qu.; Flora of Australia 45:, 1987.

10. C. australe var. rubicaule Herb. Amary. 246, 1837. North Queensland.

11. C. australe var. exaltum Herb. Bot. Mag. t.2121 p.5, 1820; Amary. 246, 1837.

12. C. brevilimbum Herb. Bot. Mag. t.2121 p.7, 1820; Amary. 266, 1837. Queensland, Cairns Coast. Blossoms white, filaments red.

13. C. asiaticum L. foliis carinatis. Fl. Zeyl. 127, 1748. = C. defixum Ker-Gawler 1817.

15. Status Unknown, indeterminable from descriptions

17. C. macrantherum Engl. in Bot. Jahrbk. Vii: 488, 1886. Syn. Lilio Asphodelus americanus serpervirens maximus polyanthus albus Commelin's, 1706, Hort. medici. rar. 14, t.14; Moninckx Atlas 6: t.29 & 30, 1706, (in error identified as C. americanum L. sensu Wijnands, 1984); Dill.,1732, elth. 194, t.161), f. 195. * Native to Bismark Archipelago = C. rumphii Merrill (?) = form near C. sumatranum Roxb.

18. C. cortifolium Hallier in Lorentz Nova Guinea Viii: 900, 1913. New Gu.

19. C. northianum Baker, Gdn. Chron. i: 761, 1882. Borneo. (Platyaster?)

20. C. pusillum Herb. Amaryl. 255, t-32, f 3. Hab. Nichobar Islands, Geerinck Flora Malesiana sect. 1, XI(2): (1990), p. 361.

21. C. macrophyllum Hallier in Lorentz Nova Guiana VIII: 899, 1913. New Caledonia.

    1. Crinum asiaticum var. cuprefolium Traub in Plant Life 16: 94, 160 from Fosters Garden, Hawaii, is not true breeding, but an apparent pigmented procerum hybrid.

       1. * Note. Linnaeus erred in Listing Dill. elth. 194, t.161 as t.160 in several publications.

23. I-B. Semi-Crinum asiaticum forms, Western Indian Ocean

    1. These two species bear 30-40 blossoms on short pedicels and may be allied to those listed under XI multiflora.

25. Crinum bracteatum Willd. Spec. Pt. IV (ii): 47; Bot. Reg. t-179.

    1. syn. C. brevifolium Roxb. Hort. Bengal (23); Fl. Ind. 129, 1832. Habitat. Seychelles. (2 ft. foliage).

26. C. ligulatum Baker Jour. Linn. Soc. xx: 270, 1883. Madagascar.

29. I-C. Sterile Semi-Crinum asiaticum forms

32. C. amabile Donn. Hort. Cantab. VI: 83; Ker-Gawler Bot. Mag. T-1605. Hawaiian 'Queen Emma Crinum'. Native Sumatra.

    1. Syn. C. superbum Roxb. Flora Indica ii: 133, 1832.

       1. (Note. Blossoms in Bot. Mag. illustration distorted. Use Flora Indica descrip­tions:

       2. Tall leafy trunk 7-12 cm. dia., foliage resembles C. moorei, scape 80-120 cm. Long, umbel 20-30 blossoms with semi lorate wine red-keeled tepals 7-12 mm. wide x 12-15 cm. long. Sumatra. Considered a natural Stenaster hybrid by Herbert.)

34. C. augustum Roxburgh Hort. Bengal 23; flora Indica ii: 136, 1832; Bot. Mag. t.2397; Bury Hexand. pts. t.64, 1834; Bot. Register t.670. (Carey in Flora Indica describes the bulb as large, conical, 15 to 20 cm. dia. x 30 cm. long, foliage like small C. procerum, scape heavy 80-100 cm. long, heavily pigmented red blossoms 30, semi declinate with long lanceolate tepals 12-17 mm. wide. Reported a triploid, possibly hybrid. Tepals near Platyaster in form. Native to Seychelles & Mauritius marshes.)

35. C. augustum Roxb.(?) Overy, Polunin & Wimblett.., pl. Kiribati iii: 8, 1982. Habitat Gilbert Islands (Possibly a red flowered C. asiaticum var. procerum like kaaawanum introduced as a garden plant during occupation.)

36. C. x 'Hana.' An apparent natural hybrid involving a red C. asiaticum found at Hana, Maui, Hawaii, bearing exteriorly pigmented red blossoms which flower in a unique random centrifugal manner. Seed sterile.

II. The Crinum series Defixa (Ker-Gawler.)

--Subgenus Stenaster Asian

As stated in the introduction Crinum defixum Ker-Gawler appears to be the plant which Linnaeus originally cited from his Flora Zeylanicus 127, 1748, as his C. asiaticum. But the literature references back to Hermann's Lugduno Batavus 682, t-683 of 1687 were too confusing for Ker-Gawler to make an identification, provided he examined edition I of Species Plantarum. As was, he selected a secondary reference to Radix toxicaria in edition II to be C. asiaticum L. Later he identified Roxburgh's Belutta-Pola-Taly specimen (Syn. Crinum, foliis carinatis, Fl. Zeyl. 127) to be C. defixum.. Herbert's illustration in Botanical Magazine t-2208 of a bulb which was furnished by Dr. Cary has been accepted as the C. defixum type. Baker cites several species as near C. defixum. These unquestionably are actual geographical variants.

All of the C. series Defixa are confined to southern Asia and Malaysia, and are small swamp or wetland plants, and appear to be the most primitive Crinum. Near all have 1 to 10 subsessile or short pedicelled blossoms with narrow, linear tepals, so some may be nearer subgenus Crinum than Stenaster.

1. Crinum defixum Ker-Gawler in Jour. Sci & Arts iii: 105, 1817; Bot. Mag. t.2301. Hab. So. India. C.asiaticum per Roxburgh in Flor. Indica. II:1128, 1832; syn. Belutta-pola-taly, Rheed, mal. 11: t. 38 fide Ker-Gawler, Rev. Jour. Sci. 3: 105, 1817.

   1. syn. C. asiaticum L. Flora Zeylanicus 127, 1748; ex parte: Spec. Pl. 1: 292, 1753; Spec. Pl. Willd. 2:45, 1799; Osbeck it 94(?)=C. defixum Ker-Gawler 1817.

   2. syn. Amaryllis vivipara Weathers Bulb Book 157, 1911.

2. C. wattii Baker in Handbk. Amary. 76, 1888. Hab. India.

3. C. stenophyllum Baker in Gard. Chron. i: 786, 1881; Amaryl.. 76, 1888. Hab. India.

4. C. ensifolium Rox. in Flora Indica ii: 129; Hooker Flor. Brit. India vi: 281, 1894. (near C. defixum Ker-Gawler)

5. C. serrulatum Baker in Gard, Chron. i: 786, 1881; Amaryl.. 76, 1888, Hab. Cambodia.

6. C. bakeri Schumann in Engl. Jahrbk. IV: 194; Bkr. Amar. 77, 1888. Hab. Marshall Islands.

7. C. woolliamsii Hann. Plt. Life. Herbertia 43 (1): 14-16 w/ f. l, 1987. New Guinea.

8. C. rumphii Merrill in Interpr. Rumph. Herb. Amboina 141, 1917.

Since preparing the above, Nicolson, Turesh, and Manilal in Regnus Vegetabile 119: 272, 1988, have concluded that the Rheede Tanghekolli represents C. defixum Ker-Gawler. Thus, Linnaeus's original C. asiaticum sensu Flora Zeylanicus 127 may be typified by Hermann's Hort. lugduno batavus 682, t.683, 1687. Or by Roxburgh's Plants of Coromandel (1795-1820) with illus. fol. (Roxburgh's specimens were also filed in the Louvain University herbarium at Meise, Belgium). A recommendation may be proposed to validate C. defixum Ker-Gawler as C. asiaticum Linn. while our present C. asiaticum will become C. toxicarium (Rump.) Linn. Spec. Pl. ii: 419, 1763. Linnaeus's lumping of references clouds identification. Ker-Gawler obviously used this confusion by Linnaeus in Species Plantarum II to his advantage.

Since Linnaeus's description of C. defixum as C asiaticum in Flora Zeylanica was the first of the subgenus Stenaster mentioned, it is the type representative of the subgenus.

We find Belutta-pola-taly spelled several ways: B.-p.-talis by Linnaeus, B.-pali-tali by Ker- Gawler in Bot. Mag. t.1073, and B. p. taly and tali by Rheede and Wm. Roxburgh while D. H. Nicolson states the name is derived from polathali in Malayan script.

III. Crinum series Amoena (Roxb.) and allied forms

-- Subgenus Crinum, ex sg. Platyaster, ex parte Asian

Commonly scattered through Tropical Asia and vicinity. Bulbs normally under 6 cm. diameter, globose. Foliage linear 30-60 cm. length. Blossoms subsessile. Tepals patent, lorate to 10-15 mm. wide. Interrelationship of some species uncertain as descriptions are relatively meager. Some forms have long subsurface rhizomes. Several variants grow up to the 2,000 meter elevation in the Himalayan mountains and are hardy. Evidence suggests that C. americanum did not originate from the series Amoena, but from the series Defixa. Thus the above subgenus Crinum's identity may be recinded as misapplied to Asian forms.

1. C. amoenum Roxby. Flora Bengal 23; Flora Indica ii: 127, 1832. Sikkim, India.

   1. var. caudicum Herb. in Amaryl. 255; Gard. Chron. xlii: 62, f., 1907. Ceylon.

   2. var. vericundum Carey ex Herb. Gard. Chron. xlii: 63, f., 1907.

   3. var. mearsii (?) Gard. Chron. xlii: 62, f., 1907; RHS Jour. 33:(Floral Comm. Rept. 177, July 1907), 1908. Burma.

   4. C. humile Herb. Bot. Mag. 53: t-2636, 1826. Habitat India. (Extinct?)

   5. C. gracile Meyer ex Presl. Rel. Haenk. ii: 120. Luzon.

   6. C. cumingii Baker in Gard. Chron. ii: 72, 1881. Hab. Philippines. Geerinich Flora Malesiana sec.1, XI(2):(1990), p.361.

   7. C. stracheyii Baker Gard. Chron. xvi: 72, 1881. Himalayan region

   8. C. pratense Herb. Amaryllidaceae 256, 1837. Hab. India.

      1. var. lorifolium Roxb. Flora Indica II: 131, 1832. Pegu.

      2. var. elegans Bot. Mag. t.2592.

   9. C. woodrowii Baker in Hooker f. Bot. Mag. 124: t-7597, 1898. India.

   10. C. venustum Herb. Carey ex Steud. Nom. Ed. II (i): 440. India.

2. The following Platyaster forms appear to be associated with the above species:

   1. C. balfourii Baker Bot. Mag. 107: t-6570, 1881. Hab. Socotra, Gulf of Aiden.

   2. C. canalifolium Herb. in Bot. Mag. subt.-2121, p. 7; Amary. 259. India.

   3. C. erythrophyllum Carey ex Herb. in Bot. Mag. subt.-2121, p. 7; Amary. 258; Plant life Herbertia 7: 92, 1940. Rangoon, w/ red foliage.

   4. C. esquirolii Leveille in Nouv. Contrib. & C. Chine, 9, 1960; Mem. Pontifica Acad. Romana dei Nuovi Lincei 24: 343, 1906. China.

   5. C. thaianum J. Schulz in Plant Life Herbertia 27: 127, 1971; 28: 33-42, 1972 With figures & map. Thailand. Aquatic species.

      All Asian, Indonesian, and Australian Platyaster and Stenaster Crinum forms (including the C. asiaticum group with its unique elevated Pandanus-like stumps), plus our native C. americanum show a significant divergance from their African counterparts listed in group XI. Their migrations to Asia and America probably occurred during lowered sea levels several million years back. And their widely scattered distribution under varied coastal environmental conditions has encouraged diversified evolutionary modifications which difffer significantly from those few now present in mainland Africa.

IV. Crinum series Australasica (Herb.)--- Subgenus Criuum, ex sg. Platyaster

Plants Platyaster in form, blossoms sessile to subsessile with semi-lorate to lorate tepals, while those inland have broader tepals of a semi-codonocrinum form. The tepaltubes of C. flaccidum are often slightly curved, but the blossoms lack bilateral symmetry so are not Codonocrinum. All are confined to continental Australia and New Guinea. For reasons stated on the previous page, the subgenus Crinum may not apply to this series.

The following series is taken from the Flora of Australia 45: 369-375, 1987, Also Amer. Hort. Mag. 41(4): 224, 1962. C. pedunculatum excluded.

2. Crinum flaccidum Herb. Bot. Mag. 47: sub. t-2121, p.7, 1820.

   1. Syn. Amaryllis australasica Ker-Gawler Bot. Reg. 5: t-426, 1820; RHS Jour. 88(1): 32, 1963.

   2. var. C. arenarium var blandum Sweet, Hort. Brit. 405, 1826.

   3. var. C. corynorrhizum Muell. Fragram. 3: 23, 1862.

   4. var. C. pestilentis Bailey, Queensland Agri. Jour. 2(3): 198 t-17, f.-1, 1898.

   5. var. C. leutoleum Traub & Hann. Plt. Life 21: 96, 1965, comb. nov. Pichi-Richi pass yellow variant. So. Australia.

3. C. angustifolium R. Br. Prodr. 297, 1810; Bot. Mag. t.2355, t.2531.

   1. var. C. australasicum Herb. Amaryllidaceae 259, 1837.

   2. var. C. confertum Herb. Bot. Mag. 51: t-2522, 1824.

   3. var. C. brisbanicum Bailey, Queens. Agri. J. 4: 47, 1899.

4. C. venosum R. Br. Prodr. 297, 1810.

   1. var. C. brevistylum Bailey, Queensland Agri. Jour. 2: 197, 1898.

   2. var. C. intermedium Bailey, Queensland Agri. Jour. n. s. I: 124, 1924.

5. C. uniflorum F. Muell. Fragm. 3: 23, 1862.

6. C. papuaensis spec. nov. Bulbo spherico, breviter columnair; foliis 10, oblongo-lanceolatis, semi-petiolata, erecto-recurvis, margine laevi, vena lineatus, subvenis reticulatis, scapo viridi; spatha 9 flora, floribus subsessilibus, tubo linearibus cylindricas, coralla laciniis patens, lanceolatus, albus, stylus brevis.

   1. A bulb received from Wau (1000 m. elev.), Papua, New Guinea via Waimea Bot. Gdns. for identification: (Specimen No. 75 p.1806/ K. H. 659). Subgenus Platyaster, 8 blossoms sessile, typical of Northern Queensland narrow petaled C venosum; tepals 10 mm. Width x 8 cm. length, patent, lorate with 13 cm. tepaltube uncurved; Foliage, 10 leaves semi-petiolated, lingulate 9 cm. wide x 60 cm. long with pronounced cross-reticular veining typical of African C. jagus (ex C. giganteum). Foliage not typical of Asian forms, while linear tepal-tubes not typical of African C. jagus.. Possible remnant African species. Holotype to be filed U. C. Davis Herbarium.

A breakdown in morphological characters is reported for all but the last of the above species where they are in contact or coexist giving rise to fertile hybrid populations. The entire series is apparently in the transitional state to the Codonocrinum since fertile hybrids have been obtained between C. flaccidum and C. yemense, and C. scabrum.

Note: The Australian C. pedunculatum and associated Stenaster forms are listed under C. series Asiatica.

1. V. Crinum series Ornatae (Herbert)--

   1. Crinum subgenus Codonocrinum Baker ex parte

Wm. Herbert, 1821, defined this taxonomic group as part of his subdivision Ornatae, while J. A. Schultes, 1830, defined it as the Ornata: "Foliis saepieus undulatis, germine sessili". We find C. series Ornatae most fitting, but exclude Herbert's inclusion of C. jagus and allied forms as evolutionary distinct..

As noted in the introduction, this polymorphic group of Codonocrinum, the Crinum series Ornatae, is widely scattered across Tropical Africa and Asia with numerous confusing ecologi­cal and genetic variants, subspecies and species, both named and unnamed. The blossoms are distinctively sessile with the petals keeled a bright pink to wine-red. They open semi--simultaneously and are fragrant. The foliage is long, ensi­form to lorate, arching and often scabrous. The African forms are often semi-aquatic in habit while the Asian are adapted to sandy banks. Nomenclatural difficulties began with Linnaeus, whom was prone to lump species; then by his son, Linnaeus fil., whom unfortunately provided confusing information on identification of his Amaryllis ornata; next by Ker-Gawler and Dryander who attempted to lump all known Codonocrinum together as a single Amaryllis ornata 'species'. Even William Herbert, J. D. Hooker, and I. Nordal had their part in the confusion, although Herbert, Baker, and Hooker clarified far more nomenclatural confusion than they created. The recent trend is to lump ecologically related Tropical West African species and reuse 'C. ornatum' without clearly defining the specific basionym or variants under consideration. This is an inconvenience to horticulturists who readily recognize the variants as distinct species for breeding purposes.

As mentioned above, in 1806 Ker-Gawler attempted to treat all then known members of the subgenus Codonocrinum, which have distinctive bilateral trumpet-shaped blossoms, as a single 'Amaryllis ornata'. This soon lead to botanical confusion. True, there are some general features which suggest that he had a possible polyglot species of Amaryllis. However, William Herbert clarified much of the confusion in 1820 by recognizing all of these 'A. ornata' as Crinum, as had Linnaeus in his 1771 Mantissa. Herbert assigned nearly a score of distinctive species names. But he also noted that there was a large interrelated morphological group from tropical Africa and Asia, and he classed these particular plants as members of his taxonomic subdivision Ornatae. We too, recognize this distinctive grouping, but treat it as a Crinum series Ornatae. All members are derived from a common primitive origin. However, for convenience, for geographical as well as morphological distinctions, we subdivide this series into Asian, East Tropical African and West Tropical African groupings. The Asian forms normally have open funnel-form blossoms and more blossoms per umbel, while the African have a number of cytological types and species, which are inclined to be semi-aquatic in habit. Several of the latter bear only two or three cernuous blossoms. (See O'Brian & Mayr, 1991, on definition of a species, also D. J. Crawford, 1985. I question whether genetically distinct forms should be lumped as a common species as recently attempted.)

Some questions exist over the validity of the Linnaean lectotypes for the Asian C. lati­folium L. and C. zeylanicum L. The issue is whether these two species are actually one, as Herbert and J. D. Hooker both believed, or if they are two distinctive classical species as claimed by Baker. Noticeably, the original Commelin illustration of Lilio narcissus ceylanicus as a lectotype is not in agreement with the Herbert broad-petaled garden forms, which come from the Coromandel Gulf Coast in S. E. India, and is now grown in Florida as C. zeylanicum. While the C. latifolium L., the Sjovanna-pola-tali Rheed Mal. II: p.27, t.33, is best described in Wm. Roxburgh's manuscript, published in Bot. Reg. t.1297, 1830, and his Flora Indica 1832.

The first illustration of a Tropical West African Crinum introduced into England is represented by Ehret's Lilio narcissus africanus, a 'Cape Coast' bulb from Ghana and published by Ehret in his 'depictae' t.5, f.2, (1748),---a near typical C. scabrum. Linnaeus called this plant a questionable 'Beta' form of A. zeylanica in his 1753 Species Plantarum. Like other allied Tropical West African forms, this plant, species-wise, is distinct from the Asian members of the series, as is Ehret's bulb the A. ornata form in Linn. fil's. manuscript. The younger Linnaeus was describing and naming a Marquis of Rockingham Crinum seen at Wimbledon as A. ornata, and in his descriptive phrase name he stated it had widely flared tepals whose apices turn upward giving a concave shape, and that its habitat was Africa. He confused the description in that he cited a polyphyletic similarity to Rumphius' Tulipa javana thinking the latter a possible associate form. He overlooked the fact that his father had cited Tulipa javana in Species Plantarum II of 1762 to be a synonym or subspecies of A. zezlanica (C. zeylanicum). Aiton flowered a C. brousenetti specimen in Kew Gardens in 1785, which was presumed the same as this Linnaeus fil's basionym. The younger Linnaeus' manuscript description was used by Aiton to describe the Kew specimen in the Hortus Kewensis 1: 418, 1789, but the Rockingham source was omitted while the Ehret reference was added, with a question mark. (Note: Linnaeus f. was obviously employed by Banks or Aiton in 1781 to describe some 20 Amaryllidaceae in Kew, including A. ornata, all for the contemplated Hortus Kewensis for King George III. His name as author was appended to the species which he described, while Aiton and Dryander edited his text making several obvious botanical and geographical corrections.)

In 1805 Ker-Gawler and Dryander were still under the erroneous Linnaean view that all known Crinum in the subgenus Codonocrinum consisted of a single polyphyletic species of Amaryllis, which Ker-Gawler then identified as his 'A. ornata' in liew of A. zeylanica in 1806, 1809, 1810, and 1817. This was a nomenclatural disaster and still causes no end of confusion. During this period Dr. Afze­lius introduced some C. distichum into England from Sierra Leone. Ker-Gawler or Dryander erroneously identified this plant as another 'A. ornata'. It was single flowered, but unlike the 1785 two flowered Kew specimen, had distinctive distichous foliage. An Afzelius herbarium specimen, (now British Museum #1), was filed with Banks Herbarium while Ker-Gawler published its plate in Bot. Mag. t.1253 as an 'A. ornata'. Eventually Herbert renamed this plant C. distichum and Afzel corrected the identity on the mounted specimen, as has F.N. Hepper. Obviously this Afzel bulb and the original Aiton Sr. 1785 specimen tend to confound botanists who do not know the historical details of Dryander and Ker-Gawler's 'A. ornata' over-consolidation confusion.

In other words, Herbert, having grown and flowered C. broussonetii recognized both this 1785 Bank's Herbarium basionym and the 1782 manuscript description by Linnaeus f. as cospecific with his C. broussonetii. Like Linnaeus f. and Aiton, he probably considered a C. scabrum (Ehret's illustration) an allied form, but on finding that C. scabrum would cross with C. bulbispermum album, while C. broussonetii and C. yuccaeflorum would not, realized that there were species distinctions (actually both morphologically and cytologically). Baker in 1888 also recognized this 1785 Bank's specimen of A. ornata as the Kew basionym for his C. yuccaeflorum--C. broussonetii consolidation. Seemingly, most other botanists have overlooked this 1785 C. broussonetii basionym for A. ornata Aiton and disregarding Aiton's (or Dryander's) question mark in Hortus Kewensis I, have often accepted Ehret's plate as an erroneous C. ornata lectotype,--J. A. Schultes, 1830, Baker, 1888, and Nordal, 1982, excluded, as they all noted this 1785 Aiton herbarium specimen of C. broussonetii.

The following comments may be repetitious, but most botanists have failed to check the Linnaeus f. manuscript, Herbert excluded, as he examined both the mounted A. ornata 'type specimen' from Kew Gardens, dated 1785, in the Bank's Herbarium, and the Linnaeus f. manuscript's detailed description of A. ornata. Herbert stated that both were cospecific with his own single-flowered C. broussonetii (see Herbert's Appendix pp. 8 & 22, 1821, also Bot. Mag. t.2466, 1824 in text, as republished in the attached compendium). Despite Herbert's cospecific recognition of the Kew type specimen as a C. broussonetii, he made no attempt to transfer this A. ornata Aiton to the genus Crinum.

In 1834, when Mrs. Bury published an illustration of a small six-flowered C. broussonetii form as 'Crinum ornatum', she made a single reference to Bot. Mag. t.1253 showing C. distichum, without citing its association. The opinion is that she was merely adding her specimen to the above Ker-Gawler hetero­geneous collection listed under t.1253, with no specific basionym other than Herbert's Ornatae forms in mind. On the other hand she included properly identified plates of C. scabrum t.32, and C. yuccaeflorum, t.21, in her Hexandria Plantae. Thus, she indicates a general knowledge of Herbert's series Ornatae. And, in a yen for accuracy, she was obviously advised by botanical associates, most likely Lindley, that she had a multiflowered A. ornata Aiton (syn. C. broussonetii) form which should be properly updated to the genus Crinum. We cannot determine the background circumstances more closely now, but her motives were obviously correct.

Several other unanswered questions remain. Herbert states that C. broussonetii bears one to six blossoms (Bot. Mag. t.2121), while A Chevalier, 1950, states that C. yuccaeflorum rarely bears more than two blossoms. Yet, Baker, 1888 lumped C. broussonetii under C. yuccaeflorum Salisbury and cited Aiton's A. ornata 1785 specimen above as the type-species. From then on C. broussonetii vanished from botanical use, despite A. broussonetii Redouté, 1802, being described first which makes it the lectotype. (Under the Paris 1867 de Candolle Code C. broussonetii should have had preference. But Baker was using the Kew Code which gave preference to C. yuccaeflorum since it was first recognized as a Crinum). Due to this consolidation Baker apparently failed to recognize that Mrs. Bury's Crinum ornatum was cospecific with both C. broussonetii and the Aiton A. ornata specimen. Most likely, in his zeal to eliminate all vestigial misuses of 'ornata' and 'ornatum' by both Ker-Gawler and Herbert from British Botany, Mrs. Bury's C. ornatum was included. Under the liberal Kew Code, Baker had no problem identifying a bulb from Sanders Nursery as C. sanderianum and then citing Mrs. Bury's C. ornatum as its lectotype when both were cospecific with C. broussonetii (Redouté) Herb.

But why did Baker fail to note this C. sanderianum interrelationship with C. broussonetii? One of the major problems in species recognition within the African series Ornatae is that various members have quite diverse chromosome counts and thus form several rather distinctive non-intercrossing cytological and ecological races which morphologically may appear to be quite similar to a herbariumist. Several desert populations have blossoms which open quite widely on the first evening while maturing, but by the following dawn the widely exposed tepals have closed down converting the blossoms into slender trumpets while the tepal tubes have become distinctly nutant to cernuous, permitting the blossoms to become completely pendent. In contrast, some of the shadeland species growing along streams like C. scabrum blossoms remain in a upturned stance for several days until they wither and the tepaltubes become nutant. Crinum broussonetii, C yuccaflorum, A. ornata Aiton, Mrs. Bury's C. ornatum, and C. sanderianum, all being cospecific, are in this distinctive cernuous class, as are some of the C. abyssinicum forms. However, the early lectotype illustrations of C. broussonetii and C. yuccaeflorum are drawn with the tepals shown fully open, while Bury's plates of C. ornatum and C. yuccaeflorum, are shown with second or third day nutant blossoms. This floral behavior would appear normal for one experienced with years of growing the plants, like Herbert or Mrs. Bury, but since this nutant or drooping behavior is not cited with plant descriptions, a herbarium botanist like Baker would be quite unaware of the habit and conclude that he had a distinct species. Yet, this is a significant factor in separating the lesser Ornatae forms from C. scabrum and its allied types.

Baker's 1888 nomenclature remained fairly stable until J.K. Morton, 1961, and F.N. Hepper, 1968 recognized the C. sanderianum Baker as the C. ornatum (Aiton) Bury. Hepper cited the Ker-Gawler Bot. Mag. t.1171, 1809, plate of C. zeylanicum with the extensive Ker-Gawler synonym list as a means of interrelating C. sanderianum to A. ornata Aiton. His logic is confusing since both are C. brousonetii. The specific basionym was not cited, nor is the type specimen. Hepper's epithet is only valid if A ornata Aiton was the specimen grown by the Marchioness of Rockinghan which Linnaeus f. named the A. ornata.

Next, we have I. Nordal and associates, whom in 1978 attempted to reconsolidate all of the African C. series Ornatae into a single species: "C. zeylanicum (L.) L. in much the same manner as Ker-Gawler. Unfortunately her decision was premature as their subsequent field and cytological studies reported in 1982 revealed several obvious cytologically and morphologically distinctive species scattered about Tropical East Africa., and her photographs show nutant forms in the Cameroons. Meanwhile, she became aware of Aiton's herbarium specimen which she recognized as the holotype for A. ornata Aiton, 1789, but no mention of C. broussonetii. We cannot accept her C. zeylanicum (L.) L. based upon Linnaeus' Ehret Depictae t.2, f.5. In 1983 Wijnands corrected this by citing the Moninckx plate tab. 30 as a valid lectotype for C. zeylanicum (L.) L.

Finally, by extended field and herbarium studies in 1995-96 Dr. David Lehmiller issued a comprehensive report on Crinum in Tchad and adjacent Tropical West African areas. He consolidated several species as mere variants of C. brousonetii and C, yuccaeoides. This was reported in the 1997 International Bulb Societies Herbertia. It settles near 200 years of confusion. After preparing the above text, Dr. Lehrmiller has proposed that A. ornata Linn. f. be recognized as C. ornatum (Linn. f.) Herb. comb. nov. since Linnaeus f. refers to Rumphius' Tulipa javana as a synonym. Linnaeus sr. had already recognized T. javana Rump. as a valid variant of A zeylanica, (syn. C. zeylanicum). Evidence furnished by the Linnean Society indicates that the Marchioness of Rockingham specialized in growing exotic plants, while Linnaeus f. identified and named several bulbs for her, including A. ornata and A. fucata (unidentified).

In 1967 Jones and Smith made a comprehensive study of the somatic chromosomes in Crinum which were available from both Kew specimens and published records. As stated, their series Ornatae disclosed an unanticipated range of genetic abnormalities whose presence is not suggested by many morphological features other than environmental adaptations or flower number. As an example C. zeylanicum has a 2n = 22 count while a Kew C. scabrum form had 44. Their 'Crinum ornatum', so called, from Coastal Ghana and Nigeria indicated counts of 24, 44, and 87, while a Sierra Leone form of 'C. ornatum', probably near the Aiton's 1785 bi-flowered basionym was 22. Crinum abyssinicum is also a 2n = 22 form, while Nordal and Wahlstrom, 1980, 1983, reported triploid forms in the Camerouns, and both 2n = 30 and 33 forms in Kenya. And I have been informed that C. yuccaeflorum may be 2n = 20.

Past hybridization attempts, both intra and outcrossings, have indicated numerous unex­plained incompatabilities and irregularities with the above cited forms as well as with other variants which have not had cytological examinations. It is apparent that these tropical forms commonly indicate genetic complexities, or irregularities which appear near unknown amongst the other subtropical Crinum series where 2n = 22 is normal. As a consequence lumping of the series Ornatae forms as a single species is ill advised where they have relatively similar morphological but different environmental habitats, without undertaking cytological studies to establish the existence of relationships. And based upon the distinctive nutant or non-nutant blossom characteristics amongst the C. series Ornatae we question the accepting of Nordal's 1980 broad nomenclature for Scabrum forms without further cytological and morphological study.

According to the 1988 Botanical Code the Crinum subgenus Codonocrinum is represented by C. zeylanicum (L.) L. only. However, we find that I. B. Ferrarii in his 1633 Florvm Cvltvra Libra IV, pages 115 and 309, cites and describes C. latifolium as Narcissus indicus lato folio (Syn. Narcissis i. é rubro croceus flore liliaceo) which Linnaeus copied in brief on page 292 in Species planrarum I. This establishes C. latifolium's prior name source, validity, and supra standing.

V-A. Crinum series Ornatae (Herbert)

-Subgenus Codonocrinum. Asian, ex parte

Asian portion of the complex originally recognized as a single species by Herbert, 1837, subdivided by J.G. Baker and reconsolidated as follows by Hooker under C. latifolium., Flora of British India 6: 283, 1894, Codonocrinum subgenus. Hooker considered Rheed's description as having priority.

Crinum latifolium Linn. Sp. Pl. 1: 291, 1753; Bot Reg. t-1297, with plate,

1830; Roxb. Flor. Indica ii: 137, 1832; Sjovanna-pola-tali, Rheed,.

Mal. II: 77, t. 39, 1678, lectotype; Hooker Flora Brit. India 6:783, 1894;

Merr. Enum. Philipp. Flow. Pt. 1(1925) 210; Herb. Amar. 253, 1837.

Narcissus indicius lato folio Ferarii 1633 Florvm cvltvra p. 115.

C. zeylanicum L. Mantissa 1771; L. Syst. Vege. 263, 1781; Roxb.

Flr. Indica ii: 138, 1832; Hortus Bengalis 1814. Bury Hex. pt. t-29, 1834.

Amaryllis zeylanicus 'supra' Linn. Sp. Pl. 1: 293, ex part;

Commelin, Hort. Amst. 1:73 t.37. 1697, syntype.

Lilio Narcissus ceylanicus Moninckx's Atlas I: t.34, 1689; lectotype,

by D. O. Wijnands 1983, Bot. of the Commelins p.37, t.3..

C. zeylanicum Roxburgh Plants of the Coast of Coromandel 1795-1820.

A. latifolia L'Herit. Sert. Ang. 14, 1788; Ker-Gawl. Jour. Sci. & Arts.

3: 114, 1817, ex parte.

A. ornata 'Supra alpha' Ker-Gawler ex parte, Bot. Mag. t-1171, 1809;

Jour. Sci. & Arts 2: 243, 1817. = var. C. zeylanicum (L.) Herb. (Coromandel form).

Tulipa javana Rumph. Herb. amboina 5: 306, t.105; syn. C. zeylanicum (L.) L.; C. moluccanum (Roxb.) Herb,; C latifolium L. sensu J. D. Hooker Flora British India 6: 783, 1894. (Herbarium specimen filed under C. latifolium in B. M.)

var. insigne Schultes Systems 7: 899, 1817-30; Bot. Reg. t-597.

A. insigne Ker-Gawler, Bot. Reg. 7: t-579, China.

var. speciosum Herb. Bot. Mag. 48: t-2217, 1821.

var. herbertianum Wall. Pl. As. Rar. ii: 35, t-145, 1830; Herb.

Amaryl. 263, 1837. Syn. C. wallichianum Roem. Syn. Ensatea 85.

Syn. C. linneai Roem. Syn. Ensatea 80. A. lineata Lam. Encyc. 1: 123. 1783.

syn. C latifolium Weathers Bulb Book.

Syn. C. ornatum (Ker-Gawler) Herb. Bot. Mag. 29: t-1171, 1809, ex parte; Herb. Amaryl. 263, 1837. (Rejected by J. G. Baker, 1781).

var. latifolium Herb. Amaryl, 263, 1837.

var. zeylanicum Herb. Amaryl. 262, 1837; Bot. Mag. t.1171, 1809. (Common Garden

form from Coromandel Gulf, India, 2n =22).

var. careyanum Herb. Bot. Mag. 51: t-2466, 1824; Weathers Bulb Book 156, 1911. Mauritius.

var. moluccanum Roxb. Fl. Indi. ii: 140, 1832; Miq. Fl. Ned. Indie. 3 (1859) 581;

Bot. Mag. 49: t-2292, 1822.

Syn. C. asiaticum Wall. Cat. 8969 B, expt.

var. longistylum Herb. ex Steud. Nom. Ed. II (i): 440; Amaryl. 264, 1837. Bengal.

D. H. Nicolson in Interpretation of Rheed's Hortus Malabaricus p. 273, 1988, has difficulty with the identification of Sjovanna-pola-tali Rheed as C. latifolium L.. Unfortunately Roxburgh's Flora Indica 2:137, 1832 has been bypassed, wherein Wm. Carey, knowing many native Asian languages, identifies, describes, and validates Sjovanna-pola-tall as the C, latifolium L. Dr. Carey has long been overlooked as a most capable field botanist who worked in close relations with Roxburgh and specialized in collecting Crinum, many which he shipped to Wm. Herbert. (See Roxburgh's Flora indica, 1832 which Cary published from the Roxburgh manuscript. Also Weathers Bulb Book p.150, 1911, Kew).

V-B. Crinum series Ornatae:

African Portion.

Forms cited below are usually considered morphologically and genetically closely inter­related. Thus those listed in lighter itallics are basically considered synonyms or subspecies. Many unrecorded transitional variants exist. However, the consolidation of the follow­ing classical species is open to question without determining the environmental as well as somat­ic chromosome relationships. Consequently the apparent lumping of C. scabrum and Guinea Gulf subcernuous and biflowered forms as an undefined 'C. ornatum' as attempted by Andrews, 1956, and F. N. Hepper in 1968, is open to question as to the actual relationships involved. Herbert, 1837, found that his C. broussonetii form and C. yuccaeflorum refused to outcross with C. bulbispermum album, whereas C. scabrum yields the well known C. x herbertii quite readily. Obviously genetic distinctions exist. Basically we can divide the West Tropical African Ornatae into several slender few-flowered cernuous types and the multiflowered open blossomed C. scabrum forms as established by Jones and Smith, 1967, namely:

Sierra Leone, 2n = 22

Various chromosome Ghana, Accra Plains, 2n = 24

counts, tropical Ghana, N. of Yendi, 2n = 44

West Africa Ghana, Near Bole, 2n = 66

Ornatae Nigeria, Jos Plateau, 2n = 87

We presume that the Aiton 1785 Kew specimen from Capecoast, Ghana, could have been an Accra Plains form while Herbert's specimen was from an 1806 Sierra Leone import. Jones and Smith cite C. scabrum as 2n = 44, but most hybrids suggest numerous 2n = 22 forms are in existence.

As stated previously, C. yuccaeflorum Salisbury was recognized by both Herbert and Chevalier as distinct. Likewise, most botanists are near universal in their opinion that C. distichum is a distinct species, but it appears related to the South African C. rautanenianum Schinz and C. euchrophyllum Verdoorn which have similar distichous foliage. Several Crinum forms resembling C. fimbriatulum have been reported in Angola and about the swampy headwaters of the Zambezi River. These near unknown forms have been associated with C. scabrum, C. kirkii of Tropical East Africa, and even C. zeylanicum of India. Excluding C. scabrum, none of these tropical and south-tropical African forms in the C. series Ornatae appear hardy enough to survive severe winters in Florida, although some sterile hybrids suggest C. fimbriatulum and C. broussoneti have been crossed with hardier Cape forms back in 1920, the 'Empress of India' being one example, the 'Jamaica Lily' another.

One would assume that after a hundred years of miscellaneous cross breeding with C. scabrum and its allied forms that a number of intraspecific fertile hybrids should exist. At best we can cite some intraspecific crosses with C. abyssinicum or C zeylanicum and outcrosses with C. moorei, C. bulbispermum, and C. americanum as relatively hardy. All yield pollens which are semi-active and capable of backcrossing onto parental forms. On one rare occasion a C. x herbertii (Syn. C. scabrum x C. bulbispermum) yielded several seed which resulted in giant forms of the F-1 hybrid. These mutations were semi-fertile and obviously polyploids. On other occasions C. x herbertii variants may produce an occasional typical F-1 maternal seedling, and in some instances C. scabrum pollen will cross on some C. x herbertii.

The general conclusion is that the Tropical East African forms of the C. series Ornatae, including some of the Ethiopian forms, have all evolved from a protoform which had experienced a major form of reciprocal translation. Thus subsequent populations have experienced additional translocations and heteroploidy. As a result we have a number of genetically isolated species which are often mutually incompatible, but in some instances will outcross with unrelated species having a near normal 2n = 22 chromosome composition.

Crinum series Ornatae (Herbert) ex parte, Appendix p. 22, 1821:

C. broussonetii (Red.) Herb., Bot. Mag. t-2121, 1820, Sierra Leone.

A. broussonetii Redouté Les Liliacees t-62, 1802. Sierra Leone. (Holotype) = A. ornata Aiton sensu Herb. 1821.

(Syn. A. ornata L. f. Ms 1782 sensu Herbert, 1821= A. ornata 1785 Kew specimen?

See Herb. appendix p. 8 & 22, 1821, in accompanying compendium.)

C. broussonetainum Herb. Amar. 260, t-44, f.-13, 1837. Sierra Leone. = C. broussonetii.

C. sanderianum Baker, Gard. Chron. 1884 (2): 102, 1884. So. Cameroun.

Syn: C. ornatum Bury; Hex. pts. t.18, 1834. J. Weather's, 1911,

Bulb Book, 162, Sierra Leone.

Syn. Amaryllis ornata Aiton Hort. Kew ed. 1: 418, 1789; also Kew '1785 specimen' in British Museum, identified as C. broussonetii Herbert, 1821; (type-specimen

sensu Baker, Amar. 1888, p. 89-90). Holotype.

Syn. A. ornata ex Aiton. Willd. Sp. Pl. III: 55, 1799.

Syn. A. ornata (alpha var. uniflora) Ker-Gawler ex part, Bot. Mag. 31: t.1253, 1810.

(Hort. Kew I ref. only. Exclude fig. and descript.)

Syn. C. ornatum (Aiton) Bury sensu Hepper, Fl. W. Trop. Afr. ed. 2, 3(1): 134. 1968.

A. ornata Linn f. 1782 manuscript. (See diagnostic phrase name.)

C. ornatum Bury, Hex. pt. t-18, 1834. Guinea.(?): C. sanderianum Baker, Amat: 90, 1888. Ghana, Tchad. Also Camerouns fide Nordal, 1980. = C. broussonetii.

C. yuccaeflorum Salis. Prad. London 2: 52, 1805. Sierra Leone. = C. broussonetii.

C. yuccaeides Bot. Mag. t.2121.6, 1820; Treatise p22, 1821; Bank's Herb. A yuccaeides.

C. yuccaeflorum Salis. (Note. Species 1 to 2 flowered.) = C. broussonetii.

C. scabrum Herb. Bot. Mag. 47: t-2180, 1820. Guinea to Ethiopia, Native to moist areas;

Lilio narcissus africanus Ehret depict. t. 5, f. 2, 1748. Also Bury Hex. Pts. t. 32, 1834.

(2n = 22, 44). Known as Lilios rajados in Brazil. lectotype,

C. ruppelianum Fresen in Vehr. Befored Garten b. iii:265; Bot. Mag. t.2180.

P.p. Amaryllis zeylanica 'Beta' Linn. = Lilio Narcissus Africanus Ehret, 1748, Depictae rar.

tab. 5, f.2; Linn. Sp. Pl. i: 293, 1753. Gulf coast, Ghana. = C. scabrum Herb. ex parte.

A. liturara Reich. Icon. des. Pl. I: t.82, 1822; Plt. Lfe. 33: 36, 1977.

C. tanganyikense Bakr. Dyer Fl. Trop. Afr. viii: 400. Lake Tanganyika. (?)

C. boehmii Baker, Bull. Herb. Boiss. Ser. II, iii: 66, 1903. Tanzania.

C. corradii Chiov. Webbis 8: 6, 1951.

C. fimbriatulum Baker, Jour. Bot. XVI: 196, 1887. Syn. Nov. Angola.

C. toxicarium Chev. non Roxby. Bull. Sci. Fr. 2 (8): 212, 1912; La. Soc. Hort. Research III

(5): 321, 1970-71. Sudan, Chari Cent. = C. yuccaeflorum Salis, fide Chev. 1950.

C. doriae hort. Sprenger Soc. Tosc. Ortie Bull. 19: 18-24, 1894. Ethiopia.

C. ornatum (L. f.) Bury sensu F. W. Andrews Fl. Pts. Sudan III: 287, 1956;

C. scabrum forma? (A questionable species.)

C. distichum Herb. in Bot. Mag. sub, t-2121, p. 7; t-1253; A. Afzelius: Genera Plantarum

guineensum revisa et aucta (1804); Nordal & Wahlstrom in Adansonia ser. 2, 20 (2): 188,

1980. See Brit. Museum #1 Herbarium Specimen; holotype*, Sierra Leone, Tchad, Ghana,

2n = 20.

Syn. A. ornata Ker-Gawler non Linn. f. Bot Mag. t.1253, 1810, inc. fig. excl. synonyms.

Non C. biflorum Rottboll sensu Mabberley, 1948, see under XI, Platyaster.

C. Pauciflorum Baker in Jour. Botany 195, 1878; Hand. Bk. Amaryl. 88, 1888; Chev. Bot. Appliq. p. 620, 1950.** Trop. West Africa. (Broad tepals).

syn. C. rautanenianum Schinz?

According to the Manuscript of Linnaeus f. the A. ornata Linn. f. was being grown in Lord Rockingham's garden at Wimbledon, and had been imported from Capecoast, Guinea (now Ghana). The geographical source took precident over Linnaeus fil's unreliable reference to Rumphius' Tulipa javana as a synonym form since Linnaeus f. was geographically ignorant. On the other hand Lord Rockingham, as Prime Minister was familiar with the British colonies, Ports of Call, and world geography. We take his word as to the source.

For clarification , the synonyms and variants for C. broussonetii (Red.) Herb. are as follows: Amaryllis broussonetii Redouté, A. ornata Aiton, C. yuccaeflorum Salisbury, C. ornatum (Aiton Bury, C. sanderianum Baker, and C. zeylanicum (L.) L. sensu Nordal. In turn, C. yuccaeides (Thompson) Herb. has the folowing synonyms: A yuccaeides Thomp., C. yucccaeflorum Chevalier and C. zeylanicum (L.) L. sensu Nordal. Several of the above variants may be classed as biflowered.

In 1961 J. K. Morton, then F. N. Hepper, elected to recognize C. ornatum (Aiton) Bury as a valid name for C. broussonetii and several unnamed associates based upon Ker-Gawler's Botanical Magazine t.923 and t.1171 listing of ornata forms, namely Aiton's A. ornata, Ehret's Lilio narcissus Arficanum (a C. scabrum form), A. broussonetii Redouté and others. The basonym was not cited. The naming of other species (several) as A. ornata by Ker-Gawler clouds the use.

In reviewing the biflowered Crinum: C. distichum, C. humile (Syn. C. nubicum), C.. yuccaeflorum, Aiton's A. ornata and others, we note a unique cytological interrelationship to several other Subtropical biflowered South African species, namely C. pauciflorum Baker, C. piliferum Nordal, C. rantanenianum Schinz, C. minumum M. -R. & S., , and C. acaule Baker. Further study is necessary to clarify this groups diversified interrelationship. We suggest a subseries status of C. ornatae pauciflora, as follows.

Crinum subseries Ornatae pauciflorum

C. euchrophyllum Verdoorn in Flow. PL. Afr. 42: t-1642, 1972.

C. rautanenianum Schinz Consp. Flora Africa V:250, 1895; Herbertia 48(1&2):

82, 1992. Trop. Africa; Bull. Herb. Boiss 4, app. III: 48, 1896.

Herbertia 52:62, 1997. (May be allied to C. distichum Herb. in V-B)

C. minimun Milne-Redhead Kew Bull. 1947: 33, 1947; Fl. Pts. Afr.

40: t-1577, 1969, Norweg. J. Bot. 26: 149-156, 1979; Herb. 52:101, 1997. Rhodesia.

C. acaule Baker in Fl. Cap. 6: 532, 1897; Herbertia 48(1&2): 81, 1992, w/fig. Zululand.

Syn. C. parvibulbosum Dinter ex Over. in Mitt. Bot. Much. i:

444, 1954. (Separate species sensu Nordal 1977).

C. walteri Overkott in Mitt. Bot. Munchen 1: 444, 1954.

C. minimum Solch non M. -R. & S. in Prod. Flora S. W. Africa 150: 4-8, 1954.

C. minimum Verdoorn non M. -R. & S. Bothalia 11: 40, 1973.

C. minimum Rosseler, non M. -R.. & S. Mitt. Bot. Munchen 11: 523-544, 1974;

Herbertia 48(1 & 2): 81, 1992.

C. giessi Lehmiller Herbertia 52: 50, 1997; C. Harmsii Baker affin.

C. harmsii Baker in Warb. Kunene Sambesi Exp. 565, 1903; Affin, C. graminicola

Solich. non Verdoorn Herb. 52: 53, 1997.

C. nubicum (A. Chev.) Hannibal in La. Soc. Hort. Research III

(5): 315, 317, 1970; Bull. Soc. Bot. Fr. 134, Lettres Bot.

1987 (2): 201. Southern Sahara, Guinea, now Ghana, to Chad.

Syn. C. humilis A. Chev: Rev. Bot. Appliq. 30: 620, t-29, 1950; w/ fig.

Nordal Norw. Jr. Bot. 26: 150, 1979; Herbertia 52: 49, 1997.

Rite: C. humile A. Chev. non Herb. Fl. W. Tr. Afr. 3 (1): 126, 1968.

Syn. C. humile Hepper Fl. W. Trop. Afr. ed. 2 (3): 136, 1968.

non. C. humile Herb. in Bot. Reg. t-2636. (subgenus Platyaster)

C. parvum Baker in Kew Bull. 1897: 284, 1897. Trop. W. Africa.

Transferred to Ammocharis, by M.-R. & S. (Ammoc. heterostyla).

C. piliferum Nordal in Norweg. J. Bot. 26: 149-156, 1979. (umbel often subterrestrial,

as with C. acaule).

C. majakallense Baker in Dyer Flr. Trop. Afr. vii: 399, 1898.

(Foliage slender, 3 blossoms, allied to C. yuccaeflorum (?)).

In the 1979 Norweg. J. Bot. 26: 149-154 I. Nordal identified C. piliferum as a new species from Kenya having very slender foliage and tepals, and allied to C. nubicum (Chev.) Hann. (Syn. C. humile Chev. non Herb.) ---Also allied to C. minimum Milner-Redhead, C. acaule Baker (? syn. C. parvum Baker), and C. parvibulbosum Dienter ex Overcutt. ---While Crinum rautenanianum (syn. C. euchrophyllum Verdoorn) is obviously a variant of C. distichum Herb. Most sessile flowered members cited above appear allied to the C. series Ornatae, ---and the biflowered to a distinct subseries of same.

Crosses of C. scabrum variants (all 2n=22) by Mrs. D. C. Sheppard have resulted in vigorous near evergreen intraspecific free-seeding hybrids of exceptional vigor. These have been used on several successful outcrosses.

Crinum biflorum Rottboll in Plant Hort. Univ. Prog. 1773: 32 is placed in Group XI under platyaster. It is not a C. distichum fide Mabberley, 1984. It does not comply with the British Museum #1 type specimen. It may be a Crinum modestum Baker as the tepal tubes are abnormally short.

Note: A. spectabilis Andr. Bot. Rep. 6: 390, 1804 from Sierra Leone is a C. jagus variant. Ker-Gawler and Herbert both listed it with the above Ornatae group in error.

The Tropical East African area is considered the original source of the C. series Ornatae. According to Kingdon's 1989 Island Africa this area has experienced a number of distinct geological and ecological changes, particularly along the Great Rift. This have resulted in numerous morphological and cytological induced evolutionary changes among many of the Central East African flora, Crinum included, where the most obvious adaptations have been from the uplift of Ethiopia. The Congo River, and a prehistoric Nile River through Tchad have distributed various Tropical East Africa Crinum westerly. Isolated colonies over this vast land area have encouraged a number of divergent genetic variants. Similarly, diploid forms migrated to the east by ice age exposed land bridges and sea giving rise to C. yemense and the various Asian Ornatae members. Crinum yemense, and the Ethiopian high desert forms, C. abyssinicum and C. schimperi have all developed excessively large near spherical seed. Ethiopian features appear among the Asian C. zeylanicum which suggests a fairly recent interrelationship.

Crinum Kirkii Baker, Bot. Mag. 106: t-6512, 1880. (2n=22). Zanzibar.

C. politifolium Wahlstrom in Nord. Jour. Bot. 2: 465, 1982.

Tanzania. (2n=30 chromosome count).

C. zeylanicum var. reductum Baker Gard. Chron. 1883 ii: 618. Zanzibar.

C. massaianum N. E. Brown in Kew. Bull. ii: 100, 1888. Zanzibar.

C. abyssinicum Hochst. ex A. Rich. Tent. Fl. Abyss. ii: 311; Schimp. Fl. Abyss. no. 1347; Gartenflora---'Garten and Blumenkunde' 41: Abb. 89, 1892. (2n = 22). Ethiopia.

C. abyssinicum sensu F.W. Andrews Fl. Pts. Sudan: 287, 1956. (tepals w/ red midrib.)

C. schimperi Schumann by Baker. Curtiss Bot. Mag. 121 new ser.:

t-7417, 1895. ( a major form of C. abyssinicum) Ethiopia.( a polyploid).

C. verdoorniae Lehmiller spec. nov. Int. Bulb Soc. Herbertia 48(1&2): 75, 1992; 52, 57, 1997.

Syn. C. kirkii sensu Verdoorn in Bothalia 11(1&2): 1973. Caprivi Strip, S. W. Africa.

Morphologically and genetically the following species is closely allied to C. abyssinicam, and the Asian C. latifolium, as well as the African portion of the series Ornatae, except like C. abyssinicum and C. schimperi it lacks the pigmented midstripe to each tepal. However, the bulb and tall 30 cm. leafy stem show pigmentation. The foliage is tall, 100 cm. long x 6-8 cm. wide, semierect and winter dormant. This desert Crinum is known for its huge seed, 3 to 6 cm. diameter. All appear to be diploids.

C. album (Forskael) Herb. Herbertia 50: 20, 1994.

Syn. C. yemense Deflers, Voyage to Yemen, p. 209. 1889. Syn. C. jemense Hort.

Syn. C. latifolium var. yemense Parey, Blumengarteneri 328, 1931.

Bot. Reg. T.1297, 1830(?).

Syn C. latifolium album Hort. sensu Kew Herb. index 1915.

Syn. Amaryllis alba Forskael, 1775, Flor. Aeg. & Arabia 209; Herb.

Amary. 272, 1837.

Crinum abyssinicum and C. schimperi are morphologically mere close variants of C. album and should be so recognized. The distinctions appear genetic.

For years I have grown four forms of C. abyssinicum and C. schimperi which differ slightly in plant and blossom sizes, and depth of brown pigmentation about the umbels. The flowers, normally 10 to 12, are pure white. Crinum schimperi appears to be a polyploid due to the larger size with its distinctive cernuous blossoms (much like those of C. sanderianum), and large elongated bulbs. This latter outcrosses with tetraploid C. moorei and yields huge white hybrids identified as 'White Mogul'. No significant difference is evident between the C. abyssinicum and C. yemense. The latter is almost pigment free about the umbel while the former is distinctly pigmented. Both obviously come from a common origin. The ribbed foliage of C. verdoorniae suggest a relationship to both C. moorei and C. schimperi. (Note, the Fleischmann illustration of C. abyssinicm in http://muse. bio.cornell.edu./delta/angio/www/amaryl138.gif has a misleading blossom suggesting lorate, not elliptical tepals).

VI. Crinum series Umbratica*

Subgenus Codonocrinum ex parte.

Crinum moorei is one of the few Crinum species adapted to a relatively dry environment where the bulbs thrive in a well drained soil in part shade. Its blossoms are pedicelled and open sequentially. The species is common to Natal. The variants produce few seed and seemingly fail to intercross. It is my opinion the plants are so homozygous genetically that all are near self incompatible, and that most seed produced is purely maternal. A high yield of chlorophyll abnormalities occur. Better seed yields occur with outcrossings. In most F-1 outcrosses the C. moorei characteristics often predominate, particularly with C. americanum.

C. moorei Hooker. Bot. Mag. 100: t-6113, 1874; 104: t-6371, 1878. Natal.

var. makoyanum Carrier in Bakr. Hand. Amaryl.. 93.

var. platypetala Hort.; C. colensoi Hort.; C. natalense Hort.;

var. roseum Hort. C. mackenii Hort.

var. schmidtii Regal in Bkr. Hand. Amary. 93; Plant Life 24:

39, 1968. (Syn. C. schmidtii Regal.)

var. rubra Hannibal subsp. nov. A Gordon McNiel discovery from the Transvaal.

var. tetraploideum Hannibal, La. Soc. Hort. Research Bull. III

(5), 317, 1970-71. Common in trade. Accepts polyploid pollens.

C. imbricatum Baker in Handbk. Amary. 92. (non C. jagus var sensu Baker),

Tepals shorter, broad, pink.

* Crinum moorei, a subtropical Codonodrinum with pedicelled blossoms, is adapted to shaded conditions, thus the series name Umbratica, meaning 'shade loving'. Like all pedicelled crinum forms the umbels have a sequentially--flowering spiral-cyme determinate structure as illustrated by the appended Muller-Doblies' plate. In contrast, sessile--flowered Crinum like C. scabrum have indeterminate umbels where the spiral cyme structure has grouped blossoms which flower in sets on several-day intervals. Genetically, the Pedicelled--Determinate and Sessile-Indeterminate forms are quite distinct. This accounts for much of the ecological adaptation of various species or series, as well as sterility amongst many outcrossed sessile x pedicelled hybrids. In some instances these hybrids will backcross with parental forms giving semi fertile stock.

Crinum x 'Ellen Bosanquet' is a C. scabrum x C. moorei cross which readily accepts either C. moorei or C. scabrum pollens under humid conditions and produces near throwbacks or recurrent hybrids representing the pollen parent involved, but with fully pigmented red blossoms in either instance. The conclusion is that during meiosis the female moorei and scabrum gametes separate out as near wholes giving semi pure-line genomes. Thus, when these maternal gametes fuse with the paternal C. moorei a near typical C. moorei is attained,-or a near C. scabrum forms come with Scabrum pollen. Such backcross seedlings are called recurrent parental types, and usually contain a few bridged-in scabrum or moorei genes, particularly pigmentation. These crosses provide hybrid vigor since the resulting seedlings bear full heads of seed. Crinum moorei from the wild are normally so homozygous that they set very few seed. Crinum x 'Ellen Bosanquet' pollen seems semi-functional. These backcross results suggest most C. scabrum are diploids, contrary to Kew reports. In contrast, wide, unrelated outcrosses often develop hybrids with catenated chromosomes where one set acts as a single dominant entity in practically all features. Thus C. moorei's features are phenotypically dominant in all C. moorei x C. americanum outcrosses. Such wide outcrosses are seed and pollen sterile.

VII. The Crinum series Jagi (Thomp.) --

Subgenus Codonocrinum Ex parte.

The members of this clade are known for their large broad-faced fragrant white blossoms and unique reticularly veined foliage. Those forms, which grow in wet locations along the Ivory Coast, Congo, and Angola have erect well petioled foliage while the inland savanna forms as in Nigeria and Uganda have erect near saber-like foliage. The blossoms are sessile and several buds may open simultaneously. Ivory Coast forms were taken to Surinam and Brazil in 1600-1650 when Africans were imported to work in the cane fields. The bulbs were known for their medicinal value and were carried as charms. They spread as escapes. Ingar Nordal considers all known African C. jagus forms a single species. Such obviously includes the South American escapes which have been named erroneously as native species. Crinum gigantea is the older name encountered in the literature. Chromosome counts show 2n = 22 plus B chromosomes which account for numerous sterile forms. Aquatic forms from Zaire (Belgian Congo) have been found quite cool climate adaptable but experience foliage damage above 38 deg. Celsius

Crinum jagus (Thompson) Dandy. Jour of Bot. 77: 664, 1939.

(Subspecies and synonyms:)

Amaryllis jagus Thomps. Bot. Dispt.: t-6, (1798), from Sierra

Leone. = C. jagus (Thompson) Dandy. Holotype.

Crinum giganteum Andr. Bot. Rep. 3: t-168, (1801). Sierra Leone.

A. spectabilis Andrews Bot. Rep. 6: t.390, 1804. Sierra Leone.

A. ornata 'Beta' Ker-Gawler Bot. Mag. t-923, (1806).* Sierra Leone.

C. petiolatum Herb. Amar. 260, (1837).

C. vanillodorum Wild. ex Baker Jour. Bot. 17: 196, (1878). Angola.

C. podophyllum Baker, Bot. Mag. 106: t-6483, (1880). Old Calabar.

C. laurentii Durand & De Wilder Rev. Hort. Belg. 23: 97, (1897). Zaire.

C. congolense De Wilder, Miss. E. Laurent. (19-3-4) 1: 370, (1905-7).

C. sauveolens A. Chev. Mem. Soc. Bot. Fr., 2(8): 212 (1912).

Iv. Cst.; Bot. Appliq. 30: 623, Fig. 32, 1950.

C. bequaertii De Wilder, Pt. Bequaert 1: 46 (1921). Zaire.

C. nobile Hort. ex Gentil, Pt. Cult. Serres Jard. Bot. Brux. 65 (1907).

C. spectabile extorre Herb. Amar. 261 (1837). Brazil. Ssp. comb. nov.

C. virginieum Mart. in Schultes Syst. 7: 850 (1830). Brazil, Ssp. comb. nov.

C. brasilenese Traub, Plant Life 22: 45 (1966); 23:

70, t.27(1967). Ssp. comb. nov.

C. amazonicum Rav., Pt. Life 37: 81 (1981). Ssp. comb. nov.

C. surinamense Rav., Pt. Life 37: 81 (1981). Ssp. comb. nov.

C. jagus (Thomps.) Dandy, Jour. Botany. 77: 664 (1939).

C. rattrayii Hort. Gard. Chron. 38: 11 (1905). Lake Albert, Uganda.

C. scillifolia A. Chevalier (1912) Ivory Coast Form.

(separate sp. Syn. C. scillifolia Hort Angl.?)

*The early day distribution of C. jagus and other tropical Crinum into South America, Mauritania, India, and other locations has caused considerable confusion, particularly with Ker-Gawler who concluded that all Codonocrinum with broad white petals were the 'Beta' Amaryllis ornata Linn. f.. Herbert failed to clear this confusion until well into 1830, but he still failed to associate his C. spectabile specimens with the early descriptions of A. gigantea and A. jagus Thompson. The literature is cluttered with C. jagus synonyms listed by J. G. Baker.

(Herbert in his Amaryllidaceae, pages 260, 342, 370, (1837) confuses Cape Coast, Ghana, with Sierra Leone. He obviously is referring to the Guinea Gulf, not the old port of Cape Coast, in his combined statements that C. bulbispermum refuses to cross with Cape Coast kinds, namely C. broussonetii from Sierra Leone, and C. petiolatum or C. spectabile (both of the latter are C. jagus forms) while giving fertile seed producing hybrids with C. pedunculatum, C. canaliculatum, and C. defixum).

C. glaucum A. Chevalier. Mem. Soc. Bot. Fr. 2(8): 212, 1912; Hepper Fl, W. Trop. Afr.

2nd ed. 3: 136, 1968; Nordal & Wahlstrom. Adansonia Ser. 2, 20(2): 185-186, 1980.

"A broad leaf swamp plant with 7-15 sessile white bilateral trumpeted blossoms, quite fragrant. Placed here as no closely affiliated series seems apparent. Common from Guinea to West Uganda and the Camerouns."

VIII. Crinum series Capense (Herb.) and allied pedicelled forms

--Subgenus Codonocrinum ex parte.

This clade consists of those South African Crinum whose foliage is broad and grows vigorously taking several years to mature, while the leaf apices exhibit acute needle-like tips. The blossoms are pedicelled and normally flower sequentially. Climatic changes over the past ages have caused far more morphological speciation in these latitudes than in the tropics.

Crinum bulbispermum (Burmann) Milne-Redhead & Schweickerdt in Jour.

Linn. Soc., Bot. LII: 161, 1939. (2n = 66)

Amaryllis bulbisperma Burm. prodr. 9; C. asiaticum L.

Mantissa pt. Alt. 362, 1771. Bothalia 11 (1& 2): 49. 1873.

Crinum capense Herb. Treatise: 23,1821; Amaryl.. 269, 1837., based on

A. capensis Miller Gard. Dict. ed. 8: t.12, 1768.

var. album as A. capense alba Bot. Mag. 8: t-661. (2n = 22)

var. striatum as C. longifolium striatum.

var. farinianum Baker. Gard. Chron. i: 833, 1887.

var. riparium Burch. Bot. Mag. 53: T-2688.

(Amaryllis longifolia L. Spec. Pl. ii: 421= Cybistetes longifolia Milne-Redhead

& Schw. Jour. Linn. Soc. LII: 191, 1939).

var. longifolium Thumb prodr. 39, Bot. Mag. t.661; Red. Lil. t, 347.

C. macowanii Baker in Gard. Chron. 9: 298, 1878, w/ Macowan 2122

lectotype; Fl. Cap. 6: 202, 1896; Bothalia 11 (1& 3): 46, 1973;

Bot. Mag. t.6381; Herbertia 52: 60,1997. (2n = 22, 44).

Syn. C. gouwsii Traub in Herbertia, Plt. Life 10: 40, 1954.

Sub. sp. confusum Verdoorn in J. S.A. Bot. 32: 67, 1966.

Sub. sp. kalahariense Hann. in La. Soc. Hort. Research III (5): 256, 1970.

Sub Sp. tetraploideum Hann. in La. Soc. Hort. Research III (5): 256, 1970. (2n = 44)

Amaryllis revoluta sensu Ker-Gawler in Bot. Mag. t-1178.

Syn. C. johnstonii Bak. Bot. Mag. t-7812, 1902.

Syn. C. pedicellatum Pax. 1892, p. 142; Baker 1898, p. 399.

Syn. C. Kirkii auct. non Bakr. Lugard. 1933, p. 99.

C. graminicola Verdoorn in Fl. Pts. Africa XXIX: t-1155, 1953;

Bothalia 11, (1 & 2): 43, 1973. Transvaal, So. Africa.

C. forbesii Lindley Trans. Hort. Soc. 6: 88. 1825; Schult..

Fl. Syst. vii: 864.= C. forbesianum Herb. Amaryl.. 267, 1837. Bot. Mag. t.6545;

Herbertia 52: 57, 1997. (Plant lost, supposedly from Delago Bay, Moçambique.)

Syn. C. forbesii (Lindl.) Schultes emend Lehmiller Int. Bulb Soc. Herbertia 48: 86, 1992.

Syn. C. delagoense Verdoorn in Fl. Pts. So. Afr. 35: t-1389, 1962;

Bothalia 11, (1 & 2): 44, 1973. Transvaal, Natal.

C. stuhlmanii Baker in Dyer Fl. Trop. Afr. VII: 578, 1898; Nordal

in Norw. Jour. Botany 24(3): 187, 1977. (may be syn. of C. forbesii fide Nordal).

C. crassicaule Baker Handbk. Amaryl. 85, 1888;. S. E. Trop. Africa.

var. C. foetidum Verdoorn in Bothalia 10: 56, 1969; 11(1&2): 42, 1973;

Herbertia 48: 80, 1992.; 52: 54, 1997.

var. C. papillosum Nordal in Nor Jo. Bot. 24(3): 185, 1977;

Herbertia 48: 90, 1992. Tanzania.

Note: In several instances Cybistetes longifolia has been confused with C. capense Herb. or cited as C. longifolia L.

IX. Crinum pseudoseries Tenuifolia

Subgenus Codonocrinum ex parte.

The following species represent a diversified group adapted to a desert environment such as found between the tenth and twenty fifth parallels in South Central Africa. Rainfall is limited and most species have adapted to growing in occasionally flooded pans, as noted. The foliage is narrow, under 2 to 4 cm. by 30 to 70 cm. long and often semierect. The blossoms are normally sessile or subsessile and range from one to 10 per umbel, rarely more. Flowering is semi-sequential.

C. campanulatum Herb. in Bot. Mag. t-2121, 1820; Amaryl.. 270, 1837.

Syn. C. caffrum Herb. in Sweet Hort. Brit. III: 678, 1839.

C. aquaticum Burch. Ms; Bot. Mag. t-2352, 1822.

C. lugardiae N. E. Br. in Gard. Chron. 2: 49, 25th July 1903; Herbertia 52: 50, 1997.

syn. C. carolo-Schmidtii sensu Verdoorn fide Roessler. Herbertia 48: 74, 1992.

Syn. C. podophyllum Bak. in Bull. Herb. Boiss. 2nd ser. 3: 667, 1903; Letty Wild Flowers of the Transvaaal 1962: p. 56 w/t.5:1.

Syn C. crispum in Phillips in Flr. Pt. Afr. 14: t-532, 1934.

C. lineare L. f. Suppl. 195, 1781; Baker Amaryl. 92, 1888; Bot. Mag. t.623, t.915.

Syn. C. revolutum Herb. in Bot. Mag. t-2121, 1820; Amaryl. 267, 1837.

Syn. C. herbertii G. Don. Gard. & Bot. 1831-37. (conflicts w/ C. x herbertii)

C. variabile (Jacq.) Herb. Amaryl. 268, t-44, f. 23, 1837. Baker. in Fl. Cap. 6: 199, 1896.

var. roseum Verdoorn Flr. Pt. Afr. 36: t-1433, 1964.

Syn A. variablis Jacq. Hort. Schon. 4: 14 t-429. 1804.

Syn. C. crassifolium Herb. Append. 23, 1821, nomen nudum.

C. crassipes Baker G. C. ii: 126, 1887; Hand bk. Amar. 90, 1888. Central Africa.

C. Carolo-schmidtii Dinter, Neue Pfl. S. W. Afr. 26: 1914; Herbertia 50:61, 1997.

Syn. C. occiduale Dyer in Herbertia 15: 31 1948.

var. C. carolo-schmidtii sensu Raesler (minor form).

syn. C. stenophyllum Baker Nomen Illegit.

(C. carolo-schmidtii sensu Verdoorn = C. lugardiae var. Int Bulb

Soc. Herbertia 48: 80, 1992.)

C. paludosum Verdoorn in Flr. Pt. Afr. 39: t-1523. 1968; Bothalia 11(1&2):35,

Fig. 11& 12, (Similar C. carolo-schmidtii Dinter).

C. trifidum Nordal Norweg. J. Bot. 26: 279-281, 1979; Herbertia 50: 60, 1997.

Crinum campanulatum, C. paludosum, C. rautanenianum, C. trifidum and C. carolo-schmidtii are aquatic in habitat. For current information on these South West and subtropical African Crinum see D. J. Lehmiller's study in Int. Bulb Soc. Herbertia 48(1&2): 32, 70, 75, 86, 1992, with figs. also 50: 37, 44 and 119, 1997. Several of the above may be associated with the series Ornatae. His nomenclatural corrections are employed herein. (See the Lehmiller Abstract in the back of this publication). Unfortunately Verdoorn and Nordal confused some identities.

X. Unplaced African Codonocrinum

The proper series for these species has not been fully confirmed.

C. vassei D. Boise in Bull. Mus. Hist. Nat. Paris, xiii: 444, 1907.

Rev. Hort. 1908: 132, fig. Moçambique.

(16 blossoms opening sequentially, tubes slightly curved, allied to C. graminicola (?)).

C. brachynema Herb. in Bot. Reg. 1842: Misc. 28; Hooker f. in

Bot. Mag. t-5937; Flore des Serras t-2303. Bombay.*

(15 to 20 subsessile blossoms, short tepaltubes and tepals,

No known alliance, probably a remnant series).

C. paxtonii Herb. Amaryl.. 271, t-44, f.24. India (Hybrid?.)

(Herbert's conclusion: A possible Scabrum x bulbispermum

type of hybrid from Calcutta gardens).

C. samuelii Worsley G. C. XXVII: 42 1900; XXXII: 304, 1902 Central Africa.

(Umbel 2-flowered, sessile, near C. pauciflorum(?) Series Ornatae (?)).

C. Wimbushii Worsley in G. C. II: 303, 1902. So. Africa (allied to C. Samuelii).

(Scape 4--5 Cm. high, umbel 2-6 campanulate blosssoms, subsessile.

Series Ornatae (?) Nr. Lake Nyassa).

XI. Crina antiqua pro parte

African Subgenus Stenaster

and Subgenus Crinum (Ex Platyaster) Forms

Some members of this group may be remnant forms of primitive Crinum. At present, descriptive information is often too scant to validly recognize whether some are Stenaster or Platyaster; pedicelled or sessile forms. In fact. J. G. Baker found it impossible to separate the subgenus Stenaster from Platyaster and he ended up calling most specimens which he examined Stenaster. We obviously have scattered remnants of most distinctive clades. I note that C. buphanoides and the next four Crinum bear some 30 to 40 pedicelled blossoms typical of the Stenaster series Asiatica while C. purpurascens shows some resemblance to the Asian C. defixum. The remainder are mostly Platyaster. The tepals of C. natans are broad enough to be a Platyaster as are C. subcernuum and possibly C. purpurascens. The entire group needs careful restudy noting sessile and pedicelled forms. It is my opinion that the very short tepaltubes (5 cm.) on C. modestum suggests an Ammocharis is involved here, or else a most primative Crinum. Since East Africa is the source of Crinum, some Asian, Indian Ocean Island, and Madagascar forms obviously overlap with east coast African species. ..... Presumably, the many past glacial-period induced dry-cycles in continental Africa have hindered wide establishment of these two subgenera due to their actinomorphic blossoms.----However, these same dry periods have seemingly accentuated speciation for the trumpeted Codonocrinum and allied Amaryllidea in specific areas about Africa. With the ease of water migration were the most primative Crinum originally Asian? ---Or due to subgenus Platyaster's limited distribution about South East Africa is it of Asian evolution?

Stenaster---Multiflora

Crinum buphanoides Welwitsch ex Baker in Jour. Bot. Xvi: 195, 1878;

Herbertia: 52: 47. 1977. Angola.

C. leucophyllum Hook. in Bot. Mag. t.6783. Tropical Africa, Damaraland.

C. amboense Baker in Bull. Herb. Boiss. Ser. II, 3: 666, 1903. S. W. Africa.

C. belkianum Schinz in Bull. Herb. Boiss. iv, App. III: 47. Trop. Africa.

C. poggeii Pax in Engler Bot. Jahr. Bk. 15: 141, 1893. Quango, Tropical Africa.

C. longitubum Pax in Engler Bot. Jahrbk 15: 141-142, 1893, Angola.

C. mauritianum Lodd. Bot. Cab. t-650; Fl. Pt. Afr. t.1874, Sten.,

(as C. firmifolium in error). Herbertia 48(1& 2): 72,1992, Madagascar.

var. C. voryonii Jumelle, in Rev. Hort. n. s. XIX: 205, 1924..

var. C. braunii Harms in Notizblatt Kgl. Bot. Garten and Herb., Berlin 1: 19-21, 1895.

Madagascar. (Sten.)

Crinum hildebrandtii Vatke. in Monatsb. Aead. Berlin. 863, 1876; Emend.

Int. Bulb Soc. Herbertia 48(1&2): 70, 1992; Bot. Mag. t.6709? 4 -8 flowers,

Comoro Is. & Moçambique.

Stenaster-Platyaster---Pauciflora

C. subcernuum Baker in Gard. Chron. ii: 180, 1881; Herbertia 52: 47, 1997.

10-15 flowers. Zambesi. (platyaster?).

C. crassicaule Verdoorn non Baker in Bothalia 11(1,2): 32, t.2, 1973. = C. subcernuum Baker(?). Herbertia 48: 80, 1992.

C. manyharthii Baker Bothalia 11(1,2): 32, t.2, 1973= C. subcernuum Baker(?)

Herbertia 48(1&2): 80, 1992.

C. ondongense in Bull. Herb. Boiss Ser. II, 3: 666, 1903. S. W. Africa. Platy. (Ammocharis?)

C. welwitschii Baker in Gard. Chron. II: 40, 1884. Angola. Sten.

C. nordaliae Mabberley Taxon 33: 430, 1984.

= C. biflorum Baker. Warbe Kunene-Zambesi Expedition 565, 1903. Sten.

C. firmifolium Baker. Jour. Linn. Soc. xx: 270, 1883. 6-8 flowers. Madagascar. Sten.

C. natans Baker in Dyer Flora Trop. Africa. Vii: 396, 1898 (4-7 flowers sensu Chevalier.)

var. calamistratum Bogner/Heine, Berlin Bot. Gdn. var. nov.*

var. africanum, Berlin Bot. Gdns.

C. purpurascens Herbert in Amaryl. 250. 1837; Adansonia ii: 20 (2): 180. Isle of Camerouns.

Syn. C. koutiense Chev. Etudes Fl. Afr. 306, 1913. 5-8 flowers. Platy.

C. modestum Baker. Jour. Linn. Soc. XXII: 528, 1887. 4-6 blossoms , Madagascar.

(May be an Ammocharis due to the short tube, otherwise Platy.)

C. biflorum Rottboll Plant. Hort. Univ. Prog. 1773, 12.

Described as Platyaster type with distichous foliage. Tepaltube 2 inch long and 1 1/2 patent tepals. Such short tepaltube forms not known. Not C. distichum Herb., as listed in Taxon 33(3): 439, 1984, sensu Mabberley, and does not compare to Brit. Museum #1 specimen of C. distichum Herb. Possibly C. modestum Baker (above). (May be an Ammocharis due to the short tube.)

C. nerinoides Baker Bull. Herb. Boiss. Ser. II, 3: 555, 1903.

Identified as an Ammocharis from living plants. S. W. Africa. i.e. Ammocharis nerinoides (Baker) Lehmiller, Novon 2.1:33, 1992.

C. baumii Harms. in Warb. Kuene-Zambesii Exped. p. 199 (1903), sensu M. -R. & S.

Int. Bulb Soc. Herbertia 48(1&2): 77, 1992; Herbertia 52: 46, 1997. (Non Ammocharis).

C. harmsii Baker in Warb.-Kuene-Zambesii Exp. P. 565, 1903. (3 flrs. Platy)

J. D. Lehmiller states C. voryonii Jumell and C. braunii Harms are synonyms for C. mauritiamum Lodd. I. B. S. Herbertia 48(1&2): 82, 1992. He cites C. mauritianum is a bog plant while C. firmifolium is a desert form, and indicates that Perrier's study of Madagascar Crinum is confused regarding identities.

Note: Figure 2 of C. subcernuum Baker, erroneously labeled C. crassicaule Baker in Bothalia 11(1,2): 32, t.2, 1973, indicates drip-tips on the tepal apices similar to those on C. americanum. The two species may have a common origin. Baker considers C. subcernuum a Platyaster.

*The C. natans variants have foliage 1 to 2 cm wide and more rigid than the type species.

All are completely aquatic in habitat. Specimens are growing at the Berlin Museum.

XII. Crinum series Americana (L.)-

-Basically ex Subgenus Platyaster --

Now C. subgenus Crinum per 1988 code

The Crinum americana series is a closely interrelated semiaquatic group of Crinum which evolved from an obvious Asian entry which has been scattered by ocean currents through much of tropical and subtropical coastal or riverine North and South America. Members of the series are mostly found along overflow land, in coastal swamps and about the Caribbean Islands. The series can be recognized by the distinctive blunt-tipped apices to the recurving foliage, the semi-Stenaster to Platyaster sessile blossoms and the unique long subsurface rhizomes which encourage rapid vegetative reproduction. Those forms which have adapted to a full aquatic habitat have little need for an enlarged bulb while those native to a well drained or sandy soil produce distinctive globular bulbs often several diameters larger than their short leafy necks. Based upon C. cruentum, a sessile Stenaster form found at Oaxaxa, Mexico, this series nearest affinity appears to be associated with the C. series Defixa, a sessile Stenaster whose tepals have semi-drip tips. Due to the series Americana's questionable parental history, whether Stenaster or Platyaster, the conversion to the subgenus Crinum to comply sith the Botanical Code should definitely exclude conversion of the Asian and African Platyaster forms as has occurred.

Floods, tidal action, and ocean currents have all accounted for the abnormally wide distribution of this series whose entry apparently predates the uplift of the Panama Isthmus and the Amazon Basin. Varied ecological conditions have encouraged the formation of numerous ecospecies and innumerable variants.----Still, the resulting speciation has not been too extensive since occasional drifting bulbs or seed have led to some intercrossing and have kept many colonies from diverging too widely. Most colonies about the Caribbean and Gulf area suggest some gene pool mixing has occurred between the C. erubescens and C. americanum members---

Basically, many show evidence of introgressive hybridization. Those forms most likely to be near homozygous are essentially those plants well inland or geographically isolated,-- for example, along the west coast of Mexico. Other examples can be found in central Peru, the outer Antilles, or in the vicinity of Gatun Lake in Panama where distinctive forms exist.

As noted in the introduction Linnaeus describes his C. americanum supra and beta in 1753 as: "Crinum collarum apicibus introrfum ungviculatis" which, on translation and close checking of numerous American forms discloses that the last 3 to 7 mm. of the tepal apices normally roll up into small incurving longitudinal tubes or cylindrical claws simulating slim curved hooks which deflect inward toward the tepaltube. In other cases they appear as slightly callused in-turning drip-tip apices. This is an elusive but accurately described morphological feature typical of American Crinum, but ignored in many illustrations or post Linnaeus descriptions. Claws are evident in Commelin's Rariorum 15, t.15 illustration, but not in 14, t.14, 1706, or the Dillen Elthamensis 195, t.161, 1732, plates, or in Moninckx's Atlas 6: t.29 and t.30 plates as cited by Wijnands, 1984. These pre 1753 non complying plates are all of a Stenaster Crinum showing linear tepals, and according to Dillen have extended sequential flowering. As such they are a lesser form of C. Asiaticum, quite possibly C. macrantherum Engl.,---definitely not C. americanum. Apical claws are not present in Asian forms. (D. O. Wijnand's (1984) has erred in identifying Moninckx's Rar. & Exotic 14, t.14 as the type for C. americanum L. supra). In turn, Commelin's 15, t.15 illustration is that of a Platyaster Crinum with abruptly spreading tepals wherein all blossoms flower within a short period of 3 to 5 days, and typical of the American forms. From the tepal and tepaltube dimensions this t.15 plant is probably a Surinam small white flowered form. On the other hand the Hortus Siccus Cliffortianus 127.1 mounted specimen in the British Museum has features typical of C. americanum common about the marshes of St. Augustine, Florida. It is not a tropical form as the latter have shorter tepals. In conclusion, we are left with Linnaeus describ­ing his supra and beta forms as having hooked claws, but no clear understanding what supra was, other than a Floridan or Gulf of Louisiana form as suggested by Baker in 1881. In turn, beta has been identified as C. commelynii in preference to C. erubescens. Botanists following Linnaeus did not clear the issue until Ker-Gawler took over. He suggested erroneously in B. M. t.1034 that beta be considered the supra, dismissed making comment concerning Linnaeus's supra, and recognized the above plates t.14 and t.161 as C. asiaticum sensu Ker-Gawler. Baker, 1888, disregarded Ker-Gawler's suggestion concerning t.1034 and cited HSC 127.1 as the rightful type-specimen for Supra, but obviously ignored the character essentialis description concerning the tepal claws.

In the past, Continental botanists have had far more access to Spanish America than the English and consequently they have obtained innumerable Caribbean and Brazilian Crinum forms. Schultes, 1830, Kunth in 1843 and Roemer in 1847 all meticulously describe and identify some 20 to 25 species with numerous variants, whereas J. G. Baker lists only some 8 or 9 forms in his 1888 Handbook of the Amaryllidaceae. Unquestionably Baker passed up a number of these named plants as he lacked herbarium specimens for confirmation. Many specimens have been collected since 1882 throughout Spanish America where a number bear resemblance to the forms Kunth or Roemer describe. Some 7 or 8 distinctive forms have been named since 1890. Accordingly, there now appears to be some 33 species recorded, many with special variants as well as synonyms as cited below. Unquestionably, these 33 species are far more than justified for a closely interrelated series. A complete reevaluation is desirable but rather difficult to accomplish as many of the original holotypes are missing and their immediate point of collection has never been recorded. In addition, Roemer's excellent key needs translation from the Latin and updating, if possible. Too many present day botanists have failed to make use of his exacting study. All in all, a reassessment using modern morphological, genetic, cyme (or multicyme examinations of the umbels), and electroanalytical techniques are needed. Such studies should give us a better overall understanding of the series.

One simple procedure to aid identification is to place the plants as to geographical location, and then compare them by the relative length of the tepals. Such a procedure eliminates searching a number of non applicable descriptions while attempting specific identifications. Understandably, though, collected material may not key out exactly to any described lectotype form, but only within the scope of the general population and latitude.

As for a lectotype for C. erubescens, Linnaeus the younger's reference to Miller's Icon t.110 is unreliable. J.G. Baker cites a 1784 specimen in the British Museum from Banks home at Spring Grove as a holotype. It is reported that a mounted specimen from Panama existed in the Philip Miller collection, which needs checking. Redouté's plate 27 in Les Liliacees, 1803, is quite a distinctive specimen. But where was it obtained and does the specimen still exist? It is far more colorful than the Ker-Gawler plate 1232 in the Botanical Magazine of 1809 which lacks the deep pigmentation present in the foliage, whence the name was derived, not from the blossoms. Baker in 1888 concluded that Commelin's Rariorum 15.t.15 which Linnaeus cited as his unpigmented Beta form of the C. americanum, differed sufficiently to be accepted as C. commelynii Jacquin in lieu of C. erubescens Aiton or Ker-Gawler's West Indies form of C. americanum.

Finally we have this huge Gatun Lake, Panama form which has been listed as a C. erubescens but whose unpigmented foliage is very similar to that of the C. americanum from the Neches River in Louisiana, even to the random dentations along the leaf margins of each, a feature that Ker-Gawler viewed as specific to C. erubescens alone, but actually varies from clone to clone and colony to colony for most C. americanum forms.

Other examples could be cited. But as stated , a restudy of the entire C. americana series is quite desirable in order to clear the clutter of excess species. In a broad sense the entire series may be a single species. In a 1984 publication, D. O. Wijnand, Botany of the Commelins gives several questionable interpretations concerning identities of C americanum L. and C commelynii Jacq. It is unfortunate that he depended upon confused literature of the past and not a collection of living material. Crinum with pedicelled blossoms are definitely not American.

XII-A. Crinum series Americana (L.).

Tepals as long as tepaltube:

Crinum caribaeum Baker in Gard. Chron. ii: 40, 1881; Hnd. Bk. Amar. 80, 1888; La. Soc.
Hort. Research, III(5): 246, (1970-71). Jamaica. (Uniquely single flowered)

XII-B. Tepals about 80% as long as tepaltube:

C. Americanum L. sensu Hortus Siccus Cliffortianus 127.1, Holotype.

This Clifford specimen in the British Museum was apparently mounted by Linnaeus and was used by J.G. Baker in describing C. americanum L. in his Amaryllidaceae p.80, 1888. It is typical of the subtropical longer tepaled American Crinum as found about St. Augustine, Florida, in Tingo Maria, Peru, and the Parana River Basin, Argentina. Tropical forms which have shorter tepals like Commelin's Rariores 15, t.15 by custom have been considered C. erubescens or C. commelynii Jacq. However, most distinctions are merely environmental adaptations of subspecies status, or less. The HSC specimen is obviously the plant which Linnaeus considered his supra form in Hortus Cliffortianus. Linnaeus' assignment and objective in the 1737 Hortus Cliffortianus was to name and catalogue all living and dried specimens in Clifford's garden and herbarium at Hartecamp, Holland. Thus with the HSC 127.1 specimen being monotypic, it was merely identified as 'genus Crinum'. His character essentalis: "Crinum collarum apicibus introrfum ungviculatis descriptive name, and the trivial name "americana" were added in his 1753 Species plantarum I. (See Philosophia Botanica 1751: Aphorisms 256, 257, and 293).

Var. robustum Hannibal in La. Soc. Hort. Research. Jr. III (5): 308,1970-71;

L. H. Bailey Cyc. of Hort. I: 893 No. 5, 1812 = C. erubescens var.?

C. herbertianum Schultes. f. Syst. VII: 871. (Non Wallich in Roem. Syn. ensat 80, 1847) =

Syn. C. strictum Herbert non Hornemann in Bot. Mag. t-2635.

Mexico? (or C. Amoenum affinis, India?). Recently reported as

found along the west coast of Mexico.

Syn. C. americanum sensu HMS Sulphur Exp. west coast Mexico, 1845.

XII-C. Tepals 50 to 60% as long as tepaltubes:

C. americanum (Beta) sensu Ker-Gawler non Linn. Bot. Mag. t. 1034: A West Indies variant

which Ker-Gawler introduced as a substitute for C. americanum L. supra when he found

that Commelin's rar. 14, t.14 reference and illustration was a Javanese C. asiaticum variant.

This Ker-Gawler t.1034 does not conform to the HSC 127.1 typespecimen.

C. americanum var. longii Hann. Subsp. nov. Long in Flora of Trop. Fla., 1971.

Differs by shorter tepals. Dade to Citrus Co., Fla.

C. erubescens L. f. ex Dryander in Ait. Hort. Kew. I (1): 413, 1789;

Ed. II (2): 221; Willd Spec. Pl. II: 463; Jacq. Hort. Schon. IV: 30, t-494, 1804;

Red. Lil. t-27, 1803; Ker-Gawl. Jour. Sci. III: 104, 1817. Holotype per J.G. Baker is a Brit. Museum specimen from Banks' Home at Spring Grove, 1784. Schultes 1830 and Herbert, 1837, cites some 12 references.

var. rubilimbum Herb. Amar. 251, 1837 is nearest Jacq. lectotype.

var. erubescens Redouté Liliacee t-27, 1803. (suggested lectotype, Pigmented foliage.)

var. C. erubescens Ker-Gawler Bot. Mag. t-1232. (Not lectotype).

var. Majus Herb. Amar. 251. 1837.

var. minus Herb. Amar. 251. 1837.

var. brasiliense Wild. Ms. Herb. Amar. 251.

var. caraccense Herb. Amary. 253. White flowered, Caraccus.

var. mexicanum Roem. Fam Nat. Reg. Veg. IV: 78. San Bertolo,

Oaxaca, Mexico. Form preserved at Waimea Bot. Gdn. Haleiwa, Hawaii.

Syn. C. cruentum Herb. Amary. 251.

C. erubescens Humb., Bon. & Kunth. non Aiton = C. Kunthianum.

C. viridifolium Roem. Syn. Ensatea 78, 1847 (=C. commelynii ?)=

Syn. C. erubescens var. viridifolium Herb. append. 21, 1821.

Syn. C. lindleyanum Herb. Amar. 352 & 401, 1837, (Non Schultes).

Near C. Americanum per Lindley.

C. odoratissimum Tsch. in Flora XIV, ii: 422, 1836. Differs from above C. erubescens.

C. inodorum Tsch in Flora XIX, ii: 422. 1836. Central America.

Syn. C. americanum Ht. Prag. (See Kunth. Enum. V: 559)

Syn. C. erubescens var. octoflorum Herb. Amary. 252, 1837.

C. gordonianum Roem. in Syn. Ensatea 80.

Syn. C. erubescens var. octoflorum Herb. Amary. 252. Venezuela.

C. Corantínum Herb. ex steud. Nom. Ed. II, i: 436.

Syn. C. erubescens var. Corantínum Herb. Amary. 251; Schultes f. Syst. VIII: 861. Guiana.

C. corantínum M. Roem. Syn. Ensatea 80 =C. erubescens.

C. aquaticum Herb. in appendix 22, 1821 = C. erubescens.

C. octoflorum Herb. ex Steud. Nom. Ed. II, i: 440. = C. erubescens.

C. salsum Rav. 1982. Plant Life Herbertia 38: 42, t.15, Brazil.

C. lindleyanum Schultes f. ex Seub. in Mart. Fl. Brazil III: t-159; Roem.

Fam. nat. Regni. IV Ensat: 78, 1847.

Syn. C. revolutum Lindley in Trans. Hort. Soc. VI: 285, 1826;

Syn. C. commelynii var. viridiflora Cayenne & Demara.

C. submersum Herb. Amaryl. 262, 1837; Amer. Hort. Mag. 47(3): 311,

1968. Probably C. scabrum x C. erubescens natural hyb. Brazil.

XII-D. Tepals 40 to 50% as long as tepaltubes

C. americanum Linn. forma beta in Hort. Cliff. 135, 1738, (1988 Bot. Code alternative Crinum subgenus Crinum type form); Spec. Pl. I: 292, 1753; Sp Pl. II: 419. In Hort. Cliff. Linnaeus states: "Apices petalorun unco seu appendiculo instructi feure, exacte ut in Comm. rar. tab. 15 exhibentur, quotiescunque eundem florentem vidimus." --This suggests that Linnaeus was comparing his supra specimen's tepal apices with those of his beta since tab.15 shows these claws. Linnaeus fil. cited this white flowered C. americanum beta as his C. erubescens in his 1782 manuscript. Dryander in Aiton's Hortus Kewensis of 1789 cites both this C. americanum beta from Spec. Pl. II: 419 and C. foliis carinatis, Mill. Ic. t.110, as C. erubescens. N. J. Jacquin in Hort. Schon. II: 40, t.202, 1802 reidentified Commelin rar. tab. 15 as C. commelynii, while Ker-Gawler in Bot. Mag. t.1034, 1807, ignored Jacquin's above action and reidentified C. americanum beta as a replacement for the Linnaean supra form which he considered nomen nudum.

Baker in his 1881 Garden. Chron. study made several corrections: He reverified C. commelynii Jacquin as a valid species. He noted that a Miller specimen of C. erubescens had flowered for Banks at Spring Grove in 1847 and was on file in the Banks' Herbarium and listed this as this typespecimen for Aiton's Hortus Kewensis I of 1789. Lastly, Baker recognized the HSC 127.1 specimen as representing Linnaeus's original Supra form of C. americanum since he used the specimen's floral dimensions in describing C. americanum Linn., while citing Ker-Gawler's West Indies beta form as a mere variant of the C. americanum L., not C. commelynii Jacquin. We find Baker's nomenclatural interpretations completely valid.

C. Commelynii Jacq. Hort. Schoenbr. II: 40, t-202 1797; Commelin Rar.

15,t.15, 1706, Holotype; Ker-Gawl Jour. Sci. III: 105; Schult. f.

Syst., vii: 875; Baker Amar. 1888, Surinam, Amazon Bsn., Cuba.

Syn. C. americanum beta L. Spec. Pl. 1753; Baker Amar. 1888, p. 86;

Wijnands, Bot. of Comm. 36, 1984.

Syn. C. herbertii G. Donn ex Loud. Hort. Brit. 120.

Syn. C. Commelinianum Herb. Amaryl. 255, 1837. See C. erubescens

var. Mexicanum above.

(C. commelynii (D. C.) in Redouté Les Liliacees's 6: t-322,

1811= C. americanum, Louisiana Gulf.)

C. cruentum Ker-Gawler Bot. Reg. t-171; Lodd. Bot. Cab. t-346: Habitat: Oaxaca,

Mexico, Br. Honduras, Costa Rica, Nicaragua.

Syn; C. crucibulbum Dietr. Syn. II: 1181. Trop. Amer.

var. albidum Herb. Kunth Enumer. 5: 555, 1850; Herbertia 4: 103, 1937.

C. americanum miniata Florey, 1982, Herbertia 38: 31-33, Fig. 27, 108, West Indies.

C. undulatum Hook. in Exot. Fl. t-200. Maranham, Brazil. Liverpool Bot. Gdns.

sensu Baker. Holotype.

C. attenuatum Wild. Ms ex Schultes f. Syst VII: 873. Para, Brazil. Syn. C. commelynii (?)

C. loddigesianum Herb. Amaryl. 253 & 401, 1837. Mexico.

Near C. cruentum and C. erubescens.

C. dieterichii Schult. Obs. 65; Syst. Vii: 876. = C. commelynii.

C. graciliflorum Kunth Enum. pl. V: 565. Venezuela. Holotype,

Berlin Bot. Gdn. Herb. 1844.

var. fendleri Baker Handbk. Amary.. 87. Venezuela.

C. attenuation Wild. = C. commelynii.

C. forgetii Wright in Kew Bull. 285, 1925. Peru.

C. nicaraguense Baker in Handbk. Amary. 86. Nicaragua.

Syn. C. Kunthianum var. Nicaraguense, Nova Grenada, Nicaragua.

C. oliganthum Urban in Rep. Spec. Nov. Reg. Veg. Fas. 15: 100, 1919.

Syn. C. erubescens Griesb. Cat. Cuba 250, 1866, non Soland.,

or Ker-Gawler. Havana, Cuba.

C. palustra Urban in Sertum Antill. IV: ; in Rep. Veg. Fasc. 15: 101, 1919.

Port-au-Prince, Haiti.

C. roozenianum O'Brien in Gard. Chron. 1: 701, 1891. Jamaica.

C. stapfianum Kranzlin in Kew Bull. 191: 1913. Gayoza Prov., Brazil.

C. revolutun Lindl. in Trans. Hort. Soc. VI: 285, 1826 =

Syn. C. commelynii var. viridifolium, Cayenne and Demara.

XII-E. Tepals one third or less the length of the tepaltube:

Crinum kunthianum Roem. Syn. Ensat 80. Nicaragua, Columbia,

Surinam. Holotype cited by J.G. Baker, Amar. 1888.

Syn. C. erubescens Humb., Bon and Kunth. non Aiton Nov. Gen. I:

279; Kunth. Syn. 1: 286. Turbaco & Los Volcancilas, New Granada.

C. kunthianum Hort. Daman ex Wein. Ill. Garten 358: f-76,

1890. No habitat given.

C. kunthianum var. nicaraguense = C. nicaraguense Baker in

Handbk. Amary. 86. Nova Grenada, Nicaragua.

C. dariensis R. E. Woodson in Ann. Missouri Bot. Gdns. XXV: 824,

1938. Panama Canal Zone. (midget form)

C. brevilobatum McCue in Ann Missouri Bot. Gdn. XLVII: 73, 1960.

Costa Rica.

XII-F. Length of Tepals or Tepaltubes not known:

Crinum sesse-mocinoii (Sesse & Mocino) Hann. La. Soc. Hort.

Research III (5): 320, 1970-71. Mexico. Habitat unknown.

Syn. C. uniflorum Sesse & Moc. non Muell in Flr. of Mexico 2:

88, 1894. (Name conflicts with C. uniflorum Muell Fragm.

iii: 23, Australia.)

C. argentium Pax in Engler Bot. Jahrbk. XI: 325-326, 1890. Sierra

in Tucuman, Argentina. Flowers only infrequently.

C. strictum Hornemann non Herbert in Enum. Pl. Hort. Hafa. 13,

1807; Hortus Regius Botanicus Haef. i: 318, 1813.

(C. strictum Herbert nomen illeg. = C. herbertianum Schultes,

which see.)

Note: Since the preparation of the above study, Dr. Lehmiller has published a detailed analysis on the identities and numerous references involving Linnaeus's Crinum americanum Supra and Beta in Herbertia 49 (1&2): 58, 1993. His conclusion is that erroneous identities exist, including the above Supra. But basically he holds that the C. americanum Beta's plate in Commelin's rariorum 15, t.15 is the valid holotype for C. americanum Linn.---Plus the genus Crinum, as well as the subgenus Crinum. Likewise that the genus is a single species. He proposes that the entire existing series of species be merely treated as subspecies.

It appears that Dr. Lehmiller was not aware of the numerous nomenclatural guide lines cited by Linnaeus in his Philosophia botanica of 1751.--- Specifically the aphorisms 256 and 257 (on page 202) and particularly 293. (See Stafleu's 1971 Linnaeus ect. p. 84-89). Under aphorism 293 neither phrase or trivial names are required for Crinum [americanum (supra)] in the Hortus Cliffortianus since Linnaeus was naming a new genus represented by a single species documented by Clifford's dried specimen, ---and augmented with three references to presumed polyphyletic forms as synonyms. But in Species Plantarum I and II it was necessary to add the 'Character essentalis's nomen specificum legitimum' phrase name concerning the tepal apices's drip-tip claws, since additional Crinum species were listed in these publications. This phrase name concerning the claws clearly validates Linnaeus's C. americanum supra's identity, which is re-confirmed by the claws on the Hortus Siccus Cliffortianus 127.1 specimen which Linnaeus probably prepared for Clifford as the authentic holotype. According to aphorism 257, Commelins rar. 14, t.14 and Dillen's t.161 plates should have been recognized as distinctive and thus excluded as invalid (falsa sunt or Lubrica erunt omnia) by Linnaeus in both Species Plantarum I and II due to the obvious absence of claws, as well as other noncomplying morphological and floral features (namely pedicelled blossoms and sequential flowering typical of the Asian source). As indicated by Linnaeus, the Commelin's rar. 15, t.15 form Beta is merely a minor variant of his above supra form.

XIII. Crinum Species,-----Subgenera Unknown:

In the past some botanists failed to take note of the floral shapes, and particularly the tepals. The best we can do is to segregate these plants according to geographical source.

Crinum cochinchinense M. Roem Syn. Ensatea 71. China. (Allied to

C. latifolium var. insigne?)

C. eleonorae Blatter & McCann in Jour. of Bombay Nat. Hist. Soc.

xxxii: 733, 1928. Bombay, India.

C. norfalkianum A. Cunn. in Hook. Lond. Jour. Botany i: 123. 1842. Norfalk Islands.

C. osbeckii Schultes f. Syst. vii: 1721. Asian.

C. octobris Nakai & Tuyama in Ic, Pl. Asia Or. iii: 208, 1939. In cultivation.

C. verschaffeltianum Firminger's Gardening Manual of India, 8th

Ed.: 332; Voight. Hort Suborb. Cal. 1845 = C. procerum.

C. filifolium Perrier in Bull. Soc. Bot. Fr. LXXXVI: 90, 1939. Madagascar.

C. walteri Overk in Mitt. Bott. Staatssaamml, Munchen, Heft 9-10: 440, S. Africa.

C. cordofanum Kotschy & Peyr. Pl. Tinn. 14: t-21. Trop. Africa.

C. Koutiense A. Chevalier in Etudes Africa Centre Franc. i: 306,

1913. Sudan (Chari). = C. purpurascens Herb. sensu Chevalier.

C. lanceii Herb. in Sweet. Brit. Hort. ed. III: 677. Guiana.

C. lane-poolei Hutchinson in Hutch. & Dalziel Flora West Trop.

Afr. ii: 307 in Clavi, 372, 1936, anglice. Sierra Leone, Africa.

C. knyffii C. Morris in Belg. Hortic. ii: 373, t-63, 1825. Hab. Unknown.

C. sectitumbum Tiejsm. & Binn. Cat. Hort. Bogoriensis, 377. Habitat

C. imperiale Regal Cat. Pt. Hort. Aksakov. 34. Habitat?

NON CRINUM

(misidentified species)

Crinum falcatum Jacq. by Fritsch Bull. Her. Boiss. ser. 2, 1:1107,

1901 = Ammocharis coranica Ker-Gawler Bot Reg. t.139, 1816.

C. coccineum (Pax) Fritsch Bull. Her. Boiss. ser. 2, 1: 1108, 1901

= Am. coranica K.-G. Bot. Reg. t.139. 1816.

C. tavelianum (Schinz) Fritsch Bull. Her. Boiss. ser. 2, 1: 1108, 1901

= Am. coranica K.-G. etc.

C. tinneanum Kotschy & Peyritsch Plantae Tinneanae 44, t.21, 1867.

= Am. tinneana (K.&P.)

Milne-Redhead & Schweickerdt Jour. Linn. Soc. LII:177, 1939.

C. ammocharides Baker Hand Bk. Amar. 79, 1888.

= Am tinneana M.-R. & S. Linn. Soc. LII: 177, 1939.

C. bainesii Baker Hand Bk. Amar. 79, 1888

= Am. tinneana M.-R. & S. Linn. Soc. 1939.

C. thrupii Baker in James: Unknown Horn of Africa p. 332, 1888.

= Am. tinneana M.-R. & S. Linn. Soc. 1939.

C. lastii Baker Hand Bk. Amar. 79, 1888.

= Am. tinneana M.-R. & S. Linn. Soc. 1939.

C. rhodanthum Baker in Dyer Fl. Trop. Afr. VII: 397, 1898.

= Am. tinneana M.-R. & S. Linn. Soc. LII: 178, 1939.

C. heterostylum Bullock in Bull misc. Info. Kew. 305, 1932.

= Am. heterostyla (Bullock) M.-R. & Sch. Jour. Linn. Soc. LII: 184, 1939.

C. parvum Baker see Herbertia II: 31, 1935, non Baker.

= Am. heterostyla (Bullock) M.-R. & Sch. Jr. Linn. Soc. etc. 1939.

A. angolense Baker Hand Bk. Amar. 80,. 1888

= Am. angolensis (Baker) M.-R. & Sch. Linn. Soc. LII: 186, 1938.

C. curvifolium Baker in Warb. Kunene-Sambi. Exped. 565, 1903.

= Am. angolensis (Baker) M.-R. & Sch. Linn. Soc. LII: 186, 1938.

C. longifolium (L.) Thumb. Prod. Pl. Cap. 59, 1794.

= Cybistetes longifolia. (L.) M.-R> & Sch. Linn. Soc. LII: 192, 1938.

C. concinnum Mart. & Roem. in Schultes Syst. Veg. VII: 857, 1830

= Griffinia concinna (M/ &. R.). (See G. liboniana Morren).

C. nerinoides Baker Bull. Herb. Bois. ser. 2, 3: 666, 1903.

= Am. nerinoides (Baker) D. J. Lehmiller, Novon 2, 1: 33, 1992.

Ammocharis falcata (Jacq.) Herb. Appendix p. 17, 1821

= Cybistetes longifolia (Linn.) M.-R. & S. Jour. Linn. Soc. 190, 1939.

Ammocharis herrei

= cybistetes longifolia sensu Snijman & Williamson, Bothalia 24: 127, 1994.

Brunsvigia albiflora Ecklon (syn. Crinum latifolium L. sensu Nordenstam)

= Crinum variabile (Jacq.) Herb. sensu D. A. Snijman.

Posssible Ammocharus forms

Crinum biflorum Rottboll Plant. Hort. Univ. Prog. 1773: 12, Mabberly in Taxon 33(3):

439, 1984. Non C. distichum Herb. sensu Hannibal.

Probably an Ammocharis from Mocambique.

Ammocharis baumii (Harms) M. -R. & S.

= C. baumii Harms.

A compendium of key texts relating to Amaryllis ornata Linn. filile

Since the Younger Linnaeus' 1782 unpublished manuscript and Wm. Herbert's 1821 Appendix and Preliminary Treatise are nearly unavailable to many botanists we present several key texts exactly as originally written: Linnaeus f. obviously received a King's assignment through Banks to describe a number of bulbs in Kew Garden. The additional bulbs noted in his manuscript were from non-Kew sources and were listed for his own records. The latter usually lack proper descriptions for publication purposes when compared to those prepared for the Hortus Kewensis.

First, Linnaeus the younger's MSS, 1782 extracts:

1. "A. zeylanica (Comm. hort. 1, p. 73, t.73) Patriam suaae Africam dict. Erhwt. [i.e. Ehret]."

   [Comm. Hort. 1: 73, t.73 illustrates the Lilio narcissus ceylanicus. Linnaeus f. fails to differentiate between Aisian and Tropical West African species].

2. And separtly: "A. ornata (Rumph. Amb. V: p. 306, t.105. Tulipa javana). Patriam in Africa: Guinea [now Ghana], Vulgo Cape-Coast-Lily. Ex horto in Wimblington [i.e. Wimbledon] Lord Rockingham. Obs. figuram coloratam hujus vidi in Musaeo Dni Banks in patriae cura Smeathmanii facta. Plantam vivam ex horto Dni Rockingham habui.

"Amaryllis: floribus sessilbus, corallis basi tubulosis; tubo spathis limboque longiori curvo; limbi lacinilis oblongis, aristatis; lacini infirma divaricata, concava. Linn. fil."

This manuscript's diagnostic description: "A.- floribus sessilbus, etc.", was reprinted in Aiton's 1789 Hortus Kewensis I, p.418, while the added phrase: "foliis lato-sublatis undulatis" appears in Hortus Kewensis II, 1811, and J. A. Schultes 1830 Systema. Aiton, Dryander and Banks dropped the confusing reference to Rumpius' Tulipa Javana synonym as an obvious geographical misfit. 'Linn. f.' was appended as author to the species he described. Concurrently, the following was added by Aiton (or Dryanded) to Hort. Kew. I, (but not II), while the above Rockingham notation was omitted.

3. "Cape Coast Lily"

Nat. Guinea [i.e. now Ghana]

Cult. 1740, by Robert James Lord Petre. Ehret pict. t.5. f.2 ?

Fl. June and July S. 4. 13. A"

Joseph Banks, Aiton, and Dryander all accepted the original Linnaean broad polyphyletic mode of classification. Thus they obviously concluded that the Cape Coast, Guinean sessile-flowered Amaryllidaceae were all interrelated with A. ornata and should be lumped together as a common species. This lumping of interrelated species was also sanctioned by Ker-Gawler in the 1806 Botanical Magazine t.923 and subsequent publications,--and continued by Dryander in the 1811 Hortus Kewensis II, as noted below. Note the ? mark which Linnaeus sr. used after Ehret's reference in Species Plantarum ed. I. The A. zeylanica relationship to Ehret's Pict. (A. scabrum form) was originally questioned by Linnaeus senior, and obviously as an A. ornata synonymic form by Aiton.

4. Linnaeus f, also listed several Pancratium and an A. fucata which he saw at Wimbledon. For A. fucata he wrote: "Ex horto Marquis of Rockingham in Wimbledon. Obs. longissimo tube et coarctite differt a A. longifolia facile." [A. fucata unidentifiable. A. longifolia probably a Cybistetes]. No source cited. These were not listed by Aiton or Dryander as known Kew specimens.

5. In 1785 Aiton filed a bifloral Crinum specimen with Banks which was identified as A ornata from the Capecoast, Guinea. The specimen which still exists is described below.

6. In 1787 L'Heriter prepared his Sertum Anglicum manuscript at Kew and and redescribed 16 of the 20 bulbs described by Linnaeus f., but missed mentioning A. ornata. Sertum Anglicum was published in 1788, and Hortus Kewensis in 1789.

7. In Willdenow's 1799 Species Plantarum iii, item [#19], A. ornata was recognized as Aiton's from Hort. Kew. I., while Rumphius' Tulipa javana was listed specifically [in #22] as a Linnaeus senior's A. zeylanica's synonym, thus rejecting Linnaeus fil's views..

8. In the 1811 Hortus Kewensis II Dryander listed all known Cape Coast Crinum as A. ornata Alpha, and the Asian C. zeylanicum forms as A. ornata Beta. Why both Dryander and Ker- Gawler ignored C. zeylanicum as the Linnaean supra form in lieu of the Beta group remains unexplained.

9. Next, in Herbert's 1821 Appendix p.8: "There is a doubt whether the younger Linnaeus' A. ornata was a C. broussonetii [Bot. Mag. t.2121] or a C. yuccæides. The specimen in the Banks' Herbarium from Kew bearing the date [1785] only three years after the name was given by the younger Linnaeus in this country, appears to be Broussonetii, and Yuccæides of the same herbarium to be my Yuccaeoides. Mr. Salisbury has reason for thinking he can trace my Yuccaeoides through the Marquis of Rockingham's collection to be Linnaeus's Ornata; but, the name Ornata has been applied to so many allied species in various works [by Ker-Gawler], I think it much best to retain the specific names of Broussonetii and Yuccæides, and give the [subdivision] name of Ornatae to the group which they belong."

10. And in the Appendix page 22, 1821, Herbert cites: "1. C. broussonetii Bot. Mag. 2121. Am. ornata, Banks' Herbarium." [Obviously implying that the 1785 Aiton herbarium specimen is the valid holotype for C. broussonetii (Redouté) Herbert in Bot. Mag. plate 2121.]

11. And in Bot. Mag. t.2466, (1824), in the text Herbert states: "We could not retain his [Wm. Carey's] Ornatum [for a C. jagus Mauritianus bulb] which applied to any Crinum would belong to C. broussonetii which is the A. ornata of Linnaeus fil.".

12. Finally in Baker's Amaryllidaceae, p. 89-90, 1888: Baker writes:

"C. yuccaeflorum salis. Parad. t.52.....syn. C. broussonetii Herb. in Bot. Mag. t.2121; Lodd. Bot. Cab. t.668; Bury Hexand. t.21, etc.

"Habitat: Sierra Leone. There is a specimen in the British Museum from Kew Gardens in 1785, so that it is no doubt the original Ornata of Aiton, who cites the younger Linnaeus under his diagnosis, and says it was introduced by Lord Petre in 1740. A plant gathered by Welwitsch in Angola (4015) is probably the same species." [Baker was quoting from the Hortus Kewensis I , not Linnaeus fil's 1782 manuscript.]

13. Linnaeus fil's manuscript was finally published by S. Savage in the American Amaryllis

Society's Herbertia 4: 91-100, 1937. One requires a copy of the bulbous section of the 1789 Hortus Kewensis I for reference purposes, as well as a copy of the International Bulb Society's Herbertia 51: 112-228, 1997.

Note: Brackets [ ] in above quotes were added for clarification of original quoted statements.

Comments

Historically, King George III, the 'farmer king', likewise the Prime Minister---the Marquis of Rockingham and his wife the Marchioness, as well as Lord Petre, Joseph Banks, J. E. Smith, William Aiton (the King's gardener at Kew), and numerous individuals about London were all interested in exotic plants, animals, improved livestock, and various scientific developments or ventures. It now appears that Banks, the King, and Aiton were all contemplating an Index Kewensis when the younger Linnaeus visited London in 1781-82. Consequently, with the index in mind, they gave him an assignment to describe and even name a number of bulbous plants in Kew Gardens. Thus, all the various species descriptions which he furnished Aiton employed a typical diagnostic phrase name arrangement, also a trivia name, all quite typical in form to descriptions normally used by his father in the latter's past publications. Linnaeus f. completed these descriptions with typical synonymic references, then where first grown in England, the native habitat, and flowering period. However, for his own personal use, Linnaeus f. expanded upon his initial listing for Aiton and included some 50 non-Kew bulbous plants which he observed in herbariums and private gardens about London, the English countryside, and Europe. Here, he discarded his more formal diagnostic descriptions in favor of brief descriptive remarks,-and where possible he cited a single reference synonym, normally referring to those provided by the original authors, not from his father's citations. In these instances this appears due to a family conflict where he was not to use his father's publications or references publically.

Then a number of phonetic spellings and geographical errors appear throughout his his manuscript. These errors suggest that he avoided discussing his portion of the manuscript with Daniel Solander or Jonas Dryanded. These latter were Linnaean graduates and botanical secretaries for Joseph Banks. Linnaeus f. lived with them while in London so he had ample opportunity for discussion.

Linnaeus f. cited some six amaryllidaceae species being grown at Wimbledon by the Machioness of Rockingham. Her subsequent correspondence with J. E. Smith at the Linnean Society indicates that she received several additional specimens from Tropical west Africa via Dr. Afzelius and others, and that she was actively engages with other exotics until 1802 when she passed away. Several appeared in J. E. Smith's Icones. In 1789 Smith rated her plant collection at Hillingdon House, Uxbridge, as "amongst the earliest and choicest in the country."

The Amaryllis ornata was included amongst the six Wimbledon species cited by Linnaeus f. and is quite remontly related to Crinum zeylanicumi. In this instance he gave the bulb a complete diagnostic description, apparently for Aiton's use. He brashly cited the Rumphius' Tulipa javana as synonym form while giving the Cape Coast origin,--an obvious clash in geographical sources. As Prine Minister the Marquis of Rockingham unquestionably had numerous contacts with various British Colonies and numerous Ports of Call where many exotic imports such as plants or seed could be obtained. So he and the Marchioness, as well as King George III, Aiton, and others knew of the Cape Coast's general location and its bulbous exports, ---unquestionably far better than Linnaeus f.

We can only presume that the A. ornata specimen, the bifloral sessile flowered Crinum which Aiton filed with Banks in 1785 was similar to the Marchioness' specimen. However, at some latter date while finalizing the long delayed 1789 publication of Hortus Kewensis, several of Linnaeus fil's questionable references and errors were encountered and reedited, such as Aiton's inclusion of l'Heriter's 1787 etchings of A reticulata and A vittata (t.14 and t.15) from the 1788 Sertum Anglicum as better references. The most obvious correction was the deletion of Tulipa javana, since the Cape Coast origin for the Marchioness' A. ornata was well known. Citing Java as a source in the Hortus Kewensis would be an open blunder, quite unacceptable, even if bearing Linnaeus fil's name,--since King George III, or the relatively knowledgeable and somewhat dominant Marchioness, as well as others would critically examine the Hortus Kewensis as soon as copies were available. Too many botanically inclined individuals already knew the sources for the Marchioness' numerous bulbs, as well as Linnaeus senior's having identified the Tulipa Javana as a synonym of the A. zeylanica in his 1762 Species Plantarum II. Subsequently, Aiton and Dryander obviously identified and lumped Lord Petre's and Ehret's specimens as logical synonymic Gulf Coast A. ornata forms. One gross error escaped notice: C. asiaticum was recognized as a pedicel flowered C. americanum. Despite his serveral corrections, Aiton still listed 'Linn. f.' as author of the various descriptions which he submitted, presumably for the Linnaean prestige in Hortus Kewensis I, but not II.

We append herewith the actual text and dimensions taken from a Xerox of Wm. Aiton's 1875 herbarium specimen which Herbert identified as Crinum broussonetii, (Syn. A. ornata Linn. f.)

"Amaryllis ornata Linn. fil. Vulg: Capecoast Lily. Aiton Hort. Kew. 1: 418 (1789)."

"Crinum broussonetii (Red.) Herb." amend.

Bulb; no information.

Foliage: 27.5 cm. long x 2.5 cm. tapering, margins scabrous. (probably exterior leaf).

Scape: 5 mm. dia. x 10 cm. long, as cut. , spathe valve bified 5 cm. long.

Umbel: Two blossoms. Ovaries; sessile.

Tepaltube: 11.8 cm. long, curved x 2.5 mm. wide.

Tepals: 9.37 cm. x 2.2 cm. wide.

Comment: Foliage width nearer that of C. yuccaeflorum, otherwise it is difficult to determine if C. yuccaeflorum or C. broussonetii is the nearer affiliate. We note in later publications that Herbert favored C. broussonetii.

From the above specimen's dimensional information we believe that this Kew specimen and C. broussonetii (Redouté) Herb. are cospecific bifloral forms, but question whether the multiflowered C. sanderianum Baker (syn. C. ornatum (Aiton) Bury), or C. scabrum Herb. are genetically or cytogenetically intraspecific variants of this biflowered A. ornata Aiton, though there is a close intraseries species relationship.

Rumphius' tulipa javana has been recognized under several synonyms, originally by Linnaeus in his 1762 Species Plantarum II as an A. zeylanica (Syn. C. zeylanicum), which is still considered valid, ---then by Roxbourgh as C. moluccanum, and by Joseph Dalton Hooker as a C. latifolium variant. The Roxburgh specimen is now filed under Hooker's C. latifolium in the British Museum.

In conclusion, from the entries made in the 1789 Hortus Kewensis I, Joseph Banks, William Aiton, and Jonas Dryander all presumably recognized the Rockingham A. ornata to be a CapeCoast bulb from Guinea (now Ghana), and allied specieswise to both Lord Petre's and Ehret's pict. t.5, f.2 specimens , as cited. In no way could they accept Linnaeus fil's inconsistant personal views. They were well aware of his acute conflict with his parents. In 1799 Willdenow, in the Species Plantarum III: 55, #19, recognized Amaryllis ornata Aiton as a distinct species, quite unassociated with A. zeylanica Linn. and its polyphyletic associates. In the subsequent 1811 Hortus Kewensis II all Cape Coast forms were clearly recognized as a single species of A. ornata Alpha, while all Orientals were the Beta. Tulipa javana Rumphius is distinctly a Beta form which fails to comply geographically with the recognized source of the Rockingham A. ornata. In turn, in 1820, Herbert recognized Aiton's 1785 A. ornata specimen as a C. broussonetii (Redouté) Herbert. Then Mrs. Bury cited her illustration t.18 in Hexandria Plantae of 1831-34 as Crinum ornatum, namely C. ornatum (Aiton) Bury, which has been accepted by J. K. Morton on 1961 and F. N. Hepper in 1968 as their identification for several phyletic forms from Tropical West Africa, apparently including C. scabrum.

L. S. H.

Genus Crinum Linn. in Hortus Siccus Cliffortianus 127.1.

I. q. Crinum americanum L. lectotype specimen

(Note the drip-tips on the tepals apices as cited by Linnaeus in Species Plantarum I: 292, 1753. Few, if any past botanists ever noted this distinctive character essentalis identification feature.

Umbel of C. asiaticum shown in cross section. Two similar spirals, each with 22 blossom buds in a 3.5 revolutions are shown. the paired order of flowering is shown spiraling toward the centers. The bud brachts are shown in black in the first helix.

Cross section of the umbellar face of Crinum moorei. Ten flower buds are located between the two sheaths of the bified spathe valve in a two revolution helix. The flowering order is as numbered. The brachts are shown in black.

By permission of D. Muller-Doblies

The genetic distinctions between the pedicelled-floral and sessile floral Crinum types

The genus Crinum is normally subdivided into three subgenera: Stenaster, Platyaster (or Crinum in part, now) and Codonocrinum, all as based upon obvious floral morphological features. In turn, these subgenera can be split into two distinct umbellar morphological types: Those Crinum series with pedicelled blossoms and those with sessile or subsessile blossoms. These distinct features have never been given much attention, but morphologically and genetically they are of distinct evolutionary significance: First, as a specific feature common to each respective series involved, and secondly, as the leading cause of sterility, substerility or fertility where interspecific or interseries breeding is involved.

As indicated in the Muller-Doblies 1977 report, his appended plates illustrating the umbels of C. asiaticum and C. moorei show these pedicelled umbels represented by spirally coiled cymes bearing single apical buds at their central apex. This structure is known as a determinate floral type and the buds flower sequentially, one-a-day. In contrast, the sessile and subsessile umbels have an indeterminate spiral structure lacking a single lead or apical bud, and the flowering occurs in sequential groups of 3 to 5 blossoms over an interval of several days. Crinum americanum and C. scabrum are typical examples of the sessile-indeterminate floral types.

In classifying known Crinum species according to their umbel type it soon becomes obvious that the majority of the C. subgenus Stenaster in Asia can be subdivided into the C. series Asiatica which are pedicelled, and the C. series Defixa which are sessile. In Africa, only a few pedicelled C. sg. Stenaster exist along with one only known sessile form. A few narrow petaled C. sg. Stenaster forms appear amongst the C. series Americana, suggesting an Asian C. series Defixa origin. Seemingly most C. series Americana have evolved into broader tepaled tropical Platyaster forms since their arrival.

In turn, practically all C. subgenus Platyaster, some 45 known species, are sessile flowered and native to Asia, Australia, and America. Only one known Platyaster species is found in Africa, C. subscernum on the Zambesi River, and it is pedicelled. So effectively, both the C. sg. Stenaster and Platyaster are Asian, or Asian adapted since the Ice Age..

Lastly, the C. subgenus Codonocrinum is basically African, with the sessile forms represented by both the C. series Jagi (ex C. gigantum) and the series Ornatae, with the latter including a C. subseries biflora. These two series are strictly tropical African, with the exception of C. zeylanicum, an Asian C. series Ornatae migrant, while a few of the Ornatae subseries biflore are found in south central Africa. In contrast, nearly all pedicelled C. sg. .Codonocrinum species are distributed from the southern Congo to the Cape. These are nearly all subtropical and the various species usually indicate a less common interrelationship than the above sessile series.

Basically, the umbelar structures, pedicelled or sessile, are associated with the specific habitats for most series, and on intercrossing, various series or species will indicate genetically which hybrids will have fertile, semifertile or non viable pollens. For example, C. scabrum x C. americanum, both sessile-flowered, but of differing subgenera, will cross and the hyrid pollen is often capable of backcrossing onto the parental forms. But C. moorei x C. americanum will be completely sterile since differing umbel forms and subgenera are both involved.

No known species are seemingly found where intermediate umbellar forms exist between the sessile, or subsessile, and the pedicelled structures. These features are distinct for each series. However, two examples exist where the series are in transition from one subgenera to another. The first is in Australia where the inland C. sg. Platyaster sessile form of C. flaccidum is adapting the broad petaled, semi-typical shape of the sg. Codonocrinum due to desert conditions. The second example is the transition within the series C. Americana from a linear-petaled sg. Stenaster form to the broader petaled Platyaster. Typically, the C. cruentum of Oaxaca, Mexico, in a well isolated western mountainous area uplifted by earthquakes still represents the primitive Asian migrant and bears considerable resemblance to C. defixum, its probable ancestor. However, most C. series Americana forms have assumed the broader tropical petals of the C. subgenus Platyaster. So basically, since C. americanum represents the typespecies for the genus Crinum, the Botanical Code now dictates that its subgenus also be known as Crinum too. This introduces a conflict in Asia and Africa since the C. subgenus Platyaster members there bears no obvious direct relationship to the C. series Americana. The subgenus platyaster identification should be retained for the distinctly unrelated Asian, Australian, and African group, i.e.--the C. subgenus Crinum should consist of the series Americana only.

Attention is directed to Dr. G. L. Stebbins, a well known evolutionary geneticist, who pointed out In his 1974 Flowering Plants that the characters of a variant differs materially from those which distinguish a species: Variant features are vegetative in character and are concerned with optimum local survival needs such as leaf size and plant stature, the growing cycle and dormancy. Species' features are distinguished with race reproduction by form of flowers and seed, adaptation to different pollinating vectors, seed size for adequate food reserve and a means of maximum dispersal to assure survival of some seedlings. We note that in the case of many Crinum forms it is difficult to determine the border line between variants and similar species, but the respective 'series' show marked divisions since in most cases each is a well established evolutionary branch having many distinctive or basic morphological features in common. And in the various levels of interspecific and intraspecific hybridization these distinctions lead to various levels of semi to complete incompatability. Since most of the series have a long evolutionary history, they may intercross, but their gametes are often disinclined to exchange many genes. This discourages mixed populations in the wild and keeps the series near pure-line.

Reference Publications

Aiton, W. 1789, Hortus Kewensis Ed. Vol. I: 413-421, with Dryander as editor.

Aiton, W. T. fil. Supt. Kew Gdns. 1799-1849, Hort. Kew. II, 1811-1812. with Dryander as Ed.

Andrews, F. W. 1956, Flowering Plants of Sudan, P. 287.

Bailey, L.H. 1912. Crinum in Cyclopedia of Horticulture, I: 890-896. (Text by Henry Nehrling).

Baker, J. G. 1888. Handbook of the Amaryllidaceae. 74-95.

-------------- 1881 Crinum Synopsis in Garden Chron. 15 & 16: pages 763, 786, 39, 72, 180, 398, 495, 588, 760, 785.

Bury, E. 1831-34. Hexandria Plantae, t.18,t.21 (1831-34).

Chevalier, A. 1950. Sur quelques Crinums de l'Afrique Trop. Rev. Int. Bot. Appliq Agricult. Trop. 30, 610-625.

Crawford, D. J. 1985, Electrophoretic Data and Plant Speciation, Taxon 10(4): 405-416.

Curtiss, W. 1787-1947, Botanical Magazine, various issues as noted.

Commelin, J. & C. 1697, Horti medici rariorum & exoticum, Amsterdam.

Dassanayake, M. D. 1981, 'Typification of Crinum zeylanicum (L.) L.' in Taxon 30: 418, 1981.

Dillenius, J. J. 1732, Hortus Elthamensis, London.

Dryander, J. Editor of Aiton's Hortus Kewensis Vol. I. 1789, under Aiton, Vol. II, Part 1, 2 only, 1811, under Aiton fil.

Ehret, G. D. 1748, Plantae et Pap. rar depictae t.5, f.2.

Fay, M. F. & Chase, M. W., 1996, Resurrection of Themidaceae for the Brodiaeaalliance, and Recircumscription of Alliacraea, Amaryllidaceae and Agapanthoideae. in Taxon 45(3): 441-452.

Fica, M. and Beiser, J. A., 1995: Tropicos Manual: Angiosperms: Monocots: Amaryllidaceae: Crinum: c/o A. V. Tucker. Missouri Bot. Gdn. 1995 [25 pages of text with need of updating some 20 obsolete names, L. S. H.]

Fosberg, F. R. & Sachet, M. H. 1987. 'Noteworthy Micronesia Plants' in Micronesia 20 (1 & 2): 131, 1987.

Geerinck, D.J.L. 1993. Amaryllidaceae, in Flora Malesiana Sec. 1, Vol. XI(2): 353-373, 1993.

Ferrarii, I. B., 1633. Florvm Cvltrvra Libri IV.

Hannibal, L. S. 1970-71. 'Crinum Bulletin and Index to Genus Crinum' in Louisiana Society for Horticultural Research Bulletin III (5): 221-322, 1970-71.

-------------1970. Australian Plants VI (45): p. 20.

-------------1985. 'The Asian--South Pacific Crinum' in Waimea Arboretum & Botanical Garden Notes, 12 (I): 2-8, 1985.

--------------. 1987. 'Those Wandering Plants: Crinum jagus' in Waimea Arboretum and B. G. Notes, 14 (I): 3, 1987.

--------------, 1990, True identity or Amaryllis ornata L. f. in Waimea Arboretum & Bot. Gdn. Notes 17(1): 2-12, June 1990; (2): Nov. 1990.

Hepper, F. N. 1968, Flora of West Tropical Africa, Ed.2, 3(1): 134.

--------------, Plant Hunting for Kew, 220 p. HMSO.

Herbert, Wm. 1820, Specierum Enumerato ex pt. Crinum broussonetii in Botanical Magazine 47: t-2121, p. 4-8, 1820.

-----------. 1821. 'A preliminary Treatise and Appendix (on Amaryllidaceae) in a special edition of Botanical Register, 1821. (Ker-Gawler publisher of B. R.; 1815-1829)

-----------. 1824, Crinum careyanum, Bot. Mag. 5:t.2466, 1824.

-----------. 1837. Amaryllidaceae and Treatise on cross-bred vegetables. London. Pages 242-275, 400-402. for Crinum.

Hermann, Paul. 1698. Paradisus Batavus, Leyden.

Hooker, J. D. 1894. Flora of British India. VI: 280-294 on Crinum.

Hornemann, J. W. 1807, 1813. Hortus regius bot. haefniensis i: 318, 1813.

Huxley, J. 1942. Evolution, the Modern Synthesis.. p. 179 (on geographical subgenera).

Index Kewensis, Kew, Index of plant species; various issues.

Jacquin, N. 1797. Hortus Schonbrunnensis, Vienna. Vol. 2:40, t.202 (1797).

Jones, K & Smith, J. B. 1967, Chromosome evolution of the Genus Crinum. Caryologia 20(2): 163-179.

Ker-Gawler, J.B. 1806 Amaryllis ornata beta. Bot. Mag. 24: t-923. (Describes C. jagus as an A. ornata variant Beta.)

----------------. 1807. C. americanum in Curtiss Bot. Mag. 26: t-1034.

----------------. 1807. C. asiaticum in Curt. Bot. Mag. 27: t.1073.

----------------. 1809. C. erubescens in Curt. Bot. Mag. 30: t.1232.

----------------. 1809 A. ornata in Curt. Bot. Mag. 29: t.1171. (Actually C. zeylanicum from Coromandel C. coast, India.)

----------------. 1810. A. ornata ex part in Curt. Bot. Mag. 33: t-1253. (Actually C. distichum Herb.)

----------------. 1817. 'A review of the genus Amaryllis (Crinum)' in Journal of Science and Arts. 2:369.

----------------. 1817. The Genus Crinum in Jour. Sci. and Arts., 3: 102-115.

Kingdon, Jonathan. 1989. Island Africa, The evolution of Africa's rare animals and plants, Princeton Press, ISBN 0-691-08560-9.

Kunth, K. S., 1843. Enumerato plantarum V: 547-584, 1843. (Genus Crinum)

Lehmiller, D. J. 1987. Observations of Crinum americanum L.; the Neches River revisited, Herbertia 43 (2): 22, 1987.

----------------1992. Crinum of Namibia. International Bulb Soc. Herbertia 48 (1 & 2): 20.

----------------1992. Crinum hildebrandtii Vatke. Int. Bulb Soc. Herbertia 48 (1 & 2): 75.

----------------1992. Namabian Crinum. Int. Bulb Soc. Herbertia 48 (1 & 2): 75.

----------------1992. Restoration of Crinum forbesii (Lindl.) Schultes emend Herbert. Int. Bulb Soc. Herbertia 48 (1 & 2): 86.

----------------1992. Transfer of C. nerinoides to Ammocharis. Novon 2,1: 33, 1992.

----------------1993. The Identity of Crinum Americanum L. Amaryllidaceae. Int. Bulb Soc. Herbertia 48 (1&2): 58.

.....................1997. Nambian Field notes and synopsis of the genus Crinum. Int. Bulb Soc. Herbertia 52: 37-65, 112-118, 119-133.

L'Heriter, C. L., 1788. Sertum Anglicum. (See 1963 Facsimilie copy for historical details of the 1786-87 period about Kew).

Linnaen Society Archives 1997 Abstract of Mary Watson-Wentworth (Marchioness of Rockingham) letters to J. E, Smith of Linn. Soc. 1788-1803.

Linnaeus, C., the younger, 1781-2 Linn f. manuscript. Herbertia 4: 91-100, 1937.

------------. 1789, Crinum erubescens L. f. ex Dryander et Aiton. Hortus Kewensis Ed. 1, Vol. l: 413.

Linnaeus C. 1738. Hortus Cliffortianus

------------. 1747. Flora zeylanicus

------------. 1753. Species Plantarum I: Crinum 291-2.

------------. 1762. Species Plantarum II: Crinum 419.

Loudon, J. C. 1830. Hortus Britannicus ii: 120.

Lyell, Sir Charles & Hooker, J. D., 1858, Tendency of Species to form Varieties; and on the perpetuation of varieties and species by natural means of selection. Linnaean Soc. 1858.

Mabberley, D. J. 1984. Crinum biflorum Rottb. in 'Two and a

half Centuries of neglected binomials'; Taxon 33 (3):

439, 1984. (See under C. series XI-B).

Miller, Philip. 1760. 'CRINUM foliis carinatis (C. erubescens)' in Icon t-110, 1756 (etching date).

Moninckx; Jan. 1689. Lilio narcissus ceylanicus Atlas I: t.34. (See Wijnand's Botany of the Commelins 1984.)

Morton, J.K. 1961. West Africa's Lilies and Orchids.

Muller-Doblies, D, 1977. Ramifications of helicoid cymes in Liliaceous Flowers = (Über den geometrishen Zusammenhang der monochasialen Verzweigungen am Beispiel einiger Liliifloren). Berlin Deutch Bot. Ges, Bd. 90: S.351-362.

Nicolson, D. H., Suresh & Manilal 1988. An Interpretation of Rheed's Hortus Malabaricus, in Regnum Vegetabile 119: p. 272, 1988.

Niklas, Ko Jo., Species and speciation in The Evolutionary biology of Plants, pages 1 to 104.

Nordal, I. 1977, 'East African Revision of the Genus Crinum in Norwegian Journal of Botany 24 (3): 179-212, 1977.

-----------, 1979. A new species of Crinum from Angola (C. trifidum sp. nov.), Norw. J. Bot. Vol. 26: 279-381.