Modified 10 Feb 2001

Collectanea Botanica
John Lindley - 1821

Amaryllis Solandraeflora undoubtedly belongs to that set of plants which has been indicated by Mr. Herbert, in some remarks on the genera of Amaryllideae lately published in the Botanical Magazine, as constituting what he considers Amaryllis proper, all of which are confined to the western hemisphere, and which he thinks form a genus altogether different from the Cape plants previously included under the same denomination. In our next article we shall take occasion to examine the principles upon which the alterations proposed by the ingenious author have been founded, and endeavor to ascertain in what cases it may be advisable to adopt them.

Amaryllis, as understood by Mr. Dryander and Mr. Ker, contains those Liliaceous plants which have an hexaploid irregular flower, and six declinate filaments inserted into the mouth of the tube, and unequal either in proportion or direction. The numerous species comprised under this definition being very different from each other in habit, Mr. Herbert has divided them into several other genera, which he calls Coburgia, Leopoldia, Galatea (afterwards Nerine), and Lycoris. His Amaryllis has flat seeds; Coburgia round seeds; Leopoldia the same seeds as Coburgia, and the outer segments of the flower completely united to the tube; Nerine radiating segments to the flower, which is divided down to the ovarium, stamens forming a cup by their dilated bases, and round or angular seeds; Lycoris a simple fimbriated stigma, and seeds supposed to be roundish. A difference between the stamens of Leopoldia, Coburgia, and Amaryllis, is mentioned by Mr. Herbert, which we have in vain sought to discover. Amaryllis purpurea is transferred to Cyrtanthus, and the species of Amaryllis with great fleshy seeds to Crinum.

Now there is a rule for forming genera which has been admirably prescribed by Linnaeus, that "Genus dabit characterem, et non character genus;" a rule which is perpetually quoted, but too often neglected; on which nevertheless all the principles of generic discrimination absolutely depend. In fact, it is the only test by which the value of the characters can be ascertained with precision. On this principle, if it can be shown that the plants combined into genera by the differences indicated above are not alike in habit, but, on the contrary, extremely dissimilar; and that others with a very similar habit are disunited, it will follow that the distinction upon which they have been divided or combined must be considered modi and not formae, if we may use that expression; and consequently of importance for distinguishing species, but not genera. How far this may really be the case, we will now proceed to inquire.

In Amaryllis are included all the South American species with the habit of equestris; that is to say, with an upright spatha and divaricating flowers. So far, therefore, the genus is natural and perhaps good; But unfortunately, A. reticulata, which has nearly all the artificial characters of Coburgia, exhibits precisely the same habit, and is moreover a native of the same country, notwithstanding Mr. Herbert's ingenious conjecture to the contrary. It is said however to form a genus of itself (Leopoldia), which would in some measure be an answer to this objection, if it really had any characters to distinguish it from Coburgia: this however not being the case it destroys not only its own genus, but Coburgia, too. Under the latter title are comprised A. blanda and Belladonna, with Brunsvigia multiflora! and Josephinae! !; plants even more unlike each other than Cyrtanthus obliquus and A. purpurea, which are referred to one genus on equally insufficient grounds. Nerine we feel willing to adopt, because we think it a very natural group; Lycoris also is possibly separated with propriety; but we have never had an opportunity of examining any species of it alive.

We do not wish it to be understood, from any thing which may have been said, that we are averse to the division of Amaryllis into several genera. On the contrary, it is, probably, necessary that some further alteration should be made to it; but on very different principles from those just examined. Of these we may perhaps have an opportunity of speaking at some future time. Our limits will not allow us to enlarge upon the subject at present, further than to add that the fleshy-seeded Amaryllides referred to Crinum by Mr. H., have not been removed from the former genus with more reason than the other plants of which we have been speaking; Crinum being more nearly related to Pancratium than to Amaryllis.

CybeRose notes:

"It is to me inconceivable that a genus proposed by one author should be interpreted by others with every original species excluded." — E. D. Merrill: An Interpretation of Rumphius's Herbarium Amboinense (1917).
William Herbert attempted to undermine Heister's Brunsvigia (1753) by combining Heister's type species with the Cape Belladonna in a new genus, Coburgia. After his attempt was dismissed by other botanists, he tried again by reviving an old confusion regarding Amaryllis Belladonna L., which allowed him to unite the Cape Belladonna with Brunsvigia orientalis (L.) Heist. and others in the genus Amaryllis. If he had been successful, both Amaryllis and Brunsvigia would have been deprived of any species described by the authors in their original publications.

Subsequently, Herbert's effort was partially ignored — orientalis remains in Brunsvigia — and we are left with the Cape Belladonna as the only "true" Amaryllis even though that species was not described in Hortus Cliffortianus (1728), as Herbert admitted, nor in Species Plantarum (1753). It is not clear whether Linnaeus intended the Cape plant to be included in Amaryllis reginae, which combines a description of one plant with plates of two others. Miller excluded Hermann's plant from his interpretation of A. regina, which is the Cape Belladonna. However, it is quite clear that L'Héritier, who did not cite Linnaeus as authority for the species, included the Cape Belladonna in his Amaryllis Belladonna (non L.).

Sealy etal. have tried to confuse A. Belladonna L. with A. Belladonna (L'Herit. non L.), with unfortunate consequences.