Article 20

• The name of a genus is a noun in the nominative singular, or a word treated as such, and is written with an initial capital letter (see Art. 60.2). It may be taken from any source whatever, and may even be composed in an absolutely arbitrary manner, but it must not end in -virus.
  • Bartramia, Convolvulus, Gloriosa, Hedysarum, Ifloga (an anagram of Filago), Impatiens, Liquidambar, Manihot, Rhododendron, Rosa.
• The name of a genus may not coincide with a Latin technical term in use in morphology at the time of publication unless it was published before 1 January 1912 and was accompanied by a species name published in accordance with the binary system of Linnaeus.
  • “Radicula” (Hill, 1756) coincides with the Latin technical term “radicula” (radicle) and was not accompanied by a species name in accordance with the binary system of Linnaeus. The name Radicula is correctly attributed to Moench (1794), who first combined it with specific epithets.
  • Tuber F. H. Wigg. : Fr., when published in 1780, was accompanied by a binary species name (Tuber gulosorum F. H. Wigg.) and is therefore validly published even though it coincides with a Latin technical term.
  • The intended generic names “Lanceolatus” (Plumstead, 1952) and “Lobata” (Chapman, 1952) coincide with Latin technical terms and are not therefore validly published.
  • Cleistogenes Keng (1934) coincides with “cleistogenes”, the English plural of a technical term in use at the time of publication. Keng’s name is validly published because the technical term is not Latin. Kengia Packer (1960), published as a replacement name for Cleistogenes, is illegitimate under Art. 52.1.
  • Words such as “caulis”, “folium”, “radix”, “spina”, etc., cannot now be validly published as generic names.
• The name of a genus may not consist of two words, unless these words are joined by a hyphen.
  • “Uva ursi”, as originally published by Miller (1754), consisted of two separate words unconnected by a hyphen, and is not therefore validly published (Art. 32.1(c)); the name is correctly attributed to Duhamel (1755) as Uva-ursi (hyphenated when published).
  • Names such as Quisqualis L. (formed by combining two words into one when originally published), Neves-armondia K. Schum., Sebastiano-schaueria Nees, and Solms-laubachia Muschl. ex Diels (all hyphenated when originally published) are validly published.
  • The names of intergeneric hybrids are formed according to the provisions of Art. H.6.
• The following are not to be regarded as generic names:
  • Words not intended as names.
    • The designation “Anonymos” was applied by Walter (Fl. Carol.: 2, 4, 9, etc. 1788) to 28 different genera to indicate that they were without names.
    • “Schaenoides” and “Scirpoides”, as used by Rottbøll (Descr. Pl. Rar.: 14, 27. 1772) to indicate unnamed genera resembling Schoenus and Scirpus that, as stated on p. 7, he intended to name later, are token words and not generic names. These unnamed genera were subsequently named Kyllinga Rottb. and Fuirena Rottb., respectively.
  • Unitary designations of species.
    • Examples such as “Leptostachys” and “Anthopogon”, listed in pre-Tokyo editions of the Code, were from publications that are now suppressed (App. VI).

Recommendation 20A

• Authors forming generic names should comply with the following:
  • Use Latin terminations insofar as possible.
  • Avoid names not readily adaptable to the Latin language.
  • Not make names that are very long or difficult to pronounce in Latin.
  • Not make names by combining words from different languages.
  • Indicate, if possible, by the formation or ending of the name the affinities or analogies of the genus.
  • Avoid adjectives used as nouns.
  • Not use a name similar to or derived from the epithet in the name of one of the species of the genus.
  • Not dedicate genera to persons quite unconnected with botany, mycology, phycology, or natural science in general.
  • Give a feminine form to all personal generic names, whether they commemorate a man or a woman (see Rec. 60B; see also Rec. 62A.1).
  • Not form generic names by combining parts of two existing generic names, because such names are likely to be confused with nothogeneric names (see Art. H.6).