All the known generic synonyms of Camellia are summarized with their types clarified. Camelliastrum and Desmitus are lectotypified and Salceda is neotypified. “Kailosocarpus” and “Parapiquetia” were not validly published, Theaphylla and Tsia are illegitimate replacement names for Thea, and Kemelia and Tsubaki are illegitimate replacement names for Camellia. Nomenclatural notes on Theopsis and its type are also provided.

Camellia cuspidata, Camellia reticulata, Camelliastrum, Desmitus, lectotypification, neotypification, Salceda, Theopsis

Linnaeus (1753: 698) established the monotypic genus Camellia L. for the ornamental species, Camellia japonica L. He referenced Kaempfer’s (1712: 850, 852) description and a single illustration (Kaempfer 1712: t. 851) under C. japonica without citing any specimens in the protologue (Arts. 38.1, 38.5, 38.6 & 38.13 of the Shenzhen Code, Turland et al. 2018; hereafter ICN). Subsequently, Bartholomew (Jarvis et al. 1993: 29) designated the illustration (Kaempfer 1712: t. 851) as the lectotype of C. japonica (Art. 9 Ex. 2 of the ICN). Accordingly, this illustration serves as the type of Camellia.

The boundaries of Camellia have been gradually enlarged by several taxonomists (e.g., Sweet 1818; Seemann 1859; Sealy 1958; Chang 1981; Ming 2000). Zhang et al. (2014) argued that Camellia might be a paraphyletic group in their analyses using four plastid DNA regions and the paralogous nuclear LEAFY marker. However, the monophyly of the genus was supported by investigations using complete plastid genome data (Yu et al. 2017) and other nuclear sequences (Vijayan et al. 2009; Li et al. 2011; Zhao et al. 2022). Nevertheless, nomenclatural problems should be resolved before a comprehensive phylogenetic study of the boundaries of Camellia. Although the genus, including tea, camellias and oil camellias, has been revised several times (e.g., Sealy 1958; Chang 1981; Ming 2000), the types of many taxa in Camellia have only recently been clarified (Zhao et al. 2017a, b, 2018, 2019; Zhao 2021). However, the types of some synonyms at generic rank remain unclear and are discussed here.

Relevant collections or their images from herbaria at A, BM, E, G, IBK, IBSC, K, KUN, L, LINN, P, PE, SBT, SYS, TCD, UPS and US (acronyms following Thiers 2022), and the taxonomic literature were examined. Types were chosen based on Arts. 9 and 10 of the ICN.

Since the latest monograph of Camellia (Ming 2000), two further genera, Bembiciopsis H. Perrier (Judd 1997) and Dankia Gagnep. (Hô 1991; Zhao et al. 2017b), have been placed into the synonymy of Camellia. All 18 synonyms of Camellia (Linnaeus 1753; Adanson 1763; Blume 1825; Siebold 1832; Rafinesque 1830, 1838; Blanco 1845; Hallier 1921; Gagnepain 1939; Nakai 1940; Perrier 1940; Hu 1956, 1965) and their types (Jarvis et al. 1993; Lin et al. 2008; Zhao et al. 2016, 2017b, 2018, 2019; Zhao 2021) are listed in Table 1. “Kailosocarpus” (Hu 1957: 170) and “Parapiquetia” (Hu 1957: 170) were cited by Chang (1981: 12) in the synonymy of Camellia; however, both “Kailosocarpus” and “Parapiquetia” were not validly published since a Latin description or diagnosis was neither provided nor cited in each of the protologues (Art. 39.1 of the ICN). “Kailosocarpus” and “Parapiquetia” have no status under Art. 12.1 of the ICN and are, therefore, excluded from Table 1. The typification and nomenclatural notes of some synonyms of Camellia are elaborated below.

Cohen-Stuart (1916: 70) established Camellia sect. Theopsis Cohen-Stuart in his Ph.D. thesis, which constituted an effective publication under Art. 30.9 of the ICN. Subsequently, he translated the first two chapters of his original thesis in Dutch into English and published it (Cohen-Stuart 1919). Cohen-Stuart (1916, 1919) listed nine species, viz. C. costei H. Lév., C. cuspidata (Kochs) hort., C. euryoides Lindl., C. forrestii (Diels) Cohen-Stuart, C. henryana Cohen-Stuart, C. lutchuensis T. Itô ex T. Itô & Matsum., C. parvifolia (Hayata) Cohen-Stuart, C. punctata (Kochs) Cohen-Stuart and C. rosiflora Hook., in the key under sect. Theopsis. These nine species are presumably treated as the members of sect. Theopsis based on the structure of the key, the description of the section and the discussion. However, Cohen-Stuart (1916, 1919) did not designate a type for his sect. Theopsis. Nakai (1940) raised this section to generic rank to include 14 species and cited Cohen-Stuart’s (1919) English article (Arts. 41.1 & 41.3 of the ICN), but without selecting a type for the genus. Remarkably, four of nine species of Cohen-Stuart’s (1916, 1919) sect. Theopsis were excluded and five of them, including C. euryoides, C. forrestii, C. lutchuensis, C. parvifolia, and C. rosiflora (= C. maliflora, according to Cohen-Stuart 1916: 69, 1919: 241), were retained in the genus Theopsis (Cohen-Stuart) Nakai. Later, Sealy (1958: 14) treated Theopsis as a synonym of Camellia and resumed sect. Theopsis (Sealy 1958: 48). Chang (1981: 128) followed Sealy’s (1958) treatment and designated C. cuspidata, a species excluded from Nakai’s genus Theopsis, as the type of sect. Theopsis and this typification must be followed under Art. 10.5 of the ICN. However, when Nakai (1940) adopted Theopsis as a generic name, he did not exclude the type of Cohen-Stuart’s (1916: 70) sect. Theopsis because the section had not yet been typified, so Arts. 48.1 & 48.2 of the ICN do not apply. According to Art. 48.1 Note 1 of the ICN, this situation should be dealt with under Art. 7.3 of the ICN. Therefore, the genus Theopsis is typified by the type of its basionym, sect. Theopsis, that is, C. cuspidata, based on Chang’s (1981: 128) typification (Art. 10.5 of the ICN) even though the species was excluded from this genus (Art. 7.3 of the ICN).

However, C. cuspidata has nomenclatural problems. Kochs (1900: 586) described T. cuspidata Kochs and cited the single gathering Henry 7026. Subsequently, a name, C. cuspidata, was provided in the list of “Awards of Merit” in The Gardeners’ Chronicle (Anonymous 1912: 228). The plant was described as having “small, single, white flowers” with “pale-yellow stamens” and “narrow leaves” that were about 2–2.5 inches long. The brief description could make C. cuspidata validly published as a new species because the requirements of Arts. 32.1 & 38.1 of the ICN are likely fulfilled (also see Art. 38 Note 2 of the ICN). However, although Kochs’s T. cuspidata was neither directly nor indirectly referenced in the protologue of C. cuspidata (Anonymous 1912: 228; Arts. 41.1–41.3 of the ICN), the latter is, nevertheless, treated as a new combination based on T. cuspidata under Art. 41.4 (see Ex. 12) of the ICN.

Furthermore, two duplicates of Henry 7026 were found at K and US, viz. K000380525 (the image is available at http://www.kew.org/herbcatimg/165067.jpg) and US 00504123 (the image is available at http://n2t.net/ark:/65665/308ec5722-d414-4a4d-b733-f34c3778997b). Since Kochs (1900: 586) did not indicate a single specimen of the entire gathering as the holotype, the two duplicates at different herbaria are syntypes of T. cuspidata based on Art. 40 Note 1. When Sealy (1958: 57) cited “A. Henry 7026 (K, type-number)” under C. cuspidata, the citation could be treated as the lectotypification of the species following Arts. 7.11, 9.10 & 9.19 of the ICN. Therefore, the lectotype of C. cuspidata is that of T. cuspidata, viz. Henry 7026 (K000380525), which, in turn, serves as the type of Theopsis (Table 1).

I thank the staff of herbaria consulted for making specimens and literature available, and three reviewers and the editor, Dr. Clifford Morden, for their careful comments and corrections. The work was funded by the Scientific Research Foundation of the Central South University of Forestry and Technology (2019YJ023).