26.1. The name of any infraspecific taxon that includes the type of the adopted, legitimate name of the species to which it is assigned is to repeat the specific epithet unaltered as its final epithet, but not followed by an author's name (see Art. 46). Such names are termed autonyms (Art. 6.8; see also Art. 7.6).
Ex. 1. The variety which includes the type of the name Lobelia spicata Lam. is to be named Lobelia spicata Lam. var. spicata (see also Art. 24 Ex. 3).
Note 1. This provision applies only to the names of those subordinate taxa that include the type of the adopted name of the species (but see Rec. 26A).
26.2. A name of an infraspecific taxon that includes the type (i.e. the holotype or all syntypes or the previously designated type) of the adopted, legitimate name of the species to which it is assigned is not validly published unless its final epithet repeats the specific epithet unaltered. For the purpose of this provision, explicit indication that the nomenclaturally typical element of the species is included is considered as equivalent to inclusion of the type, whether or not it has been previously designated (see also Art. 24.3).
Ex. 2. Linnaeus (Sp. Pl.: 779-781. 1753) included 13 named varieties under Medicago polymorpha. Since M. polymorpha L. has no holotype and since no syntypes are cited, all varietal names are validly published irrespective of the fact that the lectotype subsequently chosen (by Heyn in Bull. Res. Council Israel, Sect. D, Bot., 7: 163. 1959) can be attributed to M. polymorpha var. hispida L.
Ex. 3. The intended combination "Vulpia myuros subsp. pseudomyuros (Soy.-Will.) Maire & Weiller" was not validly published in Maire (Fl. Afrique N. 3: 177. 1955) because it included "F. myuros L., Sp. 1, p. 74 (1753) sensu stricto" in synonymy, Festuca myuros L. being the basionym of Vulpia myuros (L.) C. C. Gmel.
26.3. The first instance of valid publication of a name of an infraspecific taxon that does not include the type of the adopted, legitimate name of the species automatically establishes the corresponding autonym (see also Art. 32.7 and 11.6).
Ex. 4. The publication of the name Lycopodium inundatum var. bigelovii Tuck. (in Amer. J. Sci. Arts 45: 47. 1843) automatically established the name of another variety, L. inundatum L. var. inundatum, the type of which is that of the name L. inundatum L.
Ex. 5. Utricularia stellaris L. f. (1781) includes U.
stellaris var. coromandeliana A. DC. (Prodr. 8: 3.
1844) and U. stellaris L. f. var. stellaris (1844)
automatically established at the same time. When U. stellaris
is included in U. inflexa Forssk. (1775) as a variety,
the correct name of that variety, under Art. 11.6,
is U. inflexa var. stellaris (L. f.) P. Taylor (1961).
26A.1. A variety including the type of the correct name of a subspecies, but not including the type of the correct name of the species, should, where there is no obstacle under the rules, be given a name with the same final epithet and type as the subspecies name.
26A.2. A subspecies not including the type of the correct name of the species should, where there is no obstacle under the rules, be given a name with the same final epithet and type as a name of one of its subordinate varieties.
26A.3. A taxon of rank lower than variety which includes the type of the correct name of a subspecies or variety, but not the type of the correct name of the species, should, where there is no obstacle under the rules, be given a name with the same final epithet and type as the name of the subspecies or variety. On the other hand, a subspecies or variety which does not include the type of the correct name of the species should not be given a name with the same final epithet as a name of one of its subordinate taxa below the rank of variety.
Ex. 1. Fernald treated Stachys palustris subsp. pilosa (Nutt.) Epling (in Repert. Spec. Nov. Regni Veg. Beih. 8: 63. 1934) as composed of five varieties, for one of which (that including the type of S. palustris subsp. pilosa) he made the combination S. palustris var. pilosa (Nutt.) Fernald (in Rhodora 45: 474. 1943), there being no legitimate varietal name available.
Ex. 2. There being no legitimate name available at the
rank of subspecies, Bonaparte made the combination Pteridium
aquilinum subsp. caudatum (L.) Bonap. (Notes Ptérid.
1: 62. 1915), using the same final epithet that Sadebeck had used
earlier in the combination P. aquilinum var. caudatum
(L.) Sadeb. (in Jahrb. Hamburg. Wiss. Anst. Beih. 14(3): 5. 1897),
both combinations being based on Pteris caudata L. Each
name is legitimate, and both can be used, as by Tryon (in Rhodora
43: 52-54. 1941), who treated P. aquilinum var. caudatum
as one of four varieties under subsp. caudatum (see Art. 34.2).