Hesperis rupestris Boiss. & Noë in Boiss., Diagn. Pl. Orient. ser. 2, 1: 22. 1854 [nom. illeg.] [non Pall.].

Type: “Hab. ad Bakker Maden Armeniae meridionalis in fissuris rupium ubi floriferam Maio 1852 legit cl. Noë”.

Holotypus: TURKEY: “Bakker Madem in fissuris rupium”, V.1852, Noë 810 (G-BOIS [G00332233]; iso-: BM [BM001254063], G [G00446101], KW [KW000127975], LE [LE00013091], P [P00234979], W [P18890310423]) .

= Hesperis pendula subsp. campicarpa (Boiss.) F. Dvořák in Spisy Přír. Fak. Univ. J. E. Purkinje Brně 491: 113. 1968.

Notes. – None of the duplicates above was annotated by Boissier and no other duplicates were found elsewhere. Therefore, it is safe to conclude that Boissier based the species description solely on the unicate in his herbarium.

FOURNIER (1866) treated H. rupestris as a variety of H. pendula, BOISSIER (1867a) reduced it and the previous entry to synonymy of H. pendula, CULLEN (1965: 458) recognized it as a distinct species, and DVOŘÁK (1968c) reduced it to synonymy of H. pendula var. campicarpa . Position of the last author is tentatively acccepted here.

The earliest homonym H. rupestris Pall. has recently been rejected by GERMAN (2013), whereas that of Rafinesque is an illegitimate combination in Matthiola (CANDOLLE, 1821a) . By contrast KUNTZE’s (1891) later homonym was based on the South American endemic Eudema rupestre Bonpl., and together with many New World Brassicaceae, they were transferred by him to Hesperis, a genus restricted to Eurasia.