The typification of Cordia flavescens Aubl., the transfer of Firensia Scop. from Cordia L. (Cordiaceae, Boraginales) to the synonymy of Ocotea Aubl. (Lauraceae), and the identity of the species of Firensia

Abstract Firensia Scop. was based on Cordia flavescens Aubl., a species described and illustrated from a mixed collection that Scopoli never transferred to Firensia. The genus included three additional species formally named by Rafinesque. Currently the four species are placed in three different families and none retained the epithet accepted by Scopoli or given by Rafinesque for reason of priority. A lectotype is designated for Cordia flavescens that places Firensia in the synonymy of Ocotea (Lauraceae).


Introduction
In the preparation of the treatment of the Boraginaceae for the Flora of the Guianas, I got intrigued by the history of Firensia Scop. When Scopoli (1777: 157) described Firensia , he included only Cordia fl avescens Aubl. that is therefore the type of the genus. Contrary to what is stated by Jackson (Index Kewensis 1893, as refl ected in IPNI (http://www.ipni.org/index.html; accessed 23.01.2013), in that work C. fl avescens was not transferred to Firensia . Necker (1790, 1: 276) accepted Firensia as one of his " species naturalis " in a work where the nomenclature was uninominal and that is rejected as a source of names by the International Code of Botanical Nomenclature (McNeill et al. 2006, App. VI: 483). Th ree species were placed in Firensia by Rafi nesque (1838: 40), two of them illegitimate renaming. Other authors ignored Firensia or placed it in the synonymy of Cordia (ex. : Candolle 1845: 471 [as "Firenzia"]). Th en Kuntze (1891, 2: 977) placed them in Lithocardium "L. 1735". As this is a pre-1753 name, he created Lithocardium Kuntze with Cordia L. as a synonym. It is an illegitimate renaming of Cordia .

Methods
To solve the question of the generic and familial placement of Firensia, I studied the literature and type specimens. I looked also at printed photographs found in the cited herbarium collections, and examined scanned of specimens or photos of specimens posted online, available directly through the herbarium site, or through the sites of JSTOR or Europeana. Example: "photos F, MO!, US!; scan!". Th e Internet address to the scans is given in a note below the species.
A few typifi cations had to be made. When the type collection was known, but not the holotype, the text says: "Type:... (lectotype... ; isotypes...)". When the original description was associated with several collections of equal status or syntypes, a type collection and a lectotype were selected and the text says: "Lectotype...: ... (hololectotype...; isolectotypes...)". In both cases the information on the date of lectotypifi cation is given next to the word lectotype.

Typification of Cordia flavescens and familial placement of Firensia in the Lauraceae
Th e name Cordia fl avescens Aubl. was described and illustrated based on a mixture of fruiting branches of Ocotea commutata Nees (LAURACEAE) and fl owers belonging in Cordia . Specimens by Aublet at BM and S represent only the Ocotea element (Johnston 1935: 44). Th is off ers an opportunity to give a clear identity to the name by designating as the lectotype a specimen lacking the Cordia element.
Cordia fl avescens Aubl. is the only species included in Firensia by Scopoli (1777: 157) and is therefore the type species of Firensia Scop. Firensia falls in the synonymy of Ocotea Aubl. 1775 and belongs in the Lauraceae.

Identity of the type species and those once placed in Firensia
Th e species once placed in Firensia have been named or identifi ed as species of Cordia L., one by Linnaeus (1757: 14) and three by Aublet (1775: 219, 224, 226;t. 86, 88, 89). None of the epithets used in Firensia could be kept for reason of priority. Th e type species, Cordia fl avescens , and the three species of Firensia currently belong in three different genera and families.