Nomenclatural revision of Cryptantha (Boraginaceae s. str.) names linked to South American taxa

Abstract During the preparation of the treatment of the genus Cryptantha Lehmann ex G. Don for South America, numerous names were identified as needing typification to stabilize their nomenclature. As a result, lectotypes are designated for 11 names and second-step lectotypes for 20 names. Furthermore, supporting information about the type material of the basionyms of four Cryptantha names already typified by Johnston (Eritrichiumtalquinum Phil., Eritrichiumdimorphum Phil., Eritrichiumcarrizalense Phil., and Eritrichiumsubamplexicaule Phil.) is provided.


Introduction
Boraginaceae (sensu Weigend et al. 2014;Luebert et al. 2016;Hasenstab-Lehman 2017) are widely distributed, mainly in tropical, subtropical and temperate regions, with a high number of poorly studied taxa in the Americas. These lands are among the most diverse ecoregions in the world and, in line with this, their plant diversity has been a challenge, especially for taxonomically complex and diverse genera such as Cryptantha Lehmann ex G. Don. This genus and close relatives, commonly known in English as

Materials and methods
In order to resolve typifications, the protologues of the treated taxa were studied and key literature (e.g., Johnston 1927;Pérez-Moreau 1976) was consulted to identify possible prior typifications. Type specimens and original material from the herbaria CORD, E, F, G, GH, HAL, K, MA, P, S, SGO, and US (herbarium acronyms after Thiers 2021+) were analyzed from images on the JSTOR Global Plants database (ITHAKA 2021) or by personal communication with herbarium curators. To proceed with the typifications, the rules of the ICN (Turland et al. 2018) and suggestions proposed by McNeill (2014) were followed. In selecting lectotypes, whenever choosing between syntypes (Art. 9.6 of the ICN), the one that shows the best quality of preservation of the important diagnostic features of the taxon was selected to preserve the current application of the names involved.
Concerning the names described by the Prussian botanist R. A. Philippi, who greatly contributed to the taxonomic knowledge of Cryptantha in South America (Philippi 1857(Philippi , 1860(Philippi , 1864(Philippi , 1873(Philippi , 1891(Philippi , 1895, a clarification of the material he used to describe numerous taxa might help to overcome some future difficulties. A relevant type collection of Philippi's names is well-known to be currently lodged at SGO (Stafleu and Cowan 1983). However, duplicates annotated and thus presumably studied by him can be traced in several other herbaria such as CORD, GH, and HAL, or previously lodged at B and then destroyed by the fire caused by the Allied bombing in 1943 (Hiepko 1987;R. Vogt, curator at B, pers. comm.). In this context, it is not at all clear in which instances there is a holotype for his names, unless the author made clear in the protologue that only a single specimen of the gathering existed. Thereby, if Philippi did not specify a single specimen and syntypes are (or were) available, it is possible to designate a lectotype (Art. 40.2 of the ICN, Turland et al. 2018).

Typifications
Lectotypes are here selected for 11 names, whereas 20 names already typified by Johnston (1927) or Pérez-Moreau (1976) were found to require second-step lectotypifications (Art. 9.17 of the ICN; Turland et al. 2018) because more than one specimen exists at the herbarium cited by them. It is worth noting that several names typified by Johnston are in need of second-step lectotypification given that the material held at SGO was unmounted and part of it was in storage by the time he visited the herbarium (Muñoz Pizarro 1960;Taylor and Muñoz-Schick 1994). On the other hand, supporting information related to the names Eritrichium talquinum Phil., Eritrichium dimorphum Phil., Eritrichium carrizalense Phil., and Eritrichium subamplexicaule Phil. is provided.
Type designations are organized into a single treatment arranged by accepted species, with a full accounting of homotypic synonyms, followed by the heterotypic synonyms (and their homotypic synonyms) in need of typification, and a discussion of the typification/s involved. Concerning numerous specimens currently found at GH, they mainly consist of fragments removed by Johnston from sheets at P and SGO; Amsinckia patagonica Speg. and Cryptantha argentea I.M. Johnst. constitute the only exception in which the material lodged at GH consist of whole specimens.  Johnston (1927) discussed this name and its original material, stating that the "type" was lodged at SGO, although two sheets annotated in Philippi's hand as "Eritrichium alfalfalis" are actually lodged there. The material found at SGO is in agreement with the locality and the diagnosis cited in the protologue. Thus, according to the Art. 9.17 (Turland et al. 2018), the choice of Johnston (1927) is here interpreted as a first-step lectotypification. From among the material available for typification purposes, the sheet SGO000004033 is here selected as a second-step lectotype of the name. The protologue of Eritrichium rigidum (Philippi, 1895) includes a direct reference to a collection made by R. A. Philippi in the valley of the Colorado River in Chile. Johnston (1927) referred to the original material of this name, stating that the "type" was lodged at SGO. However, two sheets annotated by Philippi as "Eritrichium rigidum" were found there in agreement with the protologue. Thus, Johnston's statement is here interpreted as a first-step lectotypification. In this context, the sheet SGO000004131 is here selected as a second-step lectotype of the name Notes. In describing Eritrichium gilliesii, Rudolph. A. Philippi (1873) cited a collection he made in Valle del Yeso, Chile. Two specimens of the gathering referred to by Philippi are found at SGO; these duplicates all bear original labels annotated by him with the identification of "E. gilliesii" and agree with the diagnosis cited in the protologue. Johnston (1927) clearly indicated a sheet housed at SGO as the "type", although more than one specimen is actually lodged there. Thus, Johnston's choice is here interpreted as a first-step lectotypification. In order to narrow this designation, the sheet SGO000004084 is here selected as a second-step lectotype. The original material of Eritrichium talquinum, as referred to by Rudolph A. Philippi (1895) in the protologue of the species, was collected by Friedrich Philippi in Talca, Chile. There are two sheets of apparent original material, which agree with the diagnosis and cited locality, at CORD and SGO. Both specimens were presumably studied by R. A. Philippi since they were annotated, in his hand, as "Eritrichium talquinum". Johnston (1927) noted the specimen at SGO as the "type", although the author of the name had not indicated any collection as such at the time he published the species. Thus, given that only one element at SGO satisfies the information given by Johnston, his statement is here interpreted as a lectotype designation. Chile. Four syntypes annotated in Philippi's hand as "Eritrichium apricum" and in agreement with the diagnosis cited in the protologue are found at GH, S, and SGO. The material held at GH and SGO consists of three specimens collected in Catemu in 1860; one from among them (SGO000004035), however, has a more precise reference to the place of collection (i.e., Cajón del Boldo). On the other hand, the sheet found at S (S12-25271) comes from a collection made in Aconcagua province, Chile, without reference to the date of collection. In this context, the sheet SGO000004035 is here chosen as the lectotype.

Cryptantha aprica
In describing Eritrichium bridgesii, Rudolph A. Philippi (1895) cited a collection he made in Lampa, Chile. Three specimens linked to the type collection are found at GH and SGO. The material kept at SGO bears original labels annotated by R. A. Philippi with the identification of "Eritrichium bridgesii" and agree with the diagnosis as referred to in the protologue. Johnston (1927) clearly indicated by direct citation that the type is housed at SGO with no further reference. In this context, Johnston's (1927) statement must thus be considered as a first-step typification since two duplicates of the gathering made by R. A. Philippi were found at SGO. In order to narrow this broad designation, the sheet SGO000004044 is here selected as a second-step lectotype (Art. 9.17 of the ICN; Turland et al. 2018).

Cryptantha argentea
Note. In the protologue of Cryptantha argentea, Johnston (1927) cited a collection he made near Aguada Grande, Chile, and explicitly stated that the type was lodged at GH. Two duplicates of the gathering Johnston 5814 are found there in agreement with the protologue as referred to by Johnston (1927). Thus, Johnston's (1927) statement is insufficiently precise since it cannot be ascertained to which of the two specimens at GH he was referring. In this context, the sheet GH00011520 is selected as lectotype of the name. Notes. In describing Eritrichium chaetocalyx, Rudolph A. Philippi (1860) cited a gathering he made near Pan de Azúcar, Chile. Two duplicates of the type collection in agreement with the protologue and annotated by him as "Eritrichium chaetocalyx" were located at SGO. Johnston (1927) discussed this name and its original material, referring to a sheet lodged at SGO as the "type" of the name. In this context, Johnston's (1927) statement must thus be considered as a first-step typification. Thus, in order to narrow this designation, the sheet SGO000004073 is here selected as a second-step lectotype.

Cryptantha capituliflora
In the protologue of Eritrichium divaricatum, Rudolph A. Philippi (1895) cited as type material a collection made by F. Philippi in Caldera, Chile. Johnston (1927) referred to a specimen kept at SGO as the "type" of the name, although two sheets annotated, in R. A. Philippi's hand, as "Eritrichium divaricatum" are found there. This situation determines that Johnston's (1927) statement must be considered as a first-step lectotypification. In this context, the duplicate SGO000004070 is here selected as a second-step lectotype (Art. 9.17 of the ICN; Turland et al. 2018).
Rudolph A. Philippi (1895) described Eritrichium pustulosum based on a specimen he collected in Caldera, Chile. According to Johnston (1927), the type material of the name is lodged at SGO. However, three sheets linked to E. pustulosum, in agreement with the protologue, were located at SGO, which can certainly be considered as original material. In this context, Johnston's (1927) statement must be considered as a first-step lectotypification. In order to narrow this earlier designation, the sheet SGO000004125 is here selected as a second-step lectotype.  (Stafleu and Cowan 1983). Both specimens are annotated, in Philippi's hand, as "Eritrichium dichita" and in agreement with the diagnosis found in the protologue. Johnston (1927) referred to a sheet housed at SGO as the "type" of the species name, although Philippi did not indicate any material as such at the time he published the species. Thus, Johnston's (1927) statement must be considered as a first-step lectotypification. In order to narrow this broad designation, the specimen SGO000004062 is here selected as a second-step lectotype.  Johnston (1927) referred to a specimen lodged at SGO as the "type" of the name, while Pérez-Moreau (1976) later stated "Chile. Coquimbo: Huanta, Baños del Toro, 1860-61, Volckmann (tipo de E. diffusum, SGO)". Despite the statements of these authors, in the general collection at SGO two sheets were found in agreement with the protologue. The specimens match the diagnosis as coined by Philippi. In this framework, the earliest statement made by Johnston (1927) is here interpreted as a first-step lectotypification. In order to narrow this broad designation, the sheet SGO000004064 is here selected as a second-step lectotype. Rudolph A. Philippi (1860) described Eritrichium micranthum based on a collection he made in Sandón, Chile, with no further reference. According to Johnston (1927), the type element of this species is housed at SGO. Johnston's (1927) statement is here interpreted as a first-step lectotype designation since two sheets in agreement with the protologue and annotated by Philippi as "Eritrichium micranthum" were found at SGO. The specimens involved match the protologue and, thus, the designation of Johnston (1927) is here narrowed by selecting the sheet SGO000004107 as a second step-lectotype.

Cryptantha cynoglossoides
In the protologue of Cryptantha famatinae, Brand (1924) (Stafleu and Cowan 1976), but duplicates of Hieronymus and Niederlein's collection were located at CORD and GH. These sheets agree with the diagnosis and cited locality found in the protologue. Thus, the sheet CORD00003765 is here selected as lectotype of the name. Note. Rudolph A. Philippi (1857) based the diagnosis of Eritrichium dimorphum on a collection he made in Santiago, Chile, without explicit further reference to any locality. Johnston (1927), in his revision of Cryptantha in South America, referred to a specimen lodged at SGO as the "type" of the name. However, an additional duplicate annotated, in Philippi's hand, as "Eritrichium dimorphum", is found at CORD. Both specimens are in agreement with the diagnosis as referred to by Philippi (1857). Therefore, the element cited by Johnston and his statement must be considered as an effective lectotypification of the name. Note. The protologue of Eritrichium filagineum (Philippi, 1895) includes a direct reference to a collection made by F. Philippi between Caldera and Copiapó, Chile. Two duplicates of a collection in agreement with the protologue and annotated, in R. A. Philippi's hand, as "Eritrichium filagineum" were found at SGO. According to Johnston (1927), the type element of this species is housed at SGO, and therefore his statement is here interpreted as a first-step lectotype designation. In this context, the duplicate SGO000004075 is here selected as a second-step lectotype of the name. Notes. Clos' (1849) described Eritrichium globuliferum based on two syntypes from Chile. The first collection was made by C. Gay (no. 47) in Coquimbo, while the other one was also made by him (s.n.) in Copiapó. Pérez-Moreau (1976) discussed this name and its original material, stating "Chile. Coquimbo, común en dunas de la costa, La Serena, C. Gay 47, IX-1836 (tipo de E. globuliferum, P)". However, Pérez-Moreau's statement is insufficiently precise since it cannot be ascertained to which of the two specimens at P he was referring. Thus, his choice is here interpreted as a first-step lectotypification. In order to narrow this designation, the specimen P00606762 is here selected as a second-step lectotype.

Cryptantha globulifera (Clos) Reiche, Anales
In describing Eritrichium glareosum, Rudolph A. Philippi (1864) cited a collection made in "alveo fluminis Aconcagua", Chile. Two specimens of this collection in agreement with the protologue are found at SGO. These duplicates all bear original labels annotated by Philippi with the identification of "E. glareosum" and agree with the diagnosis cited in the protologue. Johnston (1927) referred to a sheet lodged at SGO as the "type" of E. glareosum and, thus, his statement is here interpreted as a firststep lectotype designation. In order to narrow this earlier designation, the specimen SGO000004086 is here selected as a second-step lectotype.
Rudolph A. Philippi (1895) described Eritrichium carrizalense based on a Rosario Godoy de Callao collection made in Yerba Buena, Chile. A duplicate of this material, annotated in Philippi's hand as "Eritrichium carrizalense", was lodged at B and now destroyed but a photograph from Macbride's Berlin negatives (neg. 17371) is avail-able at F with a copy at GH. Furthermore, there is a duplicate at herbarium SGO and fragments of the syntype at B lodged at GH. Johnston (1927) referred to the original material of this name, stating that the "type" was lodged at SGO; in the protologue of the species, however, there is no indication of any particular specimen as such. Thus, Johnston's (1927) statement is here interpreted as a lectotype designation.
The protologue of Eritrichium sphaerophorum (Philippi 1895) includes a direct reference to a collection made by Rudolph A. Philippi near Caldera, Chile, in 1879. Johnston (1927) clearly indicated a sheet housed at SGO as the "type", although an additional element, also annotated in Philippi's hand as "Eritrichium sphaerophorum" was located there in agreement with the protologue. In this context, Johnston's (1927) choice is here interpreted as a first-step lectotypification. In order to narrow this designation, the specimen SGO000004137 is selected as a second-step lectotype.  Johnston (1927) referred to a specimen kept at SGO as the "type" of the species name, while Pérez-Moreau (1976) later discussed this name and its original material, stating "Chile. Coquimbo, Las Mollacas, Cordillera de Illapel, s/c, 1-1888 (tipo de E. glomeruliferum, SGO)". However, the statements made by the authors are insufficiently precise since it cannot be ascertained to which of the specimens at SGO they were referring. Thus, the earliest choice of Johnston is here interpreted as a first-step lectotypification. In order to narrow this designation, the specimen SGO000004088 is here selected as a second-step lectotype.  (1799) based the description of this species on two syntypes they had collected during their stay in the Americas. The first collection was made in Lima, Peru, while the other one consists of a collection from Chancay, Peru. Two specimens linked to this name were located at MA in full agreement with the diagnosis and localities cited in the protologue. From among the material available for typification purposes, the specimen MA814804 from Chancay is preferred over the material from Lima (MA814805) since it bears an original label annotated, in Pavón's hand, as "Myosotis granulosa". Therefore, it is here selected as the lectotype of the name. Note. Rudolph A. Philippi (1895) based the diagnosis of Eritrichium virens on a collection made by Guillermo Geisse in Copiapó, on the road to Chañarcillo, Chile. Johnston (1927), in his revision of South American Cryptantha, referred to a specimen lodged at SGO as the "type" of the name. However, two duplicates studied by Philippi were located at SGO in agreement with the protologue. Thus, Johnston's choice is here interpreted as a first-step lectotypification. In order to narrow this designation, the specimen SGO000004149 is here selected as a secondstep lectotype.  Johnston (1927) referred to a specimen lodged at SGO as the "type" of the name. However, two duplicates studied and annotated as "Eritrichium longifolium" by Philippi were located there in agreement with the protologue. Thus, Johnston's (1927) choice is here interpreted as a first-step lectotypification. In order to narrow this broad designation, the specimen SGO000004102 is here selected as a second-step lectotype. Note. In describing Eritrichium spathulatum, Rudolph A. Philippi (1895) cited a collection made by F. Philippi in Colchagua de Popeta, Chile. Two duplicates of this gathering in agreement with the protologue and annotated as "Eritrichium spathulatum" by Philippi are found at SGO. Johnston (1927) referred to the "type" as lodged at SGO and, therefore, his statement must be considered as a first-step lectotypification. In order to narrow this broad designation, the specimen SGO000004135 is here selected as a second-step lectotype. Note. In the protologue of Eritrichium subamplexicule, Rudolph A. Philippi (1860) cited a collection he made near Paposo, Chile. There are two sheets of apparent original material, which agree with the diagnosis and cited locality, at HAL and SGO. Both specimens were presumably studied by Philippi since they were annotated, in his hand, as "Eritrichium subamplexicaule". The material at SGO was referred to as the "type" of the species name by Johnston (1927). Therefore, the element cited by him must be considered as an effective lectotypification. Note. Rudolph A. Philippi's description of Eritrichium chrysanthum (Philippi 1864) was based on material collected by H. Volckmann near Illapel, Chile. Johnston (1927) indicated a specimen housed at SGO as the "type". However, two sheets linked to E. chrysanthum were located at SGO and, therefore, his statement must be considered as a first-step lectotypification. In order to narrow this broad designation, the specimen SGO000004051 is here selected as a second-step lectotype.