The Taxonomic Circumscription and Nomenclatural History of Pilosella suecica (Asteraceae): A Special Case of Grey Literature in Taxonomic Botany

The taxonomic history, nomenclature and application of the oldest species names available for the common hybrids between Pilosella caespitosa and P. lactucella are reviewed. Elias Fries created a nomenclatural and bibliographical collision when he replaced a printed label of his exsiccata Herbarium normale with its second version, distributed at a later date, in which the protologue of Hieracium suecicum had appeared. In this protologue, the new species name was validly published with a mere reference to the original description of H. auricula var. majus, thus being based on the type of the latter. In a later fascicle of the same exsiccata, Fries excluded this synonym and distributed a different morphotype of H. suecicum, which caused taxonomic confusion and re-description of the same taxon under the name H. fennicum. The surviving original material of H. auricula var. majus is rejected, and its neotype is designated, making H. suecicum the correct name for the hybrids strictly intermediate between P. lactucella and P. caespitosa. Such hybrids constitute the most common hybridogenous taxon of Pilosella in Scandinavia, Finland and neighbouring Russia, with many synonyms described from this area and partly typified here. Another hybridogenous taxon of the same origin, more similar to P. lactucella and previously known as P. cochlearis, is correctly named P. stipitiflora comb. nov. The nomenclatural value and bibliographic complexity of exsiccata, a commonly underestimated kind of grey literature in taxonomic botany, are further highlighted.


Introduction
Hybridisation is a common phenomenon in vascular plants.It has long been recognised as a fundamental evolutionary process [1,2], and many ancient taxonomic lineages may appear to have a hybrid background [3].Hybridisation, as a process of reticulate evolution, commonly leads to alloploid speciation, which is often connected with polyploidy [4,5].
In contrast to hybridogenous speciation, which leads to the formation of novel evolutionary lineages, introgression is viewed as limited gene exchange between otherwise established species [6], which does not have an evolutionary effect other than changing fitness to the current and potential distribution environment, thus leading to local success and range expansions [7].
Our study of the original collection of P. floribunda (Wimm.and Grab.)Fr. [25], which was recovered at the Komarov Botanical Institute, confirmed the historically predominant application of this species name to a triple hybrid between P. caespitosa (Dumort.)P.D.Sell and C.West s.l.(syn.Hieracium pratense Tausch), P. lactucella (Wallr.)P.D.Sell and C.West and P. praealta (Vill.ex Gochn.)F.Schultz and Sch.Bip.This lectotype agreed with the current use of the species name in Northern and Eastern Europe [22,26,27] and narrowed the broader taxonomic concept used in Central Europe [28].
The name P. suecica was incorrectly synonymised with the broadly defined P. floribunda on the belief that the latter is applicable to the binary hybrids between P. caespitosa and P. lactucella [28,29].Another lumping concept, even more inclusive, subordinated P. suecica and P. floribunda to P. caespitosa on the assumption of their common hybrid origin [30].With the lectotypification of P. floribunda [25], P. suecica became the correct name for the binary hybrid, and its lectotypification and synonymy are, therefore, of primary nomenclatural interest.This research is the main subject of the present contribution.
Similar taxa of the same hybrid origin, P. fennica and P. cochlearis, were recently accepted in some major taxonomic works and compilations [22,28].Their identities and relationships with P. suecica are also clarified here.
Our previous work [31] has demonstrated that published exsiccata, a special kind of botanical publication and dissemination of information on plant diversity [32], have the utmost importance for the Hieracium nomenclature.The present contribution uncovers further aspects and complexities of this type of grey literature, which is still much underexplored and under-evaluated in plant taxonomy.
Diagnostic characters of the parental species were used for the inference of the hybrid origin already by Nägeli and Peter [34], and this approach is still considered useful in the diagnostic of Pilosella (cf.[25]).The most important diagnostic characters of P. caespitosa are a violet colouration of the stem base with abundant red-based hairs, dark green leaves with simple hairs on both surfaces and stellate pubescence on the lower side, and dark styles.The hybridisation with P. lactucella is recognisable through the yellowish-glaucous colour of leaves, which tend to be spathulate, undulate, glabrescent, with shorter and broader petioles, by low and slender stems, and by numerous slender glandular hairs in the synflorescence and along the stem; a prominent feature of this species is the sulphureous tint of its flowers, which is often recognisable in hybrids.
These characters are variously expressed in hybrids, with the dominance of diagnostic features of either of the parental species.Taking this dominance into account, three main variants can be distinguished among the hybrids between P. caespitosa and P. lactucella, which can be designated using hybrid formulas and binary names according to the standard adopted after Zahn's monograph [20]: P. caespitosa < P. lactucella: P. stipitiflora (Nägeli and Peter) Sennikov; P. caespitosa × P. lactucella: P. suecica; P. caespitosa > P. lactucella: P. colliniformis (Nägeli and Peter) Dostál (Table 2).In short, P. colliniformis is very similar to P. caespitosa but differs from the latter in the involucres with broad pale margins and in the leaves with a glaucous tint [33], which are partly subglabrous above.This morphotype is a hybrid with P. lactucella, which is most similar to P. caespitosa [34].Pilosella stipitiflora closely resembles its other parent, P. lactucella, but differs in the greenish (vs.yellowish-glaucous) leaves and dark or black (vs.purely yellow) styles.It also differs from P. suecica, the intermediate variant, in its subglabrous (vs.ciliate along margins) leaves with the yellowish-glaucous (vs.greenish to green) colour, synflorescence branches with very few (if any) simple hairs and sulphureous (vs.yellow) flower colour.
The hybrids between P. caespitosa and P. lactucella are polymorphic, and there is no feasible limit between the members of this complex, neither in morphology nor in crossing barriers.However, their distinction stands for nearly 150 years, and its practical utility is therefore proven by time.The need for this taxonomic distinction is based not only on the contrasting morphology of the hybrids (some of them are hardly distinguishable from the parents by non-experts) but also on their different ecological preferences, usually coinciding with those of the parental species.
The correspondence between the historical applications of these hybrid names is given in Table 3.A broad variety of names used for these hybrids, also in various senses, indicate that the historical taxonomic circumscriptions in Pilosella were quite vague, and many (if not most) species names were variously misapplied and misinterpreted during their history.The nomenclature of Pilosella has been highly unstable, and no really established usage can be claimed in Europe.Stems: 30-60 cm tall, erect, firm; lowermost internode short, pale to deep violet; with numerous simple hairs up to 3.5 mm long and with little or without stellate pubescence below, with scattered glandular hairs 0.2-0.4mm, rare simple hairs and very lax stellate pubescence under the inflorescence.Creeping stolons: short or rather long, with spathulate leaves almost lacking stellate pubescence on the lower surface.Leaves: pale to intensely green, basal in a rosette; rosulate 5-13 × 0.9-1.2cm, spathulate to oblong, shortly acute, shortly narrowed towards the base, with spiculiform teeth; cauline 1-2 (mostly in the basal part of the stem), like the basal but smaller and less narrowed to the base; all glabrous above, with some rigid, violet-based simple hairs along the middle nerve and the margin and solitary to rare stellate hairs along the middle nerve beneath, uppermost subglabrous.Inflorescence: corymbiform, usually compact, with 7-10 capitula; branches: rather firm, with abundant blackish glandular hairs 0.3-0.5 mm long, solitary simple hairs 2-2.5 mm long and dense stellate indumentum.Involucral bracts: inner 7-8.5 × 0.9-1.1 mm, olive or blackish-green with dark margins, broadly acute at the apex, with rare thin dark simple hairs ca. 2 mm long, rather dense thin black glandular hairs 0.5-0.6 mm long and scarce stellate pubescence throughout.Flowers: 12-13 mm long; ligules yellow.Styles: dark or pale dark.Achenes: ca. 2 mm long.Ecology: Natural and disturbed meadows and grasslands in secondary habitats (roadsides, yards, and pastures).
Distribution: Native in Central and Northern Europe [28]; in Finland northwards up to Ostrobottnia ultima and Regio kuusamoensis, abundant [29]; in Eastern Europe northwards up to Russian Karelia [22,37].Alien outside Europe: in Russian Siberia [38] and North America [39].Stems: 15-25 cm tall, erect, slender, pale green; with few to scattered simple hairs (dark-based in the lower part) and very sparse stellate pubescence, with numerous glandular hairs 0.2-0.4mm.Creeping stolons: short or rather long, with narrowly spathulate leaves lacking stellate pubescence on the lower surface.Leaves: yellowish-green or pale green, basal in a rosette; rosulate 3-10 × 0.5-1 cm, narrowly spathulate or oblanceolate, obtuse, gradually narrowed to a broad or narrow petiole, with spiculiform teeth; cauline 1-2 (mostly in the basal part of the stem), like the basal but smaller and less nar- Stems: 15-25 cm tall, erect, slender, pale green; with few to scattered simple hairs (darkbased in the lower part) and very sparse stellate pubescence, with numerous glandular hairs 0.2-0.4mm.Creeping stolons: short or rather long, with narrowly spathulate leaves lacking stellate pubescence on the lower surface.Leaves: yellowish-green or pale green, basal in a rosette; rosulate 3-10 × 0.5-1 cm, narrowly spathulate or oblanceolate, obtuse, gradually narrowed to a broad or narrow petiole, with spiculiform teeth; cauline 1-2 (mostly in the basal part of the stem), like the basal but smaller and less narrowed to the base; all glabrous except for the basal part, with some rigid simple hairs along the middle nerve and the margin.Inflorescence: loosely corymbiform or irregularly branched, with 3-7 capitula; branches: slender, with abundant blackish glandular hairs 0.2-0.4mm long and with or without solitary simple hairs and thin stellate indumentum.Involucral bracts: inner 6-7 × 0.8-1 mm, olive-green with pale margins, broadly acute at the apex, with few to sparse thin dark simple hairs 1-2 mm long, rather dense thin black glandular hairs 0.4-0.5 mm long and scarce stellate pubescence throughout.Flowers 10-11 mm long; ligules sulphureous.Styles pale dark or dark.Achenes: ca. 2 mm long.
Ecology: Natural and disturbed meadows and grasslands in secondary habitats (roadsides, yards and pastures).
Distribution: As for Pilosella suecica.This hybrid often occurs in the presence of P. lactucella.

Confusing History of Hieracium suecicum
Fries [35] separated Hieracium suecicum Fr. and placed it between Hieracium auricula L. (now Pilosella lactucella) and Hieracium floribundum (Pilosella floribunda).According to our comparisons of his diagnostic characters (Table 4), Fries differentiated this new species from P. lactucella using flat, broader, obovate leaves (vs.undulate, narrower, ligulate), larger corymbose inflorescence and prominently dark styles.Pilosella floribunda was said to differ from the new species by narrower (lanceolate or spathulate) leaves and yellowish styles; this description was based on the plants belonging to P. floribunda in the sense of its lectotype, i.e., hybrids with P. praealta [25].In 1843, Fries [40] distributed a specimen from eastern Sweden under the name "Hieracium dubium L.", which appeared on its printed label (Figure 5a) and in the index to the exsiccata set.This set was distributed to main herbarium institutions in the first turn, e.g., to Helsinki (pers.obs.) and Berlin [34].Afterwards, Fries found this species name inappropriate because of its ambiguous application.He decided that this taxon should be recognised as a new species, which he named H.suecicum.To reflect this name change, between 1843 and 1845, Fries [41] printed and distributed a new version of the same label (Figure 5b), in which the new species name appeared and was accompanied by a synonym, "H.Auricula β. majus.Wahlenb.Suec.et Ups."This reference to the previously published descriptions of H. auricula var.majus Wahlenb., including its protologue [42][43][44], makes the new species name validly published as a replacement name for this variety (Art.6.12 [45]), thus based on the nomenclatural type of the latter (Art.7.4 [45]).As the protologue of H. suecicum was published in the most unusual and obscure place (a replacement label for the exsiccata, which was present only in the sets distributed after the name change, in a limited number of copies available in herbarium collections rather than libraries), it attracted very scarce attention from taxonomic researchers (e.g., [46]), most of whom are still used to give credit for the valid publication of this species name to the later monograph [35].The exact publication time of this label is uncertain: it postdates the original distribution of the exsiccata set [40], in which the old name appeared, but predates a checklist of Scandinavian plants by Fries [47], in which this label was cited under the changed name.
Plants 2024, 13, 1301 12 of 20 published descriptions of H. auricula var.majus Wahlenb., including its protologue [42][43][44], makes the new species name validly published as a replacement name for this variety (Art.6.12 [45]), thus based on the nomenclatural type of the latter (Art.7.4 [45]).As the protologue of H. suecicum was published in the most unusual and obscure place (a replacement label for the exsiccata, which was present only in the sets distributed after the name change, in a limited number of copies available in herbarium collections rather than libraries), it attracted very scarce attention from taxonomic researchers (e.g., [46]), most of whom are still used to give credit for the valid publication of this species name to the later monograph [35].The exact publication time of this label is uncertain: it postdates the original distribution of the exsiccata set [40], in which the old name appeared, but predates a checklist of Scandinavian plants by Fries [47], in which this label was cited under the changed name.
(a) (b) In the protologue of H. suecicum, Fries [41] also referred to "H. dubium L. meo sensu, sed nomen ambiguum mittendum".This text is not a nomenclatural reference (because Fries abandoned H. dubium as an ambiguous and therefore unusable name and did not include it as a nomenclatural synonym) but a reference to his earlier [40] misapplication for the same plant.Similarly, this synonym was cited pro parte (regarding Swedish plants only, i.e., excluding foreign synonyms, one of which subsequently provided a lectotype of this species name [17]) in the first Hieracium monograph [35].For this reason, H. suecicum is not a superfluous and illegitimate name as erroneously believed by Bräutigam and Greuter [28].
Fries incorporated the new taxonomy into his checklist of Scandinavian vascular plants and cryptogams [47] and his first monograph on the taxonomy of Hieracium worldwide [35].Other researchers (e.g., [48]) adopted the change, using the specimen in the exsiccata as a taxonomic reference.But, after a few years, Fries changed his mind and noted that the plants of "H.dubium" in his earlier opinion [40] may be identical to H. auricula var.majus, which is not a synonym of H. suecicum that differs from the aforementioned variety in its yellow flowers with red stripes beneath (vs.sulphureous flowers in H. auricula = P. lactucella).Fries published the new circumscription of H. suecicum on herbarium labels in another issue of his exsiccata [49], adding a diagnosis of the flower colour.The plants that he distributed as the new "type" of H. suecicum were similar to the previous variant but demonstrated much greater morphological proximity to P. lactucella (stems slender, less hairy, leaves smaller, less greenish-coloured and much less hairy, flowering heads smaller).
Indeed, in his second monograph of Hieracium, Fries [36] confirmed the placement of the former H. dubium among the synonyms of H. auricula = P. lactucella and restored the variety H. auricula var.majus, but this variety was accepted in his own circumscription [50] for a vigorous variant of the species.The variety started its history from his earlier In the protologue of H. suecicum, Fries [41] also referred to "H. dubium L. meo sensu, sed nomen ambiguum mittendum".This text is not a nomenclatural reference (because Fries abandoned H. dubium as an ambiguous and therefore unusable name and did not include it as a nomenclatural synonym) but a reference to his earlier [40] misapplication for the same plant.Similarly, this synonym was cited pro parte (regarding Swedish plants only, i.e., excluding foreign synonyms, one of which subsequently provided a lectotype of this species name [17]) in the first Hieracium monograph [35].For this reason, H. suecicum is not a superfluous and illegitimate name as erroneously believed by Bräutigam and Greuter [28].
Fries incorporated the new taxonomy into his checklist of Scandinavian vascular plants and cryptogams [47] and his first monograph on the taxonomy of Hieracium worldwide [35].Other researchers (e.g., [48]) adopted the change, using the specimen in the exsiccata as a taxonomic reference.But, after a few years, Fries changed his mind and noted that the plants of "H.dubium" in his earlier opinion [40] may be identical to H. auricula var.majus, which is not a synonym of H. suecicum that differs from the aforementioned variety in its yellow flowers with red stripes beneath (vs.sulphureous flowers in H. auricula = P. lactucella).Fries published the new circumscription of H. suecicum on herbarium labels in another issue of his exsiccata [49], adding a diagnosis of the flower colour.The plants that he distributed as the new "type" of H. suecicum were similar to the previous variant but demonstrated much greater morphological proximity to P. lactucella (stems slender, less hairy, leaves smaller, less greenish-coloured and much less hairy, flowering heads smaller).Indeed, in his second monograph of Hieracium, Fries [36] confirmed the placement of the former H. dubium among the synonyms of H. auricula = P. lactucella and restored the variety H. auricula var.majus, but this variety was accepted in his own circumscription [50] for a vigorous variant of the species.The variety started its history from his earlier publication [51] rather than the actual protologue [42], but the varietal name was not validly published in that place because of the lack of a separate description.
In 1862, Fries [36] maintained H. auricula var.majus of Wahlenberg [42] as a partial synonym of H. suecicum, and interpreted it on the basis of its locality, Wahlenberg's personal communication and herbarium collections.He did not refer to the suitability of the descriptions of the taxon in Wahlenberg's works, probably because of their apparent mismatch: Wahlenberg [42][43][44] described plants with regularly hairy leaves, whereas Fries [35,36] stated that his H. suecicum has subglabrous leaves that are hairy only along the basal part of their margin.Fries' description did not match the plants distributed in the first time [40], but perfectly agreed with the plants of the second distribution [49], which he considered typical of the species.
As the varietal name H. auricula var.majus Wahlenb.provides the type of H. suecicum Fr., its typification is required to establish the nomenclatural application of the latter name.Almquist [52] examined the original material of Wahlenberg in the herbarium collections at Uppsala and reported that it contains three plants on a single herbarium sheet (Figure 6), two being complete and identifiable and the third one currently lacking flowers and therefore dismissed from consideration.The specimen (probably plant 2 in particular) was identified as H. floribundum subsp.suecicum (Fr.)Nägeli and Peter by H. Dahlstedt, the most renowned Swedish expert on Hieracium s.l.Almquist disagreed on this identity and assigned the first plant (with stems and leaves abundantly covered by long simple hairs) to H. pratense subsp.colliniforme (Peter) Zahn and the second plant (with stems and leaves subglabrous) to H. auricula = P. lactucella s.str.I partly agree with both interpretations in a way that the plants are referable to a hybrid complex between P. lactucella and P. caespitosa s.l., to which the name P. suecica belongs; the first plant represents a morphotype most resembling P. caespitosa (=H.collinum subsp.colliniforme Peter [34]), and the second plant is a variant more closely approaching P. lactucella (=H.suecicum or H. floribundum subsp.suecicum [34,36,49,53]).
A comparison of diagnostic characters stated in the protologue of H. auricula var.majus [42] with the same characters of preserved specimens in Herbarium Wahlenberg at UPS and the exsiccata distributed by Fries [40,49] (Table 5) unambiguously shows that Wahlenberg described the hybrids between P. lactucella and P. caespitosa, and the first interpretation by Fries [40] is matching the protologue.However, the second interpretation by Fries [49] is fully congruent with the second specimen in the original collection of Wahlenberg, which was indirectly referred to by Fries in 1862 [36].Neither of the original specimens in Herbarium Wahlenberg agrees with the protologue, being in serious conflict with some major diagnostic characters used by Wahlenberg [42]: plant height (first specimen) or pubescence (second specimen).Specimens that are part of the original material but in conflict with the protologue cannot be used for lectotype designation because this type of choice can be superseded under Art.9.19 [45].Almquist [52] alluded that the first (hairy) specimen of the original material is an alien plant; it has, therefore, been collected by chance because of its unusual appearance.Similarly, the second (subglabrous) specimen may have been another chance collection that is linked with the protologue [45] but does not necessarily belong to the intended taxonomic circumscription.As no part of the available original material of H. auricula var.majus is suitable for lectotypification, a neotype may be designated under Art.9.13 [45].For the neotype designation, I propose a specimen distributed in the exsiccata [40] that was the first and most appropriate interpretation of the protologue.The neotype specimen (Figure 1a,c) has the most vigorous plants of the collection, with well-developed inflorescences and partly branched stems.Their leaves are ciliate along the margins and sparsely hairy on surfaces, in agreement with the original description of H. auricula var.majus in the protologue [42].

Early Synonyms of Hieracium suecicum
Norrlin [54] accepted the name Hieracium suecicum according to its reinterpretation [36,49], as a hybrid with the prominently dominating characters of P. lactucella.Ahead of the other Scandinavian authors, he held a narrow taxonomic concept and distinguished closely related and morphologically similar morphotypes of different hybrid origins.Since Fries shifted the circumscription of his H. suecicum, the hybrid between P. caespitosa and P. lactucella with intermediate morphology required a new name.
This hybrid was described as H. suecicum subsp.fennicum Norrl.[54] based on several specimens collected in three Finnish historical provinces: Tavastland (Häme), Savolaks (Savo) and Norra Karelen (Pohjois-Karjala), i.e., the central part of southern Finland.This name was quickly elevated to the species rank under Hieracium and Pilosella [53,55] and subsequently accepted in major taxonomic treatments [22,56].A good specimen collected by Norrlin from Häme, his home province and main area of botanical research at that time, is designated here as lectotype (Figure 1b,d).
In his first detailed treatment of the Scandinavian Pilosella, Norrlin [53] provided a more detailed classification, in which he separated further subspecies and varieties from P. suecica and P. fennica.One of the new segregates, P. suecica subsp.cochlearis Norrl., was characterised by essentially the same characters as P. fennica in the previous works but deviated from the latter in smaller flowering heads.This segregate was accepted for the hybrid with the dominating characters of P. lactucella by some later authors (e.g., [23]), although its original specimens have the intermediate characters of their inflorescences and leaves.A specimen with larger plants distributed by Norrlin in his exsiccata [57] is designated here as lectotype (Figure 2a,c) to confirm this taxonomy.The specimens distributed by Norrlin in the later exsiccata [58] belong to the hybrid with the dominating characters of P. lactucella and may have had an influence on the subsequent acceptance of the species name in that sense.
Nägeli and Peter [34] published the most elaborated treatment of Pilosella in "Central Europe", which, despite its restrictive title, included taxa from Scandinavia and Finland.
They made extensive use of Norrlin's exsiccata and accepted most of his taxa, albeit not necessarily in the same circumscription.They renamed Norrlin's P. suecica subsp.cochlearis but distributed its syntypes between two subspecies, H. floribundum subsp.suecicum (Fr.)Nägeli and Peter and H. floribundum subsp.cochleatum Naeg.and Peter.The latter is not a nomenclatural replacement for P. suecica subsp.cochlearis because it included a reference to a single (and untypical) syntype of Norrlin's subspecies name rather than to its protologue.For this reason, the subspecies name, corrected by Nägeli and Peter [34] to avoid later homonymy at the species rank, is the name of a new taxon with its own type material [31]; it is not a superfluous and, therefore, illegitimate replacement as incorrectly interpreted by Bräutigam & Greuter [28].In agreement with this interpretation, its lectotype (Figure 2b,d) is designated from the exsiccata [57] cited in its protologue.This specimen also belongs to the hybrid with the intermediate morphology but is represented by small plants with depauperate inflorescences.
Contrary to the treatment of Bräutigam and Greuter [28], the combination P. cochlearis was not validly published by Norrlin, although it appeared in print in his taxonomic treatment [53] and on a single label in his exsiccata [57].The definitive classification in these works was provided by Norrlin in the list of accepted taxa in the taxonomic treatment [53] (pp.173-174) and in the index to the exsiccata [57], whereas the names appearing in the other parts of the text and on the label were provisional [31].Pilosella cochlearis remained validly unpublished until its acceptance by Soják [59], who provided conditions for its valid publication and correctly noted that Norrlin did not accept the species.
Nägeli and Peter [34] included a set of Elias Fries' Herbarium normale in their work.They used many specimens from these exsiccata to establish further new taxa.The specimens distributed as H. dubium [40] were named H.floribundum subsp.ciliatifolium Nägeli and Peter (Figure 3a,c), which is, therefore, a nomenclatural synonym of H. suecicum by our typification of the latter.The specimens distributed as H. floribundum [60] were described as H. floribundum subsp.scissum Nägeli and Peter (Figure 3b,d).These specimens do not belong to H. floribundum according to its common historical interpretation and our lectotype [25] but are referable to the hybrid with the intermediate morphology, which is named H.suecicum here, only slightly deviating towards P. lactucella by the shape of leaves.As the personal collection of Peter no longer exists [61], we designate lectotypes of both names from the exsiccata set at the University of Helsinki.
These lectotypifications provide the earliest synonyms of P. suecica.Further synonyms are expected from the Finnish Hieracium works [31] and elsewhere, but this synonymy requires a thorough inventory of the original material, which is still pending.

Pilosella stipitiflora Is the Correct Name for "P. cochlearis"
The hybrid between P. caespitosa and P. lactucella, which strongly resembles the latter parent, is morphologically distinct, and its recognition is, therefore, practical (cf.[22,23].As the previously accepted names, i.e., P. suecica [22,34,53] or P. cochlearis [23], represent the morphologically intermediate hybrid, they are not suitable, and another name should be applied.
In the concept of Norrlin [53], P. suecica was accurately circumscribed to include plants resembling P. lactucella.Nägeli and Peter [34] separated a specimen distributed by Norrlin under the name P. suecica into a new subspecies, H. floribundum subsp.stipitiflorum Nägeli and Peter, which represented a minor deviation towards P. caespitosa.This taxon has been elevated to the species rank by Brenner [62] and currently provides the earliest legitimate name for this hybrid taxon.It is lectotypified here with a specimen from Norrlin's exsiccata, and a new combination in Pilosella is effected to conform to the current taxonomy.

Materials and Methods
Historical herbarium collections are examined de visu at the University of Helsinki (H) and as digital images at the University of Uppsala (UPS).As the main Hieracium collections of Albert Peter have been destroyed [61], lectotypes of plant names established by this author are selected from the exsiccata cited in the protologues (sets kept at H). Lectotypes of plant names established by J.P. Norrlin are designated from his personal collection.Plant morphology, variability and distributions are evaluated on the basis of the herbarium collections at H and personal field observations.The taxonomy of hybrid species of Pilosella follows the long-established tradition [19,20,22,23,34] in the acceptance of more than one species-level hybridogenous taxon between two parental species.
The nomenclature of the Finnish taxa is based on my previous inventory [31] with minor updates.Nomenclatural evaluations and decisions are based on the current International Code of Nomenclature for algae, fungi and plants [45].Plant names are applied according to their nomenclatural types, except for Pilosella caespitosa (Dumort.)P.D.Sell and C.West, P. colliniformis (Nägeli and Peter) Dostál and P. praealta (Vill.ex Gochn.)F.Schultz and Sch.Bip., whose types are currently lacking, which are used in agreement with the main recent treatments [22,23,33].

Conclusions
The newly discovered protologue of Hieracium suecicum is based on H. auricula var.majus, whose original material is variable and not matching the protologue, but the designated neotype agrees with the original understanding and the current use of this species name in Northern and Eastern Europe.Pilosella suecica is the earliest correct name for all hybrids between P. caespitosa and P. lactucella, to which the name P. floribunda has been misapplied.
Numerous species names have been applied to various morphotypes resulting from the crosses between P. lactucella and P. caespitosa s.l., which demonstrate more or less intermediate morphology but may approach either of the parental species in their diagnostic characters.A few of such species names (P.cochlearis, P. fennica, P. suecica) have been recently in use by some authors [22,23].I advise against the taxonomic recognition of such morphotypes because of their recurrent polytopic origin and the lack of morphological, ecological and biological delimitation.
A hybrid variant of P. lactucella with the introgression from P. caespitosa is morphologically close to P. lactucella but clearly differs from the latter in its spathulate and greenishglaucous (vs.lingulate and yellowish-glaucous) leaves, more robust stems with scattered dark simple hairs (vs.very slender stems with few pale simple hairs), inflorescences corymbose (vs.loosely branched), involucres dark (vs.pale) and styles dark or black (vs.yellow).The taxonomic recognition of such plants is justified by their close morphological and ecological proximity to their parent (P.lactucella), which makes them dissimilar to the morphologically intermediate hybrids; yet they cannot be taxonomically treated as part of the species because of their interspecific hybrid origin.The correct name for these hybrids is P. stipitiflora.Such plants can be further placed in the P. suecica aggr. in the current classification of Pilosella [28].
The exsiccata published by Elias Fries, Herbarium normale plantarum rariorum et criticarum Sueciae, were standard reference for understanding taxonomic concepts in vascular plants of Scandinavia (with the neighbouring territories of Finland and Russia) in the 19th century.However, due to their old age and obscure bibliographic information, this exsiccata, as a nomenclatural reference, clearly belong to the corpus of grey literature, whose relevant inaccessibility and complexity may significantly hinder research in plant nomenclature [63].Irregularities in their publication process have been already noted with the discovery of casual supplements [64]; the present contribution brings to light the very unusual practice of later re-issuing of the printed matter for individual numbers in this exsiccata, which may, as in the case of Hieracium suecicum, contain important nomenclatural novelties.Such a practice, if not properly explained and bibliographically deciphered, may mislead taxonomists and recorders to incorrect nomenclatural interpretations.

Table 2 .
Diagnostic characters for Pilosella stipitiflora, P. suecica and P. colliniformis, the main variants of hybrids between P. caespitosa and P. lactucella.

Table 3 .
Nomenclature of main hybrids between Pilosella caespitosa and P. lactucella in major taxonomic works (as appeared in the cited works but with plant name authorship corrected).

Table 4 .
[35]nostic characters of Hieracium suecicum and related species as used by Fries[35], with the author's emphasis in Italics.