In the early 1800's, the following three species of Campanula L. were described: C. asperrima Zuccagni, C. decurrens Zuccagni and C. ajugifolia Schult. They all referred to plants with spiny-dentate leaves which are more or less tomentose underneath and that possess sessile axiliary flowers. They were considered by all later authors as insufficiently known and none of the names had been formally typified. All have connections with Attilio Zuccagni (1754–1807), Domenico Sestini (1750–1832) and the city of Florence. From available evidence, these three names of species might all be Turkish plants, but only one can be linked to an authenticated specimen and a fairly definite provenance. They have also been considered to be related to the Mediterranean C. alata Desf. and C. peregrina L. (sect. Pterophyllum Damboldt) which we are investigating, hence our re-assessment of their status.

The fate of the specimens collected by Domenico Sestini

No specimens have been traced for the two Zuccagni names. Attilio Zuccagni was director of the Museo di Fisico e Storia Naturale at Florence and his herbarium (also the specimens of Domenico Sestini) was deposited there (FI). However, when P. Parlatore came to Florence as director in 1842 he found that these specimens were inadequately labelled and so badly destroyed by insects that he threw them away (Parlatore, 1874: 6). Recent searching at FI (Chiari Nepi, pers. comm.) failed to find any remnants of these collections. The only clue to the identification of the Campanula species is, therefore, a combination of their original descriptions, and their collector and provenance. Both were described “ex herb. orient. D. Sestini innominata“ without any other indication of provenance; we have been unable to trace any archival material which might shed light on what the Sestini herbarium might have contained.

The travels of Domenico Sestini

Abbé Sestini, born in Florence, was clearly a polymath who travelled extensively in different parts of southern Europe and western Asia. Among his varied interests (Fruttuoso, 1837) he made observations on and collected plants. How many plants he actually collected is unknown, but both Boissier (1867), though not directly citing any of his specimens, and Grisebach (1843: ix) were aware of him as a collector. Grisebach noted that many of his plants were sent to Willdenow in Berlin. The genus Sestinia Boiss. (Lamiaceae) (a synonym of Hymenocrater Fisch. & C. A. Mey.) was named for him, as was Sestinia Boiss. & Hohen. (Rubiaceae) [nom. inval.] (an invalid synonym of Wendlandia DC. [nom. conserv.]). Sestini (1786) deals with his travels from Constantinople (Istanbul) to Basra, another travelogue (Sestini, 1794) from Constantinople to Bucharest via European Turkey. Sestini (1794: 118–121) listed some plants and their uses in Galatia, central Turkey, but there is no mention of Campanula. We came to the conclusion that the Campanula plants must have been collected during the previous trip. In fact, Sprengel (1827: 378) has in the entry for Sestini: “invenerat plantas, cum Willdenovio communicavit, qui, quod Galacz (oppidum ad Danubium) ascriptum erat, Galatiam interpretatus est, falsam patriam”. Time must have elapsed before Sestini knew of Willdenow's identifications of his plants and the list of the names could only be included in the second book.

A comment here on the name “Galazia” is relevant because Sprengel (1827: 378) referred to Sestini's Galazia as Galacz (today Galati) on the Danube, in Romania. However, in his check-list Sestini gave some local names in Turkish, referred to some plants used to feed Angora goats, mentioned Elma Dagh (a mountain near Ankara) and Kaiserie (Kayseri). He also referred to Tournefort's earlier travels there. This leaves no doubt that he was writing about Turkey and not Romania. The reason for Sprengel's confusion may have been the name “Valachia” in the title of Sestini' second book. At the time of this journey, Sestini was secretary to the Prince of Valachia, the area of S-SE Romania known today as Walachia. Griesebach (1843) also referred to “Galatiae“ in Turkey.

Sestini also spent time in Sicily, the Aegean (Khalki), Cyprus, Thrace and in what today is Iraq. Therefore, the two Campanulas that Zuccagni described “ex herb. orient. Sestini“ could have come from any of these areas, but with a greater likelihood of Turkish/Iraq provenance, because of the word “orient.”.

What is the application of the names Campanula asperrima Zuccagni and Campanula decurrens Zuccagni?

What clues can be found in their brief diagnoses and descriptions? Very few, especially in such a large and polymorphic genus as Campanula, but presumably their epithets were chosen for what their author considered salient characteristics such as the scabrid, rough-to-touch indumentum of C. asperrima. Candolle (1830: 341–342) in his monograph of the family, and Candolle (1839: 485), included this species in his “species incertae sedis“ exactly repeating Zuccagni's original protologue, and giving C. ajugifolia Schult. (with the wrong authority “Sestini in Spreng. Syst. 1. p. 732”) as a synonym under C. asperrima. This was also followed by Vatke (1874). Our assessment of C. asperrima, however, is that it is impossible, based on the existing evidence, to establish what the species is.

Candolle (1830: 341–342, 1839: 485) also listed C. decurrens Zuccagni under “species incertae sedis“. Presumably he did not see any specimens. In both publications he noted “An C. alatae Desf. affinis/proxima?”. Candolle did not notice the existence of the earlier valid C. decurrens L. of 1753, as this was rectified by Steudel (1840: 269) with the new and valid name C. orientalis Steud. ( C. decurrens Zuccagni non L.). Boissier (1875) also dealt with C. decurrens, and C. asperrima, under “species ex descriptione incomplete non recognoscendae“, but, in contrast to de Candolle, he considered C. decurrens as possibly related to C. peregrina.

Feer (1890) brought yet another name into this problem. This was C. argaea Boiss. & Balansa (Boissier, 1859: 119), which he considered a synonym of C. decurrens. He found Zuccagni's description good and considered that the indumentum, decurrent leaves and glomerulate flowers on a short peduncle were enough to characterise it. However, one feature in the original description of C. decurrens is quite out of character with C. argaea: “folia … subtus reticulata et nonnihil tomentosa“. In C. argaea, the leaves are always densely tomentose- hispid. In recent times, Greuter & al. (1984) cited “? = C. decurrens Zuccagni” in the synonymy of C. peregrina. However, we found no evidence confirming the identity of C. decurrens, as for C. asperrima.

The application of the name Campanula ajugifolia Schult.

Campanula ajugifolia Schult. is another name that has lurked in the shadows. It was not dealt with by Damboldt (1979), despite it having been described from Galatia (Central Turkey). During the present investigation, we studied a specimen of it in the Willdenow herbarium (B-W 3833). Added to the sheet one finds a label with a handwritten description and the mention “Hab. in Galatia“. On the sheetWilldenow added the original collector of the specimen, “Sestini”. Vatke (1874) also cited Willdenow 3833 as an original specimen of C. ajugifolia. In the absence of any other known specimen or any illustration, this is the obvious choice of lectotype for the name.

A later handwritten annotation on this sheet, signed by Henri Feer in 1891 is: “C. decurrens 1806, C. asperrima 1806, C. ajugaefolia 1819, C. orientalis Steud. 1840, C. argaea 1859“. The assumption is that Feer considered them all to be the same, but this was never published as he died the following year (in 1892), though he had already synonymised C. argaea and C. decurrens in 1890. Feer worked from G, but there are no specimens there of the Zuccagni names or Sestini collected specimens (F. Jacquemoud, pers. comm.).

We agree with Feer that C. ajugifolia is identical to and the earlier correct name for C. argaea. Campanula ajugifolia is a clearly distinct species restricted to several provinces in SW and Central Turkey. To date, we have been unable to ascertain where Sestini collected the specimen; maybe his exact route and the type locality of C. ajugifolia will never be known, unless relevant correspondence comes to light in Florence or among Willdenow's archives.

Damboldt placed C. argaea in sect. Spicatae (Fomin) Damboldt and considered it to be without close relatives. The species has, however, some features in common with the commoner and more widespread C. peregrina (sect. Ptetrophyllum Damboldt) especially the very similar facies with basally winged/decurrent hispid leaves and spicate infloresences; but in some other features it is clearly different, much smaller flowers and a capsule opening by basal, not median, pores.

Conclusions

Fig. 1.

—Lectotype of Campanula ajugifolia Schult. [specimen in the right side, with the arrow].

[Sestini, s.n., B] [© Botanischer Garten und Botanisches Museum Berlin-Dahlem. Reproduced with permission]