Kalanchoe marmorata var. somaliensis (Baker) Gideon F.Sm. (Fig. 3)

Basionym:— Kalanchoe somaliensis Baker (1895: 214). Also treated in: Hooker (1902: Tab. 7831); Berger (1930: 405); Jacobsen (1977: 289); Jacobsen (1986: 629, Fig. 887); Sajeva & Costanzo (1994: 153).

Type:— Somali Land [SOMALIA]. Wardie, 22 February [presumably] 1895, received [presumably at Herb. K] in “5/95” [May 1895], [Miss] Edith Cole s.n. (lectotype K, accession number H608/68 41, barcode K000232777, https://apps.kew.org/herbcat/getImage.do?imageBarcode=K000232777) designated by Smith (2021a: 98).

Homotypic synonyms:— Kalanchoe marmorata f. somaliensis (Baker) Pampanini (1909: 53). Also treated in: Pampanini (1911: 405). Kalanchoe macrantha Baker ex Maire (1976: 254) var. somaliensis (Baker) Maire (1976: 255).

Nomenclatural notes on K. marmorata f. somaliensis:—Although Pampanini (1909: 53, 1911: 405), as well as, later, Jacobsen (1977: 289) and Jacobsen (1986: 629), give the author citation of the basionym of the forma he recognised as “Hook. f.”, K. somaliensis was originally described by Baker (1895: 214).

The locality from where material that prompted this change in rank was collected was given as “Hab.: « Colonia Eritrea. Altipiano del Soira (Scimenzana). Località Golò, 2720–2740 m. sul mare: ripiano alluvionale derivante dal disfacimento dell’arenaria. 4 dicembre 1905 [Dainelli e Marinelli] ».—(Herb. Centr. Florent. [Centr. extra-ital.]).” [English: “Colony of Eritrea. Soira Plateau (Scimenzana). Location Golò, 2720–2740 m above sea level: alluvial shelf resulting from the decomposition of sandstone. 4 December 1905 [Danielli & Marinelli s.n.] (Herb. FI)”]. At the time this was regarded as the first and only record of K. marmorata (as K. somaliensis, here regarded as K. marmorata var. somaliensis) known from beyond the borders of Somalia, the country from which Baker (1895: 214) described K. somaliensis, from the Golis range near Wardie (Pampanini 1909: 54).

Note that the specimen Danielli & Marinelli s.n. is not the type of K. marmorata f. somaliensis; it is merely a specimen that vouches for the occurrence of the forma in Eritrea.

Taxonomic and nomenclatural notes on K. marmorata var. somaliensis:—When Baker (1895: 214) described K. somaliensis he gave the colour of the leaves as “brunneo-marmoratis” [brownish marbled] (see also the illustration published by Hooker 1902), as opposed to the leaves of K. marmorata that Baker (1892: 300) earlier described as “maculis copiosis fuscis marmoratis” [copiously brown marbled spotted]. Therefore, material referable to K. marmorata var. somaliensis has leaves that are virtually throughout devoid of the characteristic purplish brown mottling found on one or both surfaces of the leaves of K. marmorata var. marmorata. The leaves of the autonymic variety are usually consistently sparsely to densely spotted with purplish brown marks and in some material can be lightly spotted only (see for example the discussion in Bruce 1948: t. 1049).

In his key, under “ BaαI2 ”, Berger (1930: 405 and Figure 196 D–G on p. 403) additionally recorded K. somaliensis as being characterised by: “Röhre der Blumenkrone im oberen Teile zylindrisch. Blätter grosser, seichter buchtig gezähnt, fast weiss, einiger gefleckt”. [English: “Corolla tube cylindrical higher up. Leaves large, shallowly indented, almost white, somewhat maculate”]. The contrasting statement, “ BaαI1 ”, in the key provided by Berger (1930: 405) noted: “Röhre der Blumenkrone 4kantig” for K. marmorata. However, variation in the outline of cross-sections of the corolla tubes of K. marmorata has been observed and this character is here not regarded as taxonomically useful.

The leaves of K. marmorata var. somaliensis are often basally more distinctly amplexicaul and almost auriculate (see for example Jacobsen 1977: 289, 1986: 629) and not as cuneate as those of K. marmorata var. marmorata. Additionally, leaves of K. marmorata var. somaliensis are often larger than those of K. marmorata var. marmorata. Taxonomic recognition of material of K. marmorata with often larger, virtually immaculate, basally stem-clasping leaves seems desirable, as use of the name K. somaliensis, at the rank of species therefore, has persisted, for example in Jacobsen (1977: 289), Jacobsen (1986: 629, Fig. 887), and Sajeva & Costanzo (1994: 153) for material lacking the distinct purplish maculation. However, at the rank of variety a combination has only been made under K. macrantha Baker ex Maire, nom. illeg. Further, the nomenclatural novelty K. marmorata var. somaliensis is not a new combination as K. marmorata f. somaliensis has been published previously (Turland et al. 2018: Art. 6.10 and its Example 13).