Solanum mammosum L.

Solanum mammosum is widely cultivated for its unusual fruits that are usually teat-shaped with an apical nipple and basal lobes. It is a member of the Acanthophora clade (Levin et al. 2005), of which three other species are also found escaped in Africa (S. aculeatissimum, S. capsicoides, S. viarum). Solanum mammosum is probably native to the Caribbean and Central and South America (Nee 1979), but is now widespread worldwide in cultivation. Solanum mammosum can be distinguished from other species in section Acanthophora by the combination of its purple corollas, long copious pubescence, and its usually bizarre nipple-shaped fruit. It is most similar to S.palinacanthum, which also has purple corollas, but which has short glandular puberulence and exclusively globose fruits.

The morphological and geographical reasons for considering this species native to northern South America and possibly the Caribbean have been discussed previously (Nee, 1979). Briefly, the globose fruit form is undoubtedly the primitive condition and is nearly confined to the Llanos of Venezuela and adjacent areas where the species seems well-adapted to the seasonally dry grassland climate. It is only rare and sporadic in Brazil south of the Amazon basin, the center of diversity of the section. Because of the widely acknowledged utility of its poisonous fruits in killing rats and cockroaches, it has been to a minor extent cultivated and spread.

  The specific epithet for the synonym S. platanifolium is often attributed to William Jackson Hooker (e.g., see Nee 1979, 1999; Edmonds 2012), as he was responsible for the descriptions of plants featured in Curtis’s Botanical Magazine for most of the 19th century. John Sims, however, was the author of all descriptions from volume 15 (1801) to part way through volume 53 (1826); Hooker only wrote the descriptions from plate 2684 (erroneously given as 2689 in Chittenden 1956) in volume 53 and in their entirety from volume 54 (volume 1 of the new series) onwards. Authorship of descriptions is clearly indicated on the title page of each volume and that of vol. 53 is attributed to Sims alone, with no mention of Hooker, but several descriptions are followed by the initials W.H.; these are to be attributed to Hooker. Solanum platanifolium is not one of these. The lectotype selected for S. globiferum is the herbarium specimen made from plants cultivated in Montpellier. A single fruit in the MPU collection collected by Gillet may have been the source of seeds (Nee 1979); it is syntype material as the fruit was also cited by Dunal in the protologue.

  Solanum corniculatum and S. cornigerum are likely to be homotypic; both were described in the horticultural literature from material of unknown provenance (but thought to be from Africa by Huber, see André 1868. No specimens were cited in the protologue of S. cornigerum, and reference was made to “corniculatum, propose par MM. Huber” a name that was ‘incorrect’ (“n’est pas latin”) in Latin (André 1868), hence the need for a new name, S. cornigerum; this oblique reference may make S. cornigerum illegitimate and superfluous. A plate referred to in the protologue of S. cornigerum (see below) is perhaps that distributed by the Huber brothers on request (see Mabberley 1985) and thus related also to S. corniculatum. André’s (1868: 33) mention of an ‘incorrect’ name on the plate supports this – “[l]a figure représentée ci-dessus étant imprimee lorsque nous avons fait cet article, nous n’avons pu changer le nom qui était en bas, que nous n’adoptons plus.” Several sheets said to have originated from the Huber brothers seed catalogues have been found in BM; one of these (BM000942835) was grown in 1868, has the 5-lobed fruits of André’s original description and is probably derived from the same seed source. Edmonds (2012) incorrectly cited this as type material. We have lectotypified S. cornigerum using the plate from Revue Horticole, as this is the only unambiguous material available. These plates are not numbered and in the library at BM, for example, they are disassociated from the bound volumes and kept in the plate collection. At Kew, the plates are bound with the printed volumes opposite the page where the name appears.