321 



EEVISION OF THE SPECIFIC FORMS OF THE 

 GENUS GYPSOPHILA. 



By Frederic N. Williams, F.L.S. 



The genus Gijpsophila was founded by Linnsus ; and in the 

 first edition of the * Species Plantarum ' nine species, ascribed by 

 him to that genus, are enumerated. The best classification of the 

 species which occur within a definite geographical area is that 

 proposed by Boissier in the ' Flora Orientalis,' vol. i., p. 534. 

 While he emphasizes the fact of the dissimilarity of the species, he 

 admits the absence of distinct characters satisfactorily separating 

 this genus from Saponaria. In order to facilitate the delimitation 

 of the species referable to the genus, and the more effectually to 

 circumscribe its characters in conformity with a more natural type, 

 it is proposed here to exclude those forms in which the calyx is not 

 distinctly 5-nerved with membranous interspaces between the 

 nerves, or in which the unguis of the petal is either bilamellate or 

 appendiculate, or in which the leaves are acerose or subulate- 

 spinescent. This restriction excludes three of Boissier's eight 

 sections — Bolanthus, Psendacanthophyllum, and Jordania. Portions 

 of his classification, depending as they do on the placental attach- 

 ments and the condition of the embryonic radicle, though very 

 scientific, are scarcely practicable in the examination of large 

 collections of specimens. The first-named section might return to 

 Saponaria, as in DeCandolle's 'Prodromus,' and the other two are 

 better placed in Acanthophyllum, in which genus Jordania is included 

 in the ' Grenera Plantarum.' What is to be done with G. orter/ioides, 

 which forms Boissier's section Phryna, I do not know ; but since 

 its Greek name suggests an outcast, it might become the type of a 

 new genus, if it cannot be included in Tunica. It is also proposed 

 to include the genus Ankyropetalum, founded by Fenzl in 1843, as a 

 subgenus of Gypsoyliila. The species referable to it have a cylin- 

 drical calyx, and anchor^form petals. I confess I have not been 

 able to ascertain what " anchorffiform " petals are, but presume they 

 are three-lobed, like those of G. Jiispida : unfortunately all the 

 specimens both in the British Museum and the Kew herbaria are 

 quite destitute of flowers. It has also been suggested to make the 

 generic description sufficiently elastic to include Bunge's genus 

 Allochrusa, but, if it is not to be maintained, it would be best to 

 sink it in AcantJiophyllum . 



The species as here delimitated are grouped in three subgenera, 

 of which the first consists only of the Saponaria porrigens of Lin- 

 naeus, the second coincides with Fenzl's Ankyropetalum, and the 

 third contains seventy-one species. 



In the following enumeration, it should be noted that where the 

 plant is described as a species in Boissier's * Flora Orientalis,' 

 vol. 1., pp. 534-557, this citation is given immediately after the 

 specific name ; in all other cases the reference to the original 

 description is given. Where a second citation is given, it refers to 

 a good published figure of the species. Only the principal synonyms 



Journal of Botany. — Vol. 27. [Nov. 1889.] y 



