a new Natural Order. 405 



affinities borders somewhat on divination, — Jussieu himself failed to seize 

 those of Melianthus, which in his 'Genera' ranks as an appendage to his 

 dimorphous order of Rutacece, and is more particularly compared with Dic- 

 tamnus on one side, and Tropceolum (one of his Geraniacece) on the other. 

 More recently, in his excellent monograph of Rutacece, M. Adr. de Jussieu, 

 illustrating by accurate figures the structure of the flowers of Melianthus, refers 

 again to the analogy of that genus with Tropceolum ; while however he places it 

 as a genus affine at the end of the group of Zygophyllece, which he, following 

 his illustrious father, considers as a mere section of Rutacece, but which 

 Mr. Robert Brown had already distinguished as an independent order. To 

 that view of the affinities of Melianthus subsequent authors have either im- 

 plicitly or positively acceded ; with the exception of M. Reichenbach, who, 

 perhaps better inspired by Chance, the deity who must have presided over 

 his vast bird's-eye view of the vegetable kingdom, has revived the idea of 

 Adanson as to the immediate connexion of Melianthus with Sapindacece. 



After these too long but unavoidable details, I proceed to take up the 

 question anew, by a sketch of the striking features of those plants which 

 I consider to be the real types of Melianthus ; that is to say, Melianthus 

 major, L., a well-known Cape species, and M. Himalayanu^, Wall., a truly 

 unexpected member of the flora of Northern India. All these are shrubs, 

 with simple and half-herbaceous stems, large pinnate leaves of a glaucous 

 hue, and long terminal racemes of dull reddish flowers, which spring singly 

 from the axils of coloured bractese. A strong foetid and virous smell of all 

 the parts, the widely-winged petiole, deeply serrated folioles, large intrapetio- 

 lar stipule (made up of the connexion of two), the abundance of the honeyed 

 liquor of the flowers, — all these points are highly characteristic of the genus. 

 The flowers themselves, if seen at and long before the time of their expansion, 

 appear under a deceitful attitude; since by the early torsion of their pedicels, 

 they present downwards, as regards the general direction of the erect raceme, 

 that side which normally and actually in the young bud is turned upwards, or 

 lies close to the rachis. Hence that which by all authors has been described 

 as posticum or superum, must be understood as anticum or inferum, and vice 

 versd. According to this view, each flower exhibits the following structure : 

 — A large coloured calyx, with a depressed and somewhat triangular base, 



