BANTINCULACEJ^. 65 



Of the genera admitted by Jussieu, Salisbury separated X/wicl- 

 tonia (p. 47), Eranfhis (p. 15), and Copt is (p. 17) as distinct generic 

 types. So the authors of the " Flora Altaica' established the genera 

 CaUianthemum (p. 4S) and Ojcygraj)]iis (p. 37). De Candolle had 

 already referred Atragene to Clematis and Cimicifiif/a to Acfa;a ; but 

 he admitted as distinct genera the families Tetractis of Sprengel, He- 

 pafica, and probably too, the Enemion of Raeinesque (p. 19). Hence 

 the "Prodronms' enumerates twenty-eight genera oi RanunmlacecB. 

 Siebold & ZuccARiNi added the two Japanese genera, Anemonopsis 

 (p. 22), and Glaucidiiim (p. 23) ; Hooker & Thompson inserted the 

 Indian CalatJiodes (p. 21) ; and to Nuttall is due the genus Crosso- 

 soma (p. 62), whose claims to a place in Banimculacece are somewhat 

 doubtful. In fine, as many as sixty distinct genera have been 

 admitted into this order; we have reduced their number to nineteen.' 

 Thus established by so many labours following one after another 

 for the last century, this family of plants is one of those which B. de 

 Mirbel so happily termed ''families par enchainemenf" The genera 

 follow one another, and that closely; but they are not closely 

 grouped round a common centre. Accordingly the variability of the 

 characters has allowed various authors to establish secondary divi- 

 sions in the group. In the first place Adanson" distinguishes two 

 sections ; the first with many-seeded, the second with one-seeded 

 ca23sules. A. L. de Jussieu^ established four sections ; his first 

 answers to Adanson's first; Adanson's second section is divided 

 into two, according as the petals (not the corolla) are regular or 

 irregular ; the Actece with a single polyspermous carpel constitute 

 the fourth section. De Candolle^ divides the Ranunculacese into 

 five tribes ; the first {Clematidecp) is marked chiefiy by its valvate 

 induplicate aestivation and opposite leaves ; the second {Anemonecs) is 



organization (seep. 45, note 1), and that organs hesitate to regard as entirely identical what have 



elsewhere distinct are here often assimilated, been regarded by every one since Linnieus as dis- 



degenerated, and passing easily into one an- tinct generic terms ; but in the present state of 



other. confusion of our science it seems to us advan- 



' It must be borne in mind that we do not tageous to diminish the number of generic 



pretend to impose either these genera that we divisions as far as possible, 



admit, or the limits we assign for them, as ^ Op. cil., 457, 45'J. 



absolute. In short when once the common ^ Oi). cit. ; \,Capsul(e monosperma; nondehis- 



and the differential characteristics of two ppntes ; 2, Capsulce polj/spcniiiv—Pftdla irre- 



groups of plants are known, it matters but (jularia ; 3, Capsulm poli/npermo' — Frlala rtffu- 



little whether we separate them as distinct laria ; 4, Germen uiiicinn. Bacca unilocuUinx 



genera, or unite them as sections f>f a single poli/spermn. 



family. Here custom is all-powerful. People * ^i/st., 1127 ; P<W/-. (1824), i. 2-6fi. 



VOL. I. 1' 



