MAGNOLIACEJS. 



145 



In short, the configuration of the androceum is very variable, and 

 presents so many gradual changes in this genus as to render it 

 impossible to found any exact subdivisions on it. 



It is exactly the same with Kadmira,' a genus of plants from the 

 South and East of Asia, which have been separated from Schizandra 

 on account of a single absolute character — the form of the carpels 

 as a whole. They are here united into a ball or short head, while 

 those of Scliizandra proper, form a sort of more or less elongated 

 spike. But we are unwilling for this one reason alone to separate 



Schizandra (^Kadsura) Japonica.^ 



Fig. 185. 

 Male flower. 



Fig. 186. 

 Longritudiiial section of male flower. 



Fig. 187. 

 Androceum. 



Fig. 188. 

 Stamen, isolated. 



Fig. 189. 

 Carpel. 



Fig. 190. 



Longitudinal section of 

 carpel. 



Kadsura from Schizandra, except as a section,^ because in Magnolia 



' K^MPF., ex Juss., in Ann. Mus., xvi. 340. 

 — DuNAL, Mon. Anonac, 57.— DC, Frodr., i. 

 83.— Wall., Flor. Nepal., i. 7.— Bl,, Fl. Jav., 

 Sckizandr., 7, 1. 1, 2. — Sieb. & Zucc, Fl. Japan., 

 I 40, 1. 17.— ENDL.,G'e«., n. 4731.— Benth., FL 

 Honkg., 8.— Hook. & Thoms., Fl. Lid., i. 83.— 

 H. Bn., Adansonia, iii. 43 ; vii. 11, 66. — B. H., 

 Gen., 19, n. 9.— Miq., Fl. Ind.-Bat., \., pars 2, 18. 

 — Walp., Rep., i. 92 ; v. 15 j Ann., iv. 78.— -S'ar- 

 cocarpon KjE.wv'e.,Amoen. Exot., 476, 185, t. 477. 

 — Bl., Bijdraj., 21. Although Linnjjus admitted 

 K. japonica, it was A. L. de Jussieu, who in 

 1810 (Ann. Mus., xvi. 340) considered tlie Uvaria 

 jcqwnlca of TiiUNBEiio worthy of forming a 

 special group. Of tliis plant, he says, " we pro- 

 pose to make a separate genus under the name 



VOL. T. 



of Kadsura." Hence, in uniting Kadsura and 

 Schizandra into one genus, we have had to prefer 

 tlie latter name, which dates from 1803, and 

 which we And possesses the further advantage of 

 not forcing us to suppress the name of Schizan- 

 drecE in order to substitute that of Kadsttrem. 



2 Schizandra japonica H. Bn. (nee A. Ghay). 

 — Kadsura japonica L., Spec, 756. — Drx., 

 Mun. Anonac, 57. — MlQ., Ann. Mus. Lugd. 

 Bat., iii. 91. — DC, Frodr., i. SZ.— Uvaria 

 japonica TuUNB., Flor. Japon., 237. — Fuhi 

 Kadsura, &c., K.WMrF., loc. cif. 



3 See Adansonia, vii. 10. Especially is it im- 

 possible to distinguish Kadsura and Sc/iizandra 

 by the aspect of the anthers, and authors (such 

 aa Enulicueb) are wrong in attributing extrorse 



