82 flora indica. [Schizandracece. 



scd nanquam foliifera. Stamina gyuoccium sequantia, conncctivo apice longe subu- 

 lato. Carpella in spicam oblongam 3-4-pollicarem laxe congesta, approximata, ^« 

 pollicaria, compressa, rotundata. Semiua 1-2. 



8. M. Nilagirica (Zenker, Plant. Ind. t. 20) ; foliis elliptieis utrin- 

 que acutis vel ovalibus obtuse acumiiiatis utrinque glabris vel subtus 

 secus costain pubescentibus, floribus albis, petalis cum sepalis 12 ex- 

 terioribus obovatis interioribus oblongo-lanceolatis acutis. — Wight! III. 



Iqh. t. 6. M. Pulneyensis, Wight! 1U.L 



M 



l. 14, Icon. t. 938 ! Spic. Neilgh. t. 6. M. Pulneyensis, 

 14. t. 5, excl.f. 5 et 6. M. Champaca, Wall. Cat. 969 



lit.) M. ovalifolia, Wight! III. i. 13. 



j8. Walkeri; arbuscula, foliis oblongis vel lanceola 

 subtus glaucescentibus 2-3 poll, longis, floribus minoribus. 



keri et M. glauca, Wight, JUL i. 13. 



Hab. In montibus altioribus peninsulse australis, alt. 6-8000 ped., 

 Wight! et in summis montibus Zeylanise, Walker! etc. — (v. s.) 



Arbor magnitudine varia, plerunique exeelsa, in Zeylania interdum fnrtico3ft; 

 partes uovellse sericeo-villosse. Folia forma valde varia, 3-5 pollices longa, 1^-2 

 lata, petiolo -f-poll. Cicatrix stipularis dimidium petiolum sequans. Alabadri 

 l-]|-pollicares, cum pedunculo longitudine vario dense fusco-sericei (in cinereo- 

 sericei). Spatha 2. Carpella in spicam 2-3-pollicarem disposita, subcompressa, ro- 

 tundata. Semina pleruinque solitaria. 



Our variety £ (from Ceylon) has at first sight so very different an aspect from the 

 peninsular plant, that we can scarcely persuade ourselves that it is not distinct. We 

 have, however, failed to discover satisfactory characters to distinguish these plants m 

 the dried state; but botanists who have an opportunity of observing the living plant 

 may perhaps be more successful. The Ceylon plant, of which we have seen a 

 rather extensive suite of specimens, varies much in the size of the flower and in the 

 shape of the leaves ; and the small lanceolate-leaved states appear to pass insensibly 

 into a plant with oval leaves, which, though usually more coriaceous, are sometimes 

 quite ^distinguishable from those of the typical M. Nilagirica. These small states, 

 which have sometimes nine instead of twelve petals, seem in many of our specimens 

 to be diseased, the flowers being unusually small, the stamens few and abbreviated, 

 and the young carpels abnormally swollen, as if punctured by an insect, and appa- 

 rently abortive. Perhaps, therefore, it will be found that the broad-leaved arbores- 

 cent state is the normal form in Ceylon as well as in the peninsula and that the 

 lanceolate-leaved state is an accidental variety. M. glauca of Wight is certainly 

 only an abnormal form, with broadly obovate leaves, for the glaucous hue ot tne 

 under surface is not confined to specimens with that form of leaf, but is seen equally 

 in the oval and lanceolate-leaved plants, and is often observed on the same specimen 

 with leaves not at all glaucous below. 



IV. SCHIZANDRACE^E. 



Flora unisexuales. Sepala et petala hypogyna ternatim vel qunm- 

 tim pluriserialia, seativatione imbricata. Stamina debute vel mdenmta, 

 toro depresso vel conico inserta. Fllamenta libera vel plus minus 

 coalita. Anthem adnata biloculares, plerumque vane heteromorplue. 

 Ovaria iudefinita, in capitulum oblongum vel subglobosum coalita. 

 Ovula in sutura ventrali 2-3, ampliitropa vel fere campvlotropa. Baa* 

 dissepimento spurio transverse bi- (rarius in-) locularea, (hapena*. 



