73 FLORA INDICA. [^MagnoUacea. 



somewhat obovate, sharply denticulate, 10 inches long by 5 broad, with a petiole an 

 inch long. They are slightly silky above, but probably become glabrous with age, and 

 are pubescent below, especially on the nerves. The flowers, which unfortunately are 

 in a very imperfect state, seem the same as in D. jientagyna. l^or the present it ap- 

 pears sufficient to call attention to this plant ; we really know nothing of the varia- 

 tions and mode of growth of these trees, and to found species on single specimens, 

 especially where the flowers and leaves are detached, as is almost always the case in 

 this genus, would lead to irremediable confusion. 



III. MAGNOLIACEiE. 



Flares hermaphroditi, rarissime uuisexuales. Sepala et pelala hypo- 

 gyna, ternatim (rarissime quinatim) phiriserialia, asstivatione imbricata, 

 decidua. Hlainiiia indefinita, circa torum cylindricum inserta, libera; 

 anthera basifixa?, loculis lineaiibus lateraliter vel introrse dehisceiitibus. 

 Ovaria plurima (rarissime pauoa aut solitaria), discreta sive lateraliter 

 inter se cohcerentia, uniserialiter verticillata vel saepius supra torum 

 elongatum spioatim disposita, unilocularia. Ovida in sutura veutrali 3 

 vel plura, rarius e basi adscendentia solitaria, anatropa. Embryo in 

 basi albuminis copiosi oleosi non ruminati, minutus, hilo proximns. — 

 Arboi-es vel frutioes sape aromatica ; foliis alternu simplicidus integer- 

 rimis, stipulis lateralibus petlolo adnatis cito deciduis, rarius nullis. 



In this family the petals are always imbricated in more than one row, or, in other 

 words, the perianth always consists of more than two series. The sepals are often 

 identical in texture and appearance with the petals, but sometimes they are readily 

 distinguishable from them, and they are then usually three in number. In the Indian 

 species this ternary arrangement occurs more or less distinctly in all the species 

 which we have had an opportunity of e.xamining ; but other authors descrilie the 

 perianth of some species as pentamerous. 



We follow the usual course in including Winterees as a tribe of Magnoliacecc. The 

 absence of stipules, however, is so very marked a character, in an Order in which 

 these organs are so constantly and conspicuously present, that it may be questioned 

 whether it would not be more advisable to separate them. Tliis is, however, a mat- 

 ter of little consequence, till the systematic value of natural groups is better esta- 

 blished, as their position would in any case remain the same, then' affinity being much 

 greater with Magnolia and its allies than with any other group. 



The stipulation of IlagnoUnceip is very peculiar. In the leaf-bud each scale is 

 composed of a pair of stipules at first united throughout their whole length, but lat- 

 tcrlyjmore or less split. From the dorsum of the scale, at a distance below the apex, 

 which varies in each species, rises the rudimentary leaf, which is longitudinally folded 

 inwards in vernation. In the outermost scale of the bud the foliaceous portion of the 

 leaf is usually very small, and falls away at a very early period, leaving a distinct 

 cicatrix at the top of the very evident petiole, along which the two stipules, which 

 are united to form the scale, are adherent. After the development of the branch, the 

 stipules remain at first adherent to the petiole on each side, but very soon wither and 

 fall off, leaving an elongated cicatrix on the petiole, which varies in proportional 

 lenglli to the petiole on the different species. 



In the flower-bud the spathes are exactly analogous to the scales of the leaf-bud ; 

 but the tendency to development is in reverse order: the innermost, which is ad- 

 pressed to the flower, rarely shows any tendency to leaf-development, but splits to 

 the base before falling off; wliile in the outer spathes the petiole is generally distinct, 

 witli a scar at its apex marking the spot from wliich the rudimentary leaf has fallen 



