13U 



NATURAL mSTOItT OF PLANTS. 



(ti«;s. 1().'), 100, lO.s, IGU), and iWr///' (figs. 107, 171) cultivated 

 every where, oi' wliicli the former has persistent leaves, the latter leaves 

 wliieh fall every year, and magnificent flowers produced at the end 



of the winter before the leaves. On 

 examinincT these flowers we first notice 

 that the axis or receptacle has the form 

 of a cylindro-conoidal branch, bearing 

 successively from below upwards a pe- 

 rianth, a large number of stamens, and 

 carpels inserted on a spiral.' In the flower 

 of M.grandijlora the perianth first presents 

 three more or less greenish^ free leaves, 

 imbricated in the bud (fig. 106), so that 

 most usually one is quite outside and one 

 quite inside, wliile the third is overlapped 

 by the former on one side, and overlaps 

 the latter on the other. These leaves, 

 usually described as sepals, fall early. Within are two corolline 

 wliorls, the one consisting of three petals alternate with the 

 sepals, the other of three interior to these and alternate with 



Magnolia grandijlora. 

 Fig. 166. 

 Diagram. 



* Desf., Arhres, ii. 6. — DC, Prodr., n. 10 

 (sect. Gwillimia Rottl.). — M.consj)iev.aSALlSB., 

 Par. Lond., t. 38. — Yulania conspieua Space, 

 Siiif. a Bvffon, vii. 464. 



- The fraction of the phyllotaxy of the Magna- 

 liacere is usually ^. Accordingly, in the arrange- 

 ment of the rtoral apjjcmdages we meet with the 

 fractions derived from this up to ~ and ^. 



■' Their colour varies with the individual leaf 

 and with the age of the flower. Very often the 

 sepals when adult are as white as the petals, or 

 nearly so. When young they arc usually of a de- 

 licate green. These facts show how characters of 

 colorati<jn and consistency may he sometimes un- 

 reliable and insufficient to distinguish a calyx from 

 n corolla. It would no doubt be more correct to say 

 tliat in M. gmmlijlora L., the perianth is triple, 

 and that the leaves of the two inner whorls are 

 usually more peialoid when adult than those of the 

 outer one. Inother species the difference of colora- 

 tion between the se])als and petals is no longer 

 appreri;djle when adult. Thus, in many plants of 

 M. Yi tail Deof. and Soulangiana (hybrid), all the 

 leaves of the perianth are so similar that one may 

 well pay that these flowers possess a triple corolla 

 and no calyx. It is sometimes the same with the 

 nine yellowish green primrose leaves of the peri- 

 anth of M. acttminata L. In M. glanca L. they 



are sometimes all white and similar, sometimes the 

 two or three outermost leaves are green. M. macro- 

 ■pliylla MiCHX.has usually three green or greenish 

 sepals and six white petals. In the flowers of M. 

 purjmrea Cuet., we almost invariably find a 

 great difterence between the six petals, which are 

 broad, erect, and wine-red on the outside, and the 

 three sepals, which are small, and are early re- 

 flexed on the peduncle and become brownish. 

 The total number of pieces in the perianth is also 

 very variable in cultivated species ; as many as 

 twenty andupwards mny sometimesbe counted, as if 

 the flower showed signs of becoming double. We 

 have demonstrated [Adansonia, vii. 3) that these 

 variations have no real importance ; that in cer- 

 tain species those very leaves have been called 

 sepals, which in others have been named petals ; 

 while the sepals of authors are often only bracts 

 preceding the flower, representing the sheaths or 

 dilated petioles of leaves, and continuing the 

 spiral series of the leaves properly so-called. The 

 sepals too are inserted along the same spiral. 

 Hence it is that, as shown in fig. 166, there is 

 no sepal exactly opposite or alternate with the 

 bract which immediately surrounds the flower. 

 Moreover, the nature of the pieces of the perianth 

 is shown by these facts ; they are leaves reduced 

 to the basilar portion. 



