44 MALVACE^. 



monly articulated above the middle, or just beneath the flower. In many 

 cases each flower is subtended by an involucel of three or several, or rarely 

 only one or two bractlets, forming what is usually denominated an exterior 

 calyx ; the importance of which has beea over-estimated in the systematic 

 arrangement of the order. 



The calyx and the corolla are almost without exception pentamerous. 

 The former is herbaceous and persistent, and the sepals, which are strictly 

 valvate in the bud, are more or less united towards the base. The petals 

 are commonly more or less oblique or inequilateral, as is usually the 

 case when their aestivation is convolute. Their insertion is hypogynous; 

 but their short claws are connate with the base of the stamineal column, 

 which union also gives to the corolla the appearance of being slightly gamo- 

 petalous. 



The explanation of this union is given by the investigations of M. Du- 

 chatre upon the organogeny of the flower in Malvaceae.* He has shown 

 that the petals and stamens (at least those which ordinarily appear in 

 Malvaceae) are identical in origin, both being developed from five original 

 papilte alternate with the calyx-segments and next within them, therefore 

 morphologically representing the corolla. These, by parallel and collater- 

 al deduplication, give rise each to a petal and a cluster of stamens ; and the 

 union of these five clusters constitutes the stamineal column. This view is 

 beautifully exemplified by the genus Sidalcea (Plate 120) , recently proposed 

 by myself,! i" vs'hich the column is not resolved into simple filaments, but 

 bears five petaloid lobes or phalanges of stamens, situated opposite the petals, 

 into the base of which a vascular communication may be traced. That the 

 anthers of each lobe are the result of the collateral deduplication of a single 

 organ is evident on inspection of those cases in which the phalanges are 

 two-cleft, and their divisions again two-forked, &c., until we reach the sep- 

 arate anthers; as in Plate 120, Fig. 9. Such stamens, perfectly resolved 

 down to the column, compose the andrcBcium of Modiola (Plate 128), in 

 which the five component phalanges are more or less discernible, of Nap^a 

 (Plate 119), &c. The same, further multiplied by transverse deduplication 

 so as to form several series usually becoming free at more or less unequal 

 heights, constitute, perhaps, the entire androecium of most other Malveae. 

 But what has become of the true stamineal verticil, the parts of which 

 should alternate with the petals'? M. Duchatre has detected this in the five 

 lobes or teeth which terminate the naked apex of the column in such Mal- 

 vacete as Pavonia, Hibiscus, Malvaviscus (Plate 131, Fig. 7), &c., and 

 which, when the column is short, may be seen to alternate with the petals. 

 This, again, is confirmed by Sidalcea, in which the column, prolonged above 

 the sympetalous phalanges, terminates in antheriferous filaments, or in pha- 

 langes the principal, or five exterior, lobes of which apparently alternate 

 with the phalanges of the outer column, and therefore with the petals them- 

 selves. 



* In Jlnnales des Sciences JVaturelles, 3™^ ser. 4. p. 123. 

 i Plantar Fcndhrianm, p. IS. 



