11 



ticulation with the stem. Thirdly, the flowers of Nymphaeaceaj have eo great 

 an analogy generally with Dicotyledons, and particularly with that of Magno- 

 hace«, and their fruit with Papaveracere, that it is difficult to doubt their belong- 

 ing to the same class. ^Fourthly, the reasons which have been offered for con 

 sidering the embryo monocotjdedonous, however plausible they may have 

 appeared while we were unacquainted with the true structure of the ovu- 

 lum of plants, have no longer the importance that they were formerly supposed 

 to possess. The sac, to which I.Jaave already alluded, to which so much unne- 

 cessary value has been attached, and which was mistaken for a cotyledon by 

 Richard, is no doubt analogous to the sac of Saururus and Piper, and is nothinr 

 more than the remains of the innermost of the membranous coats of the ovulum, 

 usually indeed absorbed, but in this and similar cases remaining and covering 

 over the embryo. Mr. Brown {Appendix to King's Voyage) considers it the 

 remains of the membrane of the amnios. M. Decandolle assigns a further 

 reason for considering Nymphseaceee Dicotyledons, that they are lactescent, a 

 property not known in Monocotyledons. But in this he is mistaken ; Limno- 

 charis, a genus belonging to Butomeaa, is lactescent. Independently of the 

 peculiarities to which I have now alluded, this order is remarkable in some 

 other respects. It offers one of the best examples which can be adduced of 

 the gradual passage of petals into stamens, and of sepals into petals : if atten- 

 tively examined, the transition will be found so gradual that many intermediate 

 bodies will be seen to be neither precisely petals nor stamens, but both in part. 

 The development of the disk, which is so remarkable in Nelumboneae, takes 

 place here in various degrees. In some, as in Nuphar, it is merely an hypogy- 

 nous expansion, out of which grow the stamens and petals ; in others, as Nym- 

 phaea, it elevates itself as high as the top of the ovarium, to the surface of 

 which it is adnate, and as the stamens are carried up along with it, we have 

 these organs apparantly proceeding from the surface of the ovarium : in another 

 genus, the Barclaya of Dr. Wallich, the petals are also carried up with the 

 stamens, on the outside of which they even cohere into a tube, so that in this 

 genus we have a singular instance of an inferior calyx and a superior corolla in 

 the same plant. Supposing this order to be exogenous and dicotyledonous, a 

 fact about which there appears to me to be no doubt, its immediate affinity will 

 be with Papaveraceae, with some genera of which it agrees in the very 

 compound nature of the fruit, from the apex of which the sessile stigmas radiate, 

 in the presence of narcotic principles and a milky secretion, and in the great 

 breadth of the placentae. They are also closely akin to Magnoliaceae, with which 

 they agree in the imbricated nature of the petals, sepals, and stamens; to 

 Nelumboneae their close resemblance is evident ; with Ranunculaceae they 

 are connected through the tribe of Preonies, with which they agree in the dilated 

 state of the discus, which, in Pasonia papaveracea and Moutan, frequently rises 

 as high as the top of the ovaria, and in the indefinite number of their hypogynous 

 stamens ; but in Ranunculaceae the placentae only occupy the edge of each of 

 the carpella of which the fruit is made up ; so that in Nigella, in which the 

 carpella cohere in the centre, the seeds are attached to the axis, while in Nym- 

 phaeaceae the placental occupy the whole surface of each side of the individual 

 carpella of which the fruit is composed. But if such are the undoubted imme- 

 diate affinities of Nymph aeaceaj, it is certain that some strong analogies exist 

 between them and Hydrocharideae, to the vicinity of which they are referred 

 by those who believe them to be Monocot_yledonous. Taking Nelumboneae 

 for a transition order, they have some relation to Alismacere, the only monoco- 

 .tyledonous order in which there is an indefinite number of carpella in each 

 flower, and to FtydrocharideaB, with which they agree in the structure, though 

 not the vernation, of their leaves, and their habit. An analogy of a similar 

 nature with this last may be also traced between them and Menyantheae. 



