12 Mr. B. Clarke on Relative Position; 



variations in its position ; but this is not the case, as the carpels 

 have regularly the ventral suture turned outwards ; and this sug- 

 gests the idea that carpels (like leaves rcvolute in vernation) may 

 be formed by the margins of the carpellary leaves being turned, 

 and meeting outwards instead of inwards. May not stamens 

 also, being turned outwards or inwards, be analogous in some 

 cases to the vernation of leaves ? In Tormentilla officinalis I have 

 observed carpels among the stamens (stamens metamorphosed ?) 

 having the ventral suture appai'ently outwards, the tendency of 

 the anthers being also to open outwards. But as in Cerato- 

 phyllum and Pijjeromia the carpels are all posterior, the uni- 

 formity of the position of the carpel in Nelumbium forms only a 

 partial objection, and the former is probably the true solution of 

 the question. 



A further argument in support of the hypothesis that in Ne- 

 lumbium the ovary has the ventral suture turned outwards, is de- 

 rived from the fact that in the Nymphal Alliance the plaeentation 

 is, as Mr. Brown has remarked, dorsal, variations even when 

 they do take place being only partial ; for supposing the carpel 

 to be so placed, Nelumbium agrees in this character also in the 

 adherent funiculus being always at the inner angle of the carpel 

 (PI. III. figs. 1, 2, 3 & 4). The occurrence of dorsal plaeentation 

 in other families allied more or less nearly also makes it probable 

 that this is the true structure of Nelumbium, and on this account 

 an especial notice of them may be the more interesting, passing 

 over those in which the ovules are numerous, viz. Orobanchaceaa 

 (in some genera only), Nymphfeacese, Butomacese, and possibly 

 Hydrocharidese. 



1. Hydi-opeltis purpurea. Ovules two, pendulous, anatropal, 

 having the raphe turned away from the placenta, and attached 

 one above the other to the dorsum of the carpel*. (PI. III. fig. 6.) 



2. Cabomba aquaticu. Ovules three, pendulous, anatropal, 

 having the raphe turned away from the placenta (occasionally 

 lateral ?), one attached to the ventral suture near the apex of the 

 carpel, and the two others to its sides midway between the dor- 

 sal and ventral sutures. These two ovules are attached to two 

 cord-like ribs which originate in the base of the cell, and are 

 continued upwards to the attachment of the third ovule. (PI. III. 

 fig. 7.) 



3. Cei-atophj/llum demersum. Ovule single, pendulous from 

 the apex of the cell in consequence of the funiculus to the apex 

 of which it is attached being firmly adherent to the dorsum of 

 the carpel. This funiculus originates in the base of the carpel 



* That the ovule is aiiatiopal is further shown by the embryo being nc.\t 

 the hihim. 



