90 DEVELOPMENT OF THE FLOWER 
the stigmata, b. carpel leaf of style, c. line of union of ditto, 
presenting distinct fissures. 
In Gloriosa the alternation is complete. 
Pitcairnia bromeliefolia. Pl. 34.—Fig. 18, a. a. bract. 
Crinum, Pl. 34.— Fig. 19. This is the same as Polianthes ; 
the ovule or seminal cavities are opposite to the petals, but 
lines of communication or stigmatic surfaces are opposite the 
sepals, hence these analogically correspond to the cavities of 
the carpellary leaves, which are prolongations inwards and 
downwards of the furrows of the stigma, hence they are 
spaces pointing out the composition of the pistillum. They 
exist in Funkia, which is amarylled in appearance. 
Calosanthes. Pl. 39.—Fig. 1. 
[Stage 2. a. Stamen, b. plane of verticillus solid. 
h . a. Carpel leaves, have no right and left disposition. 
»  4.b. anterior, and c. posterior. (a) No disc. 
» 9. a, Upper, b. disc. 
» 6. Stamens still hypogynous. Stigma now evident, lamellz anterior 0, 
posterior. ?] 
It would perhaps seem to follow as a matter of course, 
that a corolla developing in protuberances from a solid plane 
must originally be composed of as many parts as there are 
protuberances. And as these points correspond to the fu- 
ture petals, it would hence follow that all corollas must at 
some period or other be polypetalous. And this I think I 
have observed to be the case, although the appearances are 
obscured by the likeness, when viewed externally, the inter- 
node has to the tube of a young corolla. A doubt has been 
also suggested to me by the earliest development of the 
ovarium which has first appeared to me not unfrequently in 
the shape of an entire annulus, its protuberances whereby 
early periods the formation of the septum, of which I do 
not see any notice has been taken. It is complete even 
"S 
* 
