42 LETTER XXIX. 



the unexpanded flower-buds of this plant {jig. ^2.). 

 If you cut them across you will find their appearance 

 in a transverse section sufficiently curious. They 

 consist of several green leaves wTapped one over the 

 other (Jig- 3.), and enclosed within a couple of concave 

 bracts ; within these lie the petals, enwrapping the 

 stamens, which are closely packed round either the 

 g}Tiophore or the ovary. 



Our gardens contain nothing included in the same 

 group as the Caper, except certain annuals called 

 Cleomes, a few of which have gay starry flowers, and 

 long stamens, far less numerous than in the Caper itself. 



It is obvious that this plant accords with the Mig- 

 nonette tribe more than any others yet examined. It 

 has, independently of its polypetalous flowers, a con- 

 siderable number of stamens, a disk adherinof to the 

 part in which the stamens originate, a gjTiophore on 

 which the ovary is elevated, an ovary with parietal 

 placentation, and kidney-shaped seeds, with a curved 

 dicotyledonous embryo. These circumstances un- 

 doubtedly indicate a near alliance between the Caper 

 and Mignonette, and, in reality, the general opinion 

 now seems to be in favour of their standing next each 

 other, only in distinct groups. 



With regard to Cleomes, I must refer you to the 

 Hothouse for information concerning them. They 

 are considered to stand, as it were, between the Caper 

 tribe and the Cruciferous tribe (Vol. I. p. 55.) ; con- 

 necting, in a very conspicuous manner, plants that 

 otherwise would not have been readily brought near 

 each other. 



