M. Treviranus on the Structure of the Fruit of the Cruciferse. 367 



to the general mode of origin of fruit, and if we regard, with 

 Brown and DeCandolle, the many-celled fruits as originally com- 

 posed of as many carpels as they have chambers, we must assume 

 that the silique also consists regularly of two, very rarely, as in 

 Tetrapoma, of four carpels, united in an apparently simple 

 manner. But the question then arises, whether the placenta 

 is an appendage of these carpels or an independent part, that 

 is, a member of an inner whorl, the members of which alter- 

 nate with the carpels as belonging to a more external whorl ; — 

 whether therefore it is, as some like to express themselves, an 

 appendicular organ or a product of the axis. Assuming the 

 former, the greatest difficulty is found in attempting to explain 

 the production of the septum in the silique. De-Candolle re- 

 garded it as an expansion of the two placentas, and he sought 

 to support this view by pointing out that the septum is always 

 found to be thinner in the middle; that here it is readily torn 

 in the direction of its length as in Thlaspi arvense, or is ac- 

 tually slit as in Cheiranthus Cheiri, or has a hole in the middle 

 as in Farsetia cegyptiaca, Octadenia lybica, Vesicaria gracilis, &c, 

 or is wanting in the middle and only present along the placentas 

 as in Tetrapoma. But leaving out of the question that this by 

 no means justifies such a conclusion, such an expansion would 

 not agree with the original assumption, which settles the placenta 

 to be a marginal production of the mere carpels. A. Brongniart* 

 has described a monstrous form of the silique of Cheiranthus Cheiri, 

 where instead of ovules small leaflets and indeed free leaves were 

 produced. But how the formation of the silique is to be con- 

 ceived so as to agree with this observation has not been stated by 

 the author, who also holds that the theory of the production of the 

 ovules from marginal teeth of the carpels is not applicable to all 

 fruits, for instance not to the Primulacese, Myrsinacea?, &c. Kunth 

 has set forth the opinion that the silique is a structure composed 

 of four leaves grown together at their margins, two of which have 

 been perfectly developed outwards, but two are only developed 

 inwards, because their growth outward is prevented by the press- 

 ure arising from the crowded position upon the axis, so that 

 they produce seeds and may easily become united and form a 

 septum f. Bernhardi also, citing a remarkable malformation of 

 the fruit of Ricotia observed by himself, considered the silique as 

 the product of four leaves grown together, two of which are ar- 

 rested in their development J. But more recently he has given 

 up this theory, since perfectly formed siliques are found with four 



* Ann. d. Sc. Nat. 3 s6r. i. 29. 



t Abhandl. d. K. Acad. d. W. zu Berlin f., 132; Lehrb. d. Botanik, i. 

 397, 458. 



% Ueb. den Begriff d. Pflanzen, art. 47. 



