Cruciferous plants, the generic differences of vvhigli are diffij 

 cult to seize perhaps beyond all others, vv'as made by the 

 gentleman to whom we are obliged for our article ; and 

 that the soundness of those characters has' never yet been 

 questioned, even by theorists. 



In illustration of these remarks, we cannot do better 

 than transcribe the words of M. Decandolle, who (H^-^n. vrg. 

 141) observes, that '• A methodical arrangemeut of cru- 

 " ciferous plants is, on account of the close afil"Hy of 

 " the genera, both difficult and ambiguous. The division 

 " of the order into S'lUquoscc and Siliculoscv, "which was 

 " first established by Ray, and afterwards adopted by Lin- 

 " nscus and most other botanists, although it has the ap- 

 *' pearance of being convenient, is not only occasionally 

 " uncertain, but has the great defect of not according with 

 " the anatomical structure and natural affinities of the ge- 

 " nera. In the first place, the silique and the silicule differ 

 ' " from each other only in length ; and every degree of 

 " length is to be found, not only in allied genera, but even 

 " among species of one and the same genus; so thatDuABA, 

 *' for instance, Nasturtium, Heliophila, Erysimum, Arabis, | 



" Stevenia, Eruca, and others, even after the most rigid 

 " examinatioUjUeccssarily include examples of Siliquosce and 

 " Siliculoficc mingled together. Secondly, there are certain 



€t 



tt 



cruciferous genera, such as Rapistrum, Cakile, Crambe, 

 &c. the fruit of which being neither silique nor silicule, is 

 so anomalous as to have received a distinct appellation 

 " from some modern writers, and to have been termed nn- 

 " camentaceous. In the third place, the length of the 

 " style is not always in an invTrse ratio to the length of the 

 " pod, but has been observed to be long in many SlUquosce, 

 and very short or nearly obsolete in other S'lliculosce. 

 Characters, however, of a more constant and important 

 description, having been remarked by Gsertner, were suc- 

 " cessfully introduced as the basis of the generic characters 

 " of the order by Mr. Robert Brown. These characters are 

 " deduced from the internal structure of the seed, and mcr^ 

 especially from the relative position of the ridicle and 



(( 



{{ 



a 



cc 



tc 



a 



cotyledons; and although it must be confer cd, th 



f 



4/ 



prejudices with regard to the association of species, which 

 " have become almost sanctified in our minds by habit, are 

 " in some cases attacked, the genera founded upon those 

 " principles are not only much better defined, but far more 



