North American Plants. 93 



to employ characters derived from Cotyledons, in subdividing 

 this great natural family. 



I have remarked that the genus Leavenworthia can hardly 

 be referred, with certainty, to either of De Candolle's 

 grand divisions of the Cruciferas. In the mature seed the 

 radicle is only about one-fourth the length of the cotyledons, 

 very slightly curved, or bent a little to one side, thus showing 

 a tendency to become accumbent. This is its form at the 

 earliest appearance of the embryo.* It is always directed up- 

 ward (or towards the style, as the seed lies in the pod) ; at 

 first turned a little from the umbilicus towards the opposite 

 side of the replum, but gradually becoming erect, or even 

 inclining towards the hilum, as the seed becomes more com- 

 pletely campulitropous. 



It remains for me to notice the affinities of this new genus. 

 Disregarding the peculiarity of the embryo, the only tribe of 

 Pleurorhiz.e, with which it can be compared, is Arabidece ; 

 and among Notorhiz.e, it has but little resemblance to any, 

 except to a small section (Arabidopsis) of the genus Sisym- 

 brium, all of which have white flowers. 



In the form of the silique, and in the margined seeds, it ap- 

 proaches some species of Arabis ; but the valves are not nerved, 

 and the flowers are yellow. Its resemblance to the species of 

 Cardamine with pinnated leaves, particularly to C. Ludovi- 

 ciana,-f (which deviates from the character of the genus in its 



* The embryo of Cruciferae, when first visible, is always straight, lying with its 

 radicle pointing to the foramen on one side of the campulitropous seed : viz. 

 that which would be the superior portion of it were the seed straight, with 

 the cotyledons directed to the curved part, or apparent summit. As the embryo 

 grows, it curves round, so that the cotyledons fold upon the radicle, and are thus 

 directed to the hilum, which, in seeds of this kind, is always next the foramen. 

 The gradual curvature of the embryo in Cruciferffl is clearly exhibited in plate 

 42, fig. 3 of Brongniart's admirable Memoire sur la Generation et le Developpe- 

 ment de VEmbryon dam les Veg. Pltaa. Ann. des Sc. Nat. torn. 12. I have ob- 

 served the same changes in Arabis Turrita and A. ambigua. 



f C. glauca Spreng. (Deless. ic. 2, t. 31.) also has margined seeds. 



