CBUGIFEBJE. 219 



a. Cruciferce hypogynce. 



1. Cheiranthece. — Siliqua dehiscent longitudinally. 



2. Raphaiiece. — Fruit elongated, (usually) indehiscent. 



3. Cakilece. — Fruit elongated, more rarely short, lomentaceous. 



4. Isatidece. — Silicule inarticulate, indehiscent. 



5. Lunariece. — Silicule dehiscent, compressed parallel to the septum. 1 



6. Thlaspidece. — Silicule dehiscent, compressed at right angles to the septum. 2 



b. Cruciferce perigynce. 



7. Subularice. — Silicule turgid. 



Next we use for the formation of subseries the less constant and 



easy character of the relations of the radicle and cotyledons, and we 



shall thus found in certain of these series secondary divisions which 



though far less absolute in their limits, may yet have great practical 



utility. Thus we divide the Cheiranthece into three subseries as 



follows : — 



( 1. Arabidinecc. — Cotyledons (usually)' 1 accumbent. 

 CHEiRANTHEiE. < 2. Sisymbrinece. — Cotyledons incumbent. 

 ( 3. Brassicineoe. — Cotyledons conduplicate. 



So also with Lunariece : — 



( 1. Alyssince. — Cotyledons (usually) accumbent. 

 LuNARiiE. < 2. Camelinece. — Cotyledons incumbent. 

 ( 3. Succovinece. — Cotyledons induplicate. 



And Thlaspidece : — 



m ( 1. Iberidinece. — Cotyledons (usually accumbent). 



( 2. Lepidinece. — Cotyledons incumbent (or conduplicate). 



The other series are more homogeneous, and remain undivided. 

 The genera will be distinguished below by characters of the third 

 grade, such as those of the insertion of the sepals ; the details of the 

 androceum and of the gynseceum, including ovary, septum, style and 

 stigma, and the fruit, seeds, funicle, &c. 



The affinities of Cruciferce have long been recognised. 4 The order 

 contains none of those types with free carpels which occur in 



1 Even when the compression is ill marked 1), on every opportunity, on the exceptions, which 

 the breadth of the septum remains always nearly are daily increasing in number. Bentham and 

 equal to that of the valves. Hookee give them at the head of each secondary 



2 Which is always narrower than the valves. group. 



3 We have elsewhere dwelt (see p. 218, note 4 Mieb., in Ann. Sc. Nat., ser. 1, vi. 266. — 



