Cl. 12.] UMBELLIFEIUE. 133 



the presence or absence of their general or partial In- 

 volucrum. Artedi, the early friend of Linnaeus, who 

 devoted himself to the study of the Umbellifertf, sug- 

 gested, or adopted, this plan. But those parts are often 

 variable in the same species. The regularity or irre- 

 gularity of the Petals also, and the perfection or par- 

 tial imperfection of the Stamens or Pistils, have been 

 resorted to, and do perhaps often afford good marks. 

 The simple or divided form of the Petals is very ma- 

 terial. But the figure, margin, ribs, angles, and sur- 

 face of the Seeds yield excellent characters, allsuffi- 

 cient for the establishment of good genera, though not 

 yet perfectly well applied to use. The earlier syste- 

 matic botanists, and more recently Crantz and Cus- 

 son, have had this object in view. Hoffmann and 

 Sprengel are now intent upon it. The Prodromus of 

 the latter, published at Halle in 1813, does honour to 

 it's author, though his Species UmbeUiferarum minus 

 cognita, published five years later, may serve to show 

 that his ideas of genera are not yet settled. It would 

 be superfluous to give the detail of Jussieu's 4 Sections. 

 Sprengel's' are as follows : 



1. Fruit compressed, flat. Hasselquistia, Tordyllum, 

 Heracleum, Peucedanum, fig. 203, Ferula and Pasti- 

 naca are good examples. Hydrocotyle appears mis- 

 placed here. 



2. Fr. solid, winged at the margin. Drusa, De 

 Cand, Ann. du Mtis. v. 10, Mulinum Persoon, Se- 



