318 THE UMBELLATE FAMILY. 



XXXII. UMBELLATE FAMILY. UMBELLIFEILE. 



Herbs, or, in a few exotic species, shrubs, with alternate leaves, 

 often much cut or divided ; the footstalk usually dilated at the 

 base, but no real stipules. Flowers usually small, in terminal or 

 lateral umbels, which are either compound, each ray of the general 

 umbel bearing a partial umbel, or more rarely simple or reduced 

 to a globular head. At the base of the umbel are often one or 

 more bracts, constituting the involucre, those at the base of the 

 partial umbel being termed the involucel. Calyx combined with 

 the ovary, either entirely so or appearing only in the form of 

 5 small teeth round its summit. Petals 5, inserted round a little 

 fleshy disk which crowns the ovary, usually turned in at the point, 

 and often appearing notched. Stamens 5, alternating with the 

 petals. Ovary 2-celled, with one ovule in each cell. Styles 2, 

 arising from the centre of the disk. Fruit, when ripe, separating 

 into 2 one-seeded, in dehiscent carpels, usually leaving a filiform 

 central axis, either entire or splitting into two. This axis, often 

 called the carpophore, is however sometimes scarcely separable 

 from the carpels. Each carpel (often called a niericarp, and hav- 

 ing the appearance of a seed) is marked outside with 10, 5, or 

 fewer, prominent nerves or ribs, occasionally expanded into wings, 

 and underneath or within the pericarps are often longitudinal 

 channels, called vittas, filled with an oily or resinous substance. 

 Embryo minute, in a horny albumen, which either fills the seeds 

 or is deeply furrowed or excavated on the. inner face. 



A numerous family, more or less represented nearly all over the 

 globe ; but the species are comparatively few in high northern latitudes, 

 as well as within the tropics, their great centre being western Asia and 

 the Mediterranean region. Their inflorescence, and the structure of 

 their flowers, distinguish them at once from all other families, except 

 that of the Aralias, and these have either more than two styles, or the 

 fruit is a berry. But the subdivision of Umbellifers into genera is 

 much more difficult. Linnseus marked out several which were natural, 

 but without definite characters to distinguish them ; and the modern 

 genera, founded upon a nice appreciation of minute differences in the 

 fruit and seed, are often very artificial, or still more frequently re- 

 duced to single species, and require as complete a revision as the Crucifers 

 and Composites. These minute characters are moreover in many cases 



