214 Statistics of the Flora of the Northern States. 



and Middle States, north of Virginia, 1833, although the elements 

 are considerably different and the ten largest orders are not the 

 same throughout.* 



Moreover, our ten predominant families do not properly cor- 

 respond with the ten mentioned by De Candolle as generally pre- 

 dominant in the temperate regions of the northern hemisphere : 

 viz. " of the first rank, Compositce, Graminece, Cyperacece, Legu- 

 minosce ; then the Cruciferae, Umbelliferce, and Caryophyllacece, and 

 then, though less decidedly, the Labiatce, Rosacece, and Scrophu* 

 lariacece.f Nor would they do so if, by dividing the Ericaceae 

 into smaller orders, we were to exclude that family from the list 

 of those (eleven in number) which severally comprise not less 

 than two per cent of our phaenogamous species. The three most 

 predominant families accord indeed with De Candolle's conclu- 

 sion, only the Cyperacece with us are remarkable for surpassing 

 the Graminece. But the next three in our list are quite differ- 

 ent, even if we omit Ericacece, being Rosacece, Scrophulariacece, and 

 Orchidacece ; and all three of De Candolles second rank fall be- 

 low our first ten ; and one of them, the order CaryophyllacecB 

 would fall still lower, if it were not reinforced by the lllecebrece, 

 so generally regarded as a distinct family. 



It is easy to see that these differences are owing to the unusual 

 richness of our flora in Cyperacece (chiefly in Carices), and to our 

 poverty in Cruciferce, Umbelliferce, Caryophyllacece, and Labiatce, 

 especially in the second and fourth, at least as compared with 

 corresponding parts of Europe. 



* The schedule drawn from Beck's Botany is as follows : 



:1066 species out of 2125 Phsenogamous 

 plants. 



The differences are readily to be accounted for. 1. The substitution of Amentacece 

 in this list for Ericacece in the other, results from the former Jussiaean order having 

 been preserved entire by Beck, but distributed into several in the present work; while 

 I have admitted the order Ericacece in its most extensive sense. 2. The precedence 

 of Cyperacece to Graminece in my list,— *which appears not to be the case in corres- 

 ponding floras of the Old World,— -is wholly owing to the great increase in the num- 

 ber of Carices, in which the Northern United States are absolutely very rich ; which 

 increase has resulted from the remarkable attention and repeated elaboration this 

 genus has received since Dr. Beck's time, from several hands, and perhaps also from 

 a minuter discrimination of the species than in other families. 3. The order Rosacea?, 

 which strangely takes precedence of the Legvminosce, is unduly expanded by a 

 crowd of nominal or traditional species, and has four times as many introduced 

 species as the latter family. 4. The naturalized plants being included, alters the 

 proper proportion of most of these orders, and swells the number of the Pharnoga- 

 mous plants to 2125, while we count only 2091 truly indigenous species within an 

 area about one-half larger and now much more thoroughly known. 



f Alph. De Candolle : Geogr. Bot., p. 1245. 



Composite, 



265 " 



Graminese, 



169 



Cyperacese, 



157 



Rosacea?, 



97 



Amentacese, 



94 



Leguminosse, 



80 



Labiatse, 



59 



RanunculaceaB, 



50 



Scrophulariacess, 



48 



Orchidacese, 



47 , 



