OF NORTH AMERICA. , 15 



508. SoRBus RiPARiA Raf. Autikon. Branch- 

 es rugose, folioles 9 to 15 oblong sessile, base 

 oblique entire, end acute equaly niucronate 

 serrate, odd leaf petiolate broader acununate ; 

 corymb paniculate, berries pisiform globular. 

 —A small tree 10 to 20 feet high, growing on 

 the margins of Rivers, the Missouri, upper 

 Mississipp, lower Ohio, Wabash, Illinois ; but 

 rare. The fruit is very small commonly of a 

 saffron color with 3 seeds oval compressed in- 

 closed in a tough shell, and thus nearly a drupe 

 sometimes only one seed by abortion. Near 

 S. microcarpa, which however has folioles 

 acum. unequally serrate, not obliquate, berries 

 scarlet and larger. Here the folioles are 2 or 

 3 inches long, pale beneath, quite smooth, pe- 

 tiols compressed at the base. The g«ius Sor- 

 Bus must be preserved, although some writers 

 wrongly unite it to Pykus : it is known at 

 first sight by the pinnate leaves, and the calix 

 not persistent nor crowning the fruit. It would^ 

 be better to unite to it all the tristyle sp. ot 

 Cratesfus than to abolish it. 



509. TPvILOPUS Mitchell, or HAMAME- 

 LIS Linneus, name posterior? This G. has 

 puzzled the Botani.^ts, Jussieu wrongly united 

 it to Berberides. It has now be made the sin- 

 gle type of the Hamamelides ; but it is so near 

 to my ScLERANTiiiDES, that it must probably be 

 united thereto as a subfamily. Only 3 species 

 were known; besides the doubtful varieties of 

 Walter ; but I have observed 6 species, and 

 therefore shall now give their Monograph. They 

 may be called hyemal shrubs, since they blos- 

 som late in the autumn, after the leaves have 

 begun to fall. 



510. Tr. or H. virginica Raf. med, fl. tab. 



