18 FLOR. TELL. 



shapes &-c, thus belonging to 7 Unnean classes, 

 and being merely united by the fruit, as if the 

 fruit alone was to form Genera, whereby all the 

 Cruciferes should be one Genus ! all the Gen- 

 tians have a bivalve unilocular capsule, and so 

 have 100 other Genera. Meantime I have de- 

 tected in the whole group 4 important charac- 

 ters overlooked by all the botanists. 1st. All 

 the Stamens are opposed to the segments of the 

 Corolla as in the Primulacea ! this will remove 

 the Gentianides in the serial order : any one 

 with alternate stamens must be removed from 

 the family. 2d. The segments of the calix are 

 more or less unequal in length or breadth in 

 nearly all, this is a generic character. 3d. 

 The stamens are equal, when they are unequal 

 as in Lisianthus &.c, those Genera belong to 

 Lisuinthides,^s\\h^^.vci\\y of Convolvulides, 4th. 

 All the leaves are entire sessile, opposite or ver- 

 ticilate with reticulated veins besides the 

 nerves, which are easily perceived in all their 

 leaves. I shall now proceed with this fine 

 group, dividing it into many good Genera with 

 the akin Chironia, and adding some new spe- 

 cies; but many more are in my flora of North 

 America. 



436. Gentiana Ad. non Necker, cal. camp, 

 anomalo spathaceo seu truncato, l-61obato. Cor. 

 camp. 6-91oba equalis, stam. 6-91iberis. Rad, 

 crassis amara, Fol. etfl. opj?os. seu vertic. — 

 This must remain as the typical Genus, and 

 the typical Sp. is G. lutea or officinalis, to 

 which that name was given by the greeks. But 

 the Genus is thus reduced to but few species, 

 and even requires to form three subgenera. 1. 

 Picriza (bitter root) Cal. 3-61obato, Cor. camp, 

 rotata. This includes G. lutea., pannonica. 



