122 TRUE FORMATION OF THE GENERA. 



some species of another genus near of kin to it, a like caution is 

 again necessary on the other hand, lest we . should multiply the 

 genera, by parting species that should stand together: thus we 

 find, that in Sedum, Sempervivum, Rhodiola, Crasstjla, Til- 

 uea, and Cotyledon, the nectaria adhere to the base of the pi- 

 stillum;...in Epilobium and CEnothera the calyx is tubulose;... 

 in Mespilus, Crataegus, and Sorbus, the structure of the flower 

 is alike and in both Alnus and Betula, there are three florets 

 on the foliole of the amentum*. 



CHAP. XXX. 



By what Parts of Fructification the Genus may with 

 the most Certainty be determined. 



THE more constant any part of the fructification is found 

 through the several species of any genus, the more it may be 

 relied on with certainty, as a characteristic mark for that genus. 

 Thus in Hypecoum the nectarium is constant, but not the siliqua; 

 ...the Convallaria is constant in its spotted berry, but not in its 

 corolla ;... the Lobelia in its corolla, but not in its fruit;. ..the 

 Cassia in its corolla, but not in its siliqua;... and the Verbena 

 in its calyx and corolla, but not in its stamina and seeds. 



In some genera one part of the fructification is found to be the 

 most constant, and in others another ; but there is no part that 



* The alnus and betula are joined by Lhmeeus under the title of Betula. The' 

 rest of these instances he has kept separate, notwithstanding the doubt raised here 

 concerning the propriety of distinguishing them. Author. 



