BOTANICAL LITEllATUBE. 207 



and concluding period is not a part of the ''scientific" 

 character, hut answers to that supplementary note wliich, in 

 the formulary of more recent authors, is set off from the 

 tecliuical description by a dash after the period, to which 

 place are relegated all matters appertaining to habit and 



the vegetative organs. 



The next genus after Cardamine, with our author, is Dea- 

 iaiHa. The generic character is the same, except that in 

 reference to the form of the seed roiiindatis takes the place 

 of orhiciilaiis; but in the subjoined note comes out the real 

 distinction which is drawn between Cardamine and Deniaria, 

 the fleshy and scaly or notched rhizomes of the Latter. In 

 this instance, the new principle, that genera are to be deter- 

 mined by flower and fruit only, fails to give the desired 

 result; and again and again it fails. Next after Deniaria 

 follows Sis})mhrinm {W\e Naslnrtiam of more recent botan- 

 ists) ; and "the only alteration in the generic character is that 

 of the substitution of subrofnmUs for rotundatis in the 

 description of the seeds. But the real mark of the genus 

 comes out only in the superadded note : "a certain general 

 aspect peculiar to the species of this genus ;" though at the 

 end of the catalogue of species he states, very aptly, this 

 negative generic character. " Sisymhrmm differs from Car- 

 damine in that the valves of its siHque are not elastic;" but 

 adds that "from Leiicoium (our Cheiranthus), Brassica and 

 some others, it is to be distinguished by habit only." To 

 Eruca, the next in order, is ascribed precisely the generic 

 character of Sisymhrium, while the real mark of the genus, 

 as indicated at the end is, " the peculiar flavor." _ 



But these are genera of Crucifen^, an order in ^.-lllch all 

 the plants are so closely related one to another that it becomes 

 more than usually difficult to set boundaries to genera. 

 Nevertheless, Tournefort everywhere, m all families of plan s 

 and even in the classifying of trees, gives recognition to 

 natural genera, whether he finds an organographic character 

 or fails ?o find it. Sail, and Popnlus are doubtless among 

 the best of genera. They are so natural that not the least 



