100 



TAXONOMY. 



(iannot assert that he would have formed fewer genera than 

 his followers have made ; but as he put less value upon the 

 fruit in particular than was proper, many of his genera are 

 too comprehensive, not to be separated with advantage. Who- 

 ever should receive the Fumaria or Poh/podium of Linnaeus 

 even now, in their entire latitude, would evidently overlook 

 the most important and essential differences, and strive for 

 simplicity at the expence of Science, and even of Nature. 



The second error leads to too fine a discrimination of insig- 

 nificant marks, which might well furnish the foundation for 

 constructing a species, but can never be approved of as generic 

 characters, unless we would ultimately make as many genera 

 as there are species., 



1495. 



We are naturally led to ask, how we may most securely 

 avoid both these errors. It cannot be denied, that the power 

 of separating important from unimportant circumstances, and 

 of being as little misled by fancy as by refinement, in connect- 

 ing or separating things which ought not to be connected or 

 separated,, is a talent which is either born with some indivi- 

 duals, or acquired to a great extent by practice. This talent 

 belongs to some Botanists in so high a degree, that we may 

 safely trust ourselves to their glance, and to their discrimi- 

 nation. Others, with the best intentions, have never been 

 able to acquire the talent, especially as they seem to believe 

 that refinement is the only requisite for the construction of 

 species and genera. 



150. 



In the first place, in order to avoid these errors, Genera- 

 must be founded on such characters, as, compared with one 

 another, have an evidently uniform value. Thus, when a 

 certain number of genera have already been distinguished from 

 one another by the difference of one character, we may with 

 propriety avail ourselves of new differences of the same, to 

 construct new genera ; since the organ or property must, m 

 all similar cases, have a uniform value. If we have thus 

 once begun to distinguish the Umbellatae by the form of their 



