MUTUAL COXNECTIQKS OF PI.AXTS. 99 



148. ~ 

 But we must confess, that there is an iitlportarit difference 

 in genera, according as they are founded on an agreement in 

 the properties of most of the parts, or only in the marks of 

 a few essential organs. When the former is the case, we re- 

 cognise a Natural Genus. We then observe correspondences 

 in structure, in external appearance, in situation ; often in the 

 form of the roots, the leaves, the buds, in the subordinate 

 parts, or in the armour and supports ; sometimes even in the 

 composition of the sap, in the colours, in the smell and taste. 

 Such natural genera are, for example, those of the Rose, 

 Wheat, Stocks, Willows, and innumerable others. 



Artificial Genera, again, are those which, though they want 

 a correspondence in external appearance, shew the same for- 

 mation of the essential parts. Whenever the preference has 

 been given to the organs of fructification as the essential or- 

 gans, we must then be permitted to connect the agreement 

 of these with the idea of a genus. Indeed, such genera are 

 by no means adapted to strike the eyes of every person ; nay, 

 in a Natural arrangement of plants, we might even very pos- 

 sibly overlook them. But when Art has once reared a sys- 

 tem, she must necessarily assume constant differences in the 

 essential parts, as foundations for the discrimination of ge- 

 nera. If the innumerable plants which are known to us as 

 Umbellatse, and which agree more or less in their external 

 marks, were not marked out as peculiar genera by such line 

 and scanty characters of the essential parts, it would be con- 

 trary to all scientific ideas, to unite them into one common 

 genus, which would then include within it innumerable species, 

 with important differences in the fruit, and in the other es- 

 sential parts. 



149«. 



There are here two opposite errors to be avoided. If we 

 fall into the one, we then seek to simplify all things, and col- 

 lect the most different forms into a few great genera. The 

 less progress the knowledge of plants has made, the more are 

 we disposed to lose ourselves in this error. Indeed, when we 

 attend to the number of species which liinnaeus knew, we 



G '2 



