Antoine- Laurent Jussieu. 15 



genera to species ; they descend from general to particular. 

 But M. de Jussieu completely reverses the proceeding ; he 

 ascends, as he himself has stated, fro7n jyarticulars to gene- 

 rals ; and all the diiference between the artificial methods and 

 the natural method consists in this, — that the former subject the 

 species to the genera, and the genera to the classes ; the 

 latter, on the contrary, subjects the classes to the genera, and 

 the genera to the species ; the foi'mer subjects facts to ideas, 

 the latter ideas to facts. 



In the new department which the consideration of alliances 

 opened up to science, every step of M. de Jussieu excited the 

 attention of the naturalist ; the secret of his power being the 

 route he followed. The example of those families which are 

 conspicuously and remarkably natm*al, constituted his guide 

 in those which are less so. He saw in those families which, 

 in the eyes of all botanists, are natural, such as the Gra- 

 minece, the Compositce, the Legimiinosoi, Umhelliferoi, &c., that 

 ■ the species were grouped conformably to the ensemble of 

 their structm'e ; and this was a ray of light to guide his path ; 

 every character which, upon application to one of these fami- 

 lies broke up the species, required necessarily to be excluded ; 

 and hence the first condition of every character was to respect 

 the alliances of the species, grounded upon the totality of 

 their structiu-e. And this calculation of the importance of 

 characters, deduced from their relation to the totality of their 

 structm*e, is the principle upon which the work of M. de 

 Jussieu is wholly based. The especial object of this work, is 

 the distribution of genera into families. Tournefort had pre- 

 viously ai'ranged the whole of the species into genera ; Lin- 

 naeus had conferred upon these primary genera more regu- 

 larity and precision ; and what remained to be done was, to 

 effect for the groups of a more elevated order, — for the groups 

 which had even been omitted in the systems of Tom'nefort 

 and Linnaeus, in a word, for the families, that which these 

 great botanists had done for the genera. 



M. de Jussieu distributed all the genera known at the time 

 he wrote, amounting to nearly 2000, into one hundred fami- 

 lies. He founded each of these primitive families upon a fixed 

 totality of characters, and he demonstrated that the concur- 



