Antoine Laurent de Jussieu. 303 



for applying and developing these principles; and the exactness, 

 perspicuity, and precision of these characters, particularly such as 

 apply to the families, still authorize us, if we keep in mind the pe- 

 riod when they were framed, to consider them as models which few 

 authors have equalled, and none surpassed. 



Finally, the notes appended to the greater part of these families 

 form, perhaps, that portion of the work which most evinces the 

 judgment and extensive knowledge of the author. 



It was in them that he often corrected the artificial tendency 

 which a linear series always assumes, that he pointed out the mul- 

 tiplied relations of families to each other, and that he indicated the 

 doubts left upon his mind by imperfect observations which he had 

 been unable to verify, or which led to the presentiment of remote 

 affinities, a foresight which greatly outstripped, so to speak, the 

 actual state of the science. Many of the improvements subsequent- 

 ly introduced into the natural method, are in reality, foreseen or in- 

 dicated either in these notes, or in the sectional divisions of the fa- 

 milies, or by a word placed at the end of the generic character. 



This last part of the work, the characters of the genera, consi- 

 dered by some superficial authors as a simple compilation, is not in 

 our estimation, the least remarkable feature of it. Certainly the 

 work would have presented, after the characters and notes on the 

 families, a list of the genera comprised in each of these families, as 

 every one has since done who has followed in the same track, and 

 this of itself would have rendered an immense service to the science, 

 and sufficed to elucidate the natural method. Yet, without gene- 

 ric characters, a table of the families would have only been a sub- 

 ject for study and reflection, and would not have been adapted to 

 actual use, nor formed a manual, so to speak, for the botanist ; and 

 the natural method would have been disseminated much more slow- 

 ly among the learned. 



But in introducing generic characters, it may be asserted that 

 they could not in general be taken by compilation, even from the 

 most esteemed works of the period ; for characters simply distinc- 

 tive, suitable for an artificial system, would often be quite unadapt- 

 ed to a natural method ; or a character which might appear trivial 

 to the author of the former, might acquire great importance in the 

 eyes of him who studies natural relations. Accordingly, the cha- 

 racters of the Genera have been generally traced by the hand of 

 Jussieu, either after nature, or after the published or manuscript de- 

 scriptions of botanists in whom he could place confidence, and the 



