Genera and Subgenera should be established, 389 



other characters, the value of that character would, in con- 

 sequence be, to a certain extent, enhanced, and we might 

 justly transfer it from particular species to an entire group. 

 It is owing to circumstances of this nature that many of the 

 Linnaean genera are now with great propriety considered as 

 natural families, while several of the species of that day are 

 found to be types of so many natural genera ; and, in the 

 same manner, those groups which, in the present state of 

 the science, we consider as subgenera, may hereafter rise in 

 value, and be found to constitute groups of a higher de- 

 nomination. 



All that we can do is, to determine the value of a group 

 with reference to the present state of our knowledge of the 

 family to which it belongs : and, for this purpose, we must 

 carefully compare its characters with those of other ackfiow- 

 ledged genera belonging to that family. If, on such com- 

 parison, they appear to be of equal value with the characters 

 of these genera, the group in question may be considered as 

 a genus also : but, if of less, it is clear that the group itself 

 is of less importance, and that it must occupy a subordinate 

 station. We are to take care, however, that such comparison 

 be strictly confined to genera of the same family, in which 

 alone the same character is necessarily of equal value in all 

 cases.* 



What has been said hitherto relates to the principle by 

 which the value of groups is regulated, and is intended to 

 show the impropriety of applying the same term, genus, to two 

 groups, however natural in themselves, of which the values 

 are different. It is not merely, however, to the neglect of 

 this principle, that we may attribute the great mutiplication 

 of genera at the present day. Many of the groups which 

 modern naturalists have designated by this title hardly 

 deserve to rank as subgenera. They rest on characters far 

 too trivial and unimportant, and unconnected with any 

 marked peculiarity either of form or habit. In some cases, 

 one isolated character of this nature has been laid hold of, in 

 order to effect a division of those genera which it was deemed 

 expedient to break up, merely because they contained a large 

 number of species. But no supposed expediency can justify 

 such a step, which is at variance with the well known and 

 universally received Linnaean maxim, character non facit genus. 

 However allowable in an artificial system, of which the sole 



* " Les memes caracteres n'ont pas la m6me valeur dans toutes les 

 families, dans tous les genres." (Decandolle, Theorie Elementaire de la 

 Botanique, p. 214. ; a work which should be read by every student in 

 zoology as well as botany.) 



c c 3 



