250 Agassi^ Contributions to the 



as worked out in the ultimate details of structure, which distin- 

 guishes one genus from another ? Latreille, in expressing the 

 want he felt with reference to the study of genera, has given us 

 the key-note of their harmonious relations to one another. 

 Genera are most closely allied groups of animals, differing neither 

 in form, nor in complication of structure, hut simply in the ulti- 

 mate structural peculiarities of some of their parts; and this is, I 

 believe, the best definition which can be given of genera. They 

 are not characterised by modifications of the features of the fami- 

 lies, for we have seen that the prominent trait of family difference 

 is to be found in a typical form ; and genera of the same family 

 may not differ at all in form. Nor are genera merely a more 

 comprehensive mould than the species, embracing a wide range 

 of characteristics ; for species in a natural genus should not pre- 

 sent any structural differences, but only such as express the most 

 special relations of their representatives to the surrounding world 

 and to each other. Genera, in one word, are natural groups of a 

 peculiar kind, and their special distinction rests upon the ultimate 

 details of their structure." 



We could have wished in this place some remarks on the ten- 

 dency at present prevalent to sub-divide the old genera into a 

 multitude of new ones, characterised by the most trivial and ev- 

 anescent differences, a process which threatens to reduce some 

 departments of Zoology to a mere chaos ; and which, from the 

 differences of view that arise, as a matter of course, when a natu- 

 ral genus is thus broken up, loads science with an odious and 

 vexatious synonymy. There appears to be a prevalent idea that 

 a genus should necessarily contain few species; but this is obvi- 

 ously an error, since the number of species generically related to 

 each other varies between large limits in different groups of ani- 

 mals. Nor is a genus to be created merely to include species 

 related to each other in a very near degree ; but for those por- 

 tions of a natural family in which the details of execution in 

 the more important parts correspond, however many the species 

 so agreeing, may be. In such a genus there may be many 

 minor sub-divisions established for convenient reference, or to ex- 

 press minor distinctions ; but these should not be characterised as 

 distinct genera. Attention to this is all the more important, be- 

 cause the generic name is attached to the species and should tell 

 something of ils affinities. In order to appreciate natural genera, 

 some breadth of mind is required, as well as familiarity with de- 



