IDEA OF A GKNUS. 591 



326. 



The method whereby we pass through several qualities and characters 

 which constitute groups, down to the genus, is analytic. The last 

 group before the species is generally considered the genus. But it is 

 possible that one or the other of such determinate genera may be still 

 further subdivided by divarications in isolated characters : are now such 

 groups to be raised to positive genera ? In general this must be nega- 

 tived, for as the higher groups were defined by the simultaneous 

 differences of all the organs, and the peculiarities deduced from them, 

 so must the genera of a family, besides a decided difference in the 

 generic characters, exhibit likewise a general transformation in shape. 

 If this be not the case the genera will necessarily be of unequal value, 

 and it will therefore never be possible to settle the contest upon the 

 generic rights of any determinate group. Every discussion and dis- 

 pute upon any subject rests upon principles ; if upon these a difference 

 of opinion prevails all further argument is useless, and no satisfactory 

 result can ever be obtained until one of the contesting parties can be 

 convinced of the falseness of their principles. 



327- 



In the structure of new genera there are two wrong roads to be 

 avoided ; the one is too circumstantial a dividing, and the other is the 

 unnatural connexion of absolutely different groups. 



The first is most easily followed, when, upon the increase of the 

 number of the species of a genus, the survey of the whole is rendered 

 difficult. Hence has proceeded in modern times the host of genera 

 which are in general deficient in all fixed characters, and are frequently 

 exceedingly superficial, being constructed merely from the external 

 form and general impression. A distinction, as, for instance, that 

 which has been used for the separation of Ophonus and Harpalus, 

 namely, the deeper punctures of the superficies in the former, whereas 

 in the latter it is smooth, must indeed be regarded when it is extensive ; 

 but it never justifies the construction of a genus, and can at most serve 

 for a subdivision of the species within the limits of the genus. Another 

 instance is exhibited likewise in the same family, namely, the Cara- 

 bodcn, in the genus Feronia, the former subdivision of which into dis- 

 tinct genera was founded chiefly upon the form of the prothorax, and 

 which modern writers have very justly, from its being untenable, re- 



