SYSTEM OF INSECTS. 405 



the trifling difficulty occasioned sometimes by the disco- 

 very of a new group, be set against the advantage of hav- 

 ing only 2000 names to commit to memory instead of 

 10,000 a ? But if, after all, it is judged best to name sub- 

 genera, M. Savigny's excellent plan of distinguishing 

 them by a plural termination would diminish the weight 

 of the above objection, and might be used with advan- 

 tage. 



When the component parts of any minor group differ 

 from another, for the most part in important charac- 

 ters, indicating some tangible difference in their habits 

 and economy, and confirmed by peculiarities in their 

 larvae ; and these differences run through the whole, ex- 

 cept that as usual they grow weaker as it is passing off 

 to another; especially where they are striking in the 

 centre or type of the group, this is always a legitimate 

 genus : but where the characters assumed are very slight, 

 and nothing peculiar in its habits, economy or larva, 

 warrant such distinction, it ought not to be conferred. 



vii. I must next say a word concerning species and 

 varieties. A species is a natural object whose differences 

 from those most nearly related to it had their origin 

 when it came from the hands of its CREATOR; while 

 those that characterize a variety, have been produced 

 since that event. As we do not know the value and 

 weight of the momenta by which climate, food, and other 

 supposed fortuitous circumstances operate upon animal 

 forms, we cannot point out any certain diagnostic by 

 which in all cases a species may be distinguished from 



* See Bicheno in Linn. Trans, xv. 491. 



