CONCLUDING REMARKS. 87 



form. Still, it is not the distinct scutellum that constitutes 

 the family of Geotrupide, for in the genus Athyreus this 

 organ is not visible; nor yet are the exsert mandibles a di- 

 stinctive character, for in Elephastomus they are concealed. 

 Even if the antennae be characterized as 1 1-jointed, genera 

 such as Hi/bosot-us and Orphnus occur having only ten ar- 

 ticulations, or at least having the eighth joint indistinct and 

 almost obsolete. Yet Elephastomus, Athyreus, Orphnus, 

 and Hybosorus, have all the general habit and characters 

 of the Geotnipidce. In short, such is the variety of strucr 

 ture that I fear no solitary character can be considered as 

 constant; and I can only say of the maxillse, that they have 

 been found less variable than other parts of the insect. As 

 for the above characters which I have ventured to propose 

 for the families, they could only be drawn from such insects 

 as I had the means of dissecting ; and it therefore may be 

 anticipated, notwithstanding all the care bestowed on the 

 designation of these groups, that insects will soon occur to 

 prove that in this respect I am as subject as any of my 

 predecessors to the charge of inaccuracy. 



Families are then evidently artificial ; that is, they are in 

 the present state of the science considered natural only so 

 far as they may be secluded from the rest by chasms which 

 we are by no means to suppose to have been left by Na- 

 ture, but rather to be the necessary result of our imperfect 

 knowledge of species. Thus the Cetoniidce seem at first 

 sight to be a most natural family ; but how near do we ob- 

 serve the Glaphyridce to approach them? — nay, it is almost 

 impossible to determine where a chasm occurs in the links 

 which lead us on througj;i the families of MelolonthidcCj 

 Anoplognathidce, and RutelidcE. If then between the genus 

 Chasmodia and the nearest genus to it of the Cetoniidce a 



