22 4< NATURAL ARRANGEMENT OF INSECTS. 



riably shorter than the shank of the leg, or that part 

 which forms the middle joint. The different modifi- 

 cations of the tarsi will be found, also, of the greatest 

 use in determining the different sub-families : let the 

 entomologist, for instance, examine those of the true 

 ScarabfBUS and of the common Geotrupes stercorarius, 

 whose drowsy hum is so familiar to us in summer ; 

 he will then perceive that, while the first exhibits the 

 greatest possible deviation from the Cetoniada, by having 

 the first joint of the tarsus larger than any of the others, 

 and the last hardly thicker than that which precedes it, 

 the latter has a miniature resemblance to the long 

 thickened claw joint so prevalent among the floral 

 beetles, with this difference, however, that the preceding 

 joints gradually become smaller from the first, as in all 

 this family, and are defended by strong bristles, which 

 are, no doubt, highly useful in assisting the animal to 

 walk upon the ground. In proportion, also, as we leave 

 this family, and approach the confines of the last, we 

 observe a gradual approximation in the structure of 

 their tarsi, until, upon reaching the bulky Dynastince, 

 the two families insensibly blend into each other; so 

 that much more refined characters must be used to de- 

 termine their respective limits, than has yet been em- 

 ployed by any author who has written upon the lamel- 

 licorn genera. 



(196.) On looking to the anatomical construction of 

 the ScarabceidcB, we find they exhibit, in the nature of 

 their jaws, the same primary divisions as are observable 

 in the last family. In one (the Cetoniadce), these 

 organs, from being membranaceous, plainly show that 

 the insects live upon juices, and, " as it were, by lick- 

 ing their food ; " while the second, composed of insects 

 which have corneous or crustaceous maxillae, live on 

 more solid species of food, and by mastication.* But 

 without entering further into these analogies, we must 

 at once proceed to enumerate the five leading divisions, 



* Hor. Ent., 27. 



