466 GENERAL REFLECTIONS ON 



no better proof of it, than what is to be found in the first 

 Essay, and I therefore am unwilUng at present to speculate 

 on the assumption of its entire accuracy. 



I may state however my behef, that future investigation 

 will produce a table expressing the affinity of the Stirpes 

 of Coleoptera, exactly on the same plan as those we have 

 already seen. For if the groups of luliform larvae be di_ 

 rectly divisible into families such as those into which the 

 Petalocera were resolved in the first Essay, then the uni- 

 formity hitherto so observable in the distribution of the 

 animal kingdom no longer holds good, or at least the diffi- 

 culty of reconciling such an hypothesis with this unifor- 

 mity seems at present to me to be insurmountable. But 

 the strongest argument in favour of my opinion, that there 

 is an intermediate class of groups between those Avhich in 

 the last chapter I termed tribes, and those which in the 

 first Essay were called families, is, that every Entomologist 

 has acknowledged that the Linnaean genera Scarabaus, 

 Curculio, Cerambyx, &c. are natural groups ; and these 

 would evidently be lost, did we proceed at once to divide 

 the Chilognathiform larvae into families, such as the Melo- 

 lonthida, Kutelidcd, &c. 



This intermediate class of groups may, as before stated, 

 perhaps be named STIRPES; and had they been properly 

 determined, our next object would probably have been to 

 resolve them into FAMILIES, unless indeed we can sup- 

 pose that there exists still another intermediate class of 

 groups. Here, however, unusual difficulties present them- 

 selves ; for the Lamellicornes of Latreille appear to be de- 

 composable into two distinct divisions of two circles each, 

 which seems irreconcileable with what we have seen, unless 

 we can imagine the Rectoceia and Petalocera to be Stir- 



