440 'OJiT THE TRIBES 



the a bo ve three hisects should eventually prove to be the types 

 of their respective osculant orders, they may in conformity 

 to the method in common use be styled Bomboptera, Me- 

 galoptera Lat., and Raphioptera. As for Mantispa, I can 

 scarcely conceive it to offer any type of form disthict from 

 Mantis, from which indeed it only differs in having the 

 wings of a Neuropterous insect. It is therefore an insect 

 in the tribe of Phasmina, close to the osculant points of the 

 orders of Orthoptera and Neitroptera. I shall only here 

 add, that while from the inspection of this chain of affini- 

 ties several deductions may be drawn, that particular one 

 Avhich seems to be of paramount importance to Natural 

 History, is the artificial nature of the term Order as usu- 

 ally applied in an insulating sense. In Entomology, at 

 least, an order can only be called natural when the epithet 

 is assigned to a certain group, which, without being in- 

 sulated, has in it one principal ganglion, whose character 

 is in some degree imprinted on the whole of the contents. 

 Thus the types of the five orders oi Mandibulala may be 

 represented by a Carabus, Pompilus, Phryganea, Lihel- 

 luloy and Locusia. 



We have now completed a hasty sketch of the tribes 

 into which the Mandibulata may hereafter with more cer- 

 tainty be subdivided. The review of them, however, 

 affords a simple method of designating the orders, which 

 hitherto I had only indicated by their analogical relations. 



COLEOPTERA LarvcE varia . . . Metamorphosis incompleta 



Orthoptera Larvie hexapodes . Metamorphosis semicompieta 



Neuroptera Larva hexapodes . Metamorphosis varia 



Trichoptera JLarvie eruciforrnes Metamorphosis obtecta 



Hymenoptera Larva apodes . . , Metamorphosis incompleta. 



The beauty of this natural disposition is, that the above 

 three columns are so many chains of connexion, the vari- 



