SEPTENARY SYSTEM. 237 



occupies is very considerable, I must decline entering, at pre- 

 sent, on that very intricate subject; at the same time, of course, 

 relinquishing the very important assistance my views would 

 obtain from the established superiority of the Neuroptera. I 

 cannot, however, dispense with the necessity of taking a 

 cursory view of this interesting class. 



The heterogeneous contents of the class Neuroptera are 

 excessively perplexing ; but, I find, the more strenously we 

 attempt to place them naturally, the more nearly will such 

 arrangement accord with that which a Septenary System 

 requires. Latreille divided them into four " natural families." 

 Three of these appear sufficiently precise, " Libellulines, 

 Ephemerines and PhryanidesJ" The fourth, called " Plani- 

 pennes," including Nemopteryx, Bittacus, Panorpa, Boreus, 

 Ascalaphus, Myrmeleon, Osmylus, Hemerobius, Psocus, Termes, 

 Embia, Raphidia, Mantispa, Corydalis, Sialis, Nemoura and 

 Perla. A glance at these genera will show that this group is 

 a most carelessly constructed one ; its very name leads one to 

 expect as much. Psocus, Termes, and Embia, are isomorphous 

 genera, and possess a limited, though various, prothoracic deve- 

 lopment ; their tarsi are tri-articulate. Perla and Nemoura are 

 also isomorphous, but have a much more extensive prothoracic 

 development ; their tarsi are likewise tri-articulate. Ascalaphus, 

 Myrmeleon, Osmylus, Sialis, Hemerobius, Raphidia and Man- 

 tispa are necromorphous ; they have moderate prothoracic 

 development, with the exception of Raphidia and Mantispa, 

 in which that segment is exceedingly elongate, as in Collyuris, 

 Mormolyce, and other carnivorous Coleoptera: all these genera 

 have five-jointed tarsi. There is another group still remaining, 

 which consists of Bittacus, Panorpa, Boreus, and Nemopteryx. 

 These singular genera agree in the structure of the parts of the 

 mouth, which are prolonged in the form of a beak ; their meta- 

 morphosis is unknown ; their tarsi are five-jointed ; their pro- 

 thorax is restricted as in Hymenoptera. Here, then, we have 

 seven groups, some of which, though very closely allied in ap- 

 pearance, present still those striking structural differences whicli 

 render it impossible to unite them. I will attempt to sketch 

 an arrangement of these groups. 



To begin with the central or typical group, I select the 

 Libelhdina, as in every way the most perfect ; whether we con- 

 sider their powers of locomotion, or parts of manducation. It 



NO III. VOL. IV. I I 



