Erichson, impressed by the variety existing in Neuroptera, separated some of the groups into a sub-Order called Pseudoneuroptera; this sub-Order comprised our Termitidae, Psocidae, Ephemeridae, and Libellulidae. This division is still adopted in several treatises; the Pseudoneuroptera are indeed by some naturalists retained as an Order distinct from both Orthoptera {343}and Neuroptera. Gerstaecker subsequently made use of a system somewhat different from that of Erichson, uniting the Perlidae, Ephemeridae, and Odonata into a group called _Orthoptera amphibiotica_, from which the Termitidae and Psocidae were excluded. The divisions we have here adopted differ but little from those of Gerstaecker, though we have arranged them in a very different manner. It is probable that not one-tenth part of the Neuroptera existing in the world have yet been examined by entomologists, and of those that are extant in collections, the life-histories and development are very imperfectly known. We have, therefore, not considered it wise to adopt a system that would involve great changes of nomenclature, while there can be little hope of its permanency. FOSSILS.—When considering the subject of fossil Insects we briefly alluded to the discussions that have occurred as to whether the fossils of the palaeozoic period should be referred to existing Orders. Since the pages we allude to were printed, M. Brongniart's very important work[269] on the Insects of that epoch has appeared. He considers that these ancient fossils may be classified with the existing Orders of Insects, though they cannot be placed in existing families; and he assigns the palaeozoic fossil Insects at present known, to the Orders Neuroptera and Orthoptera, and to the homopterous division of Hemiptera. The greater part of the species he looks on as Neuroptera, and places in six families—Megasecopterides, Protephemerides, Platypterides, Stenodictyopterides, Protodonates, and Protoperlides. Of these he considers the ancient Protephemerides, Protodonates, and Protoperlides as the precursors, which, we presume, we may interpret as the actual ancestors, of our existing Ephemeridae, Odonata, and Perlidae. Some of the fossils restored and described by the French entomologist are of great interest. We shall notice the Protephemerides, Protodonates, and Protoperlides in connexion with the families to which they are specially allied, and shall now only allude to the quite extinct families of Neuroptera, the Megasecopterides, Platypterides, and Stenodictyopterides.