form a sort of thoracic brain. In Fig. 64 we represent a very diffuse and a very concentrated ganglionic chain; A being that of the larva of _Chironomus_, B that of the imago of _Hippobosca_. In both these sketches the cephalic ganglia as well as those of the ventral chain are shown. Turning next to the cephalic masses, we find these in the perfect Insect to be nearly always two in number: a very large and complex one placed above the oesophagus, and therefore called the supra-oesophageal ganglion; and a smaller one, the sub- or infra-oesophageal, placed below the oesophagus. The latter ganglion is in many Insects so closely approximated to the supra-oesophageal ganglion that it appears to be a part thereof, and is sometimes spoken of as the lower brain. In other Insects these two ganglia are more remote, and the infra-oesophageal one then appears part of the ventral chain. In the embryo it is said that the mode of development of the supra-oesophageal ganglion lends support to the idea that it may be the equivalent of three ganglia; there being at one {118}time three lobes, which afterwards coalesce, on each side of the mouth. This is in accordance with the view formulated by Viallanes[34] to the effect that this great nerve-centre, or brain, as it is frequently called, consists essentially of three parts, viz. a Proto-, a Deuto-, and a Trito-cerebron. It is, however, only proper to say that though the brain and the ventral chain of ganglia may appear to be one system, and in the early embryonic condition to be actually continuous, these points cannot be considered to be fully established. Dr. L. Will has informed us[35] that in Aphididae the brain has a separate origin, and is only subsequently united with the ganglionic chain. Some authorities say that in the early condition the sub-oesophageal ganglion is formed from two, and the supra-oesophageal from the same number of ganglia; the division in that case being 2 and 2, not 3 and 1, as Viallanes' views would suggest. The inquiries that are necessary to establish such points involve very complex and delicate investigations, so that it is not a matter of surprise that it cannot yet be said whether each of these views may be in certain cases correct. The supra- and sub-oesophageal ganglia are always intimately connected by a commissure on each side of the oesophagus; when very closely approximated they look like one mass through which passes the oesophagus (Fig. 66, A). The large supra-oesophageal ganglion supplies the great nerves of the cephalic sense-organs, while the smaller sub-oesophageal centre gives off the nerves to the parts of the mouth. From the lower and anterior part of the supra-oesophageal ganglion a nervous filament extends as a ring round the anterior part of the oesophagus, and supplies a nerve to the upper lip.[36] This structure is not very well known, and has been chiefly studied by Liénard,[37] who considers that it will prove to be present in all Insects.