NO. I THE INSECT HEAD — SNODGRASS 5 



ganglia that become the tritocerebral lobes of the definitive insect 

 brain. Otherwise the segment of these ganglia cannot be recognized 

 in the composition of the adult cranium, and it is but little evident 

 in the embryo, though embryonic vestiges of premandibular append- 

 ages have been observed in several insect species. A short region in 

 the embryo of Japyx between the mouth and the mandibular segment 

 is identified by Silvestri (1933) as the tritocerebral segment, since, 

 though it bears no trace of appendages, it does contain rudiments of 

 a pair of ganglia. Likewise in the embryo of a centipede, Scolopendra, 

 Heymons (1901) regarded a space between the antennae and the 

 mandibles as pertaining to the tritocerebral segment because of the 

 presence of paired coelomic sacs and ganglion rudiments within it. 

 In the symphylan Hanseniella, Tiegs (1940) says, "the pre-mandibu- 

 lar ectoderm curves round the stomodaeal opening, and forms much 

 of the inferior surface of the clypeo-labrum," but he admits this has 

 not been demonstrated in the insects. 



The development of ganglia from the postoral ectoderm that be- 

 come directly the tritocerebral lobes of the brain has been observed 

 in insects by so many writers that there can be no question concern- 

 ing the origin of the tritocerebral ganglia in the insects. These ganglia 

 are always connected by a suboesophageal commissure, and give off 

 the root nerves of the preoral frontal ganglion. The tritocerebral 

 segment itself, however, appears to be practically eliminated. Eastham 

 (1930), in his study of the embryogeny of Pieris, says that "when 

 the premandibular ectoderm has given rise to the tritocerebral neuro- 

 blasts it loses its distinctness as a segment and is no longer distinguish- 

 able." However, a premandibular segment is present as a distinct 

 somite in the crustacean embryo (fig. i E), bearing the rudiments of 

 second antennal appendages {sAnt). A corresponding segment of 

 the chelicerae is present in the embryo of Arachnida (D, Chi). It 

 may be inferred, therefore, that a fully developed premandibular seg- 

 ment was present in the ancestors of all the mandibulate arthropods, 

 and a corresponding cheliceral segment in the chelicerates. 



The cephalic nervous system of the Crustacea appears to be more 

 primitive than that of the insects and myriapods. In the crustaceans 

 small premandibular ganglia are present as swellings on the nerve con- 

 nectives between the brain and the mandibular ganglia. They are 

 united by a suboesophageal commissure, and give off the root nerves 

 of a small preoral "oesophageal" ganglion, which clearly is the frontal 

 ganglion of the insects. In the branchiopods the nerves of the second 

 antennae are given off from the connectives close to the ganglia. These 

 ganglia on the connectives in the Crustacea thus appear to be the trito- 



