l8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 140 



the esophagus and overlying the roots of the mouth parts (fig. 3). 

 The anterior margin of this ganghon is continuous with the circum- 

 esophageal connectives. It is located at a point where imaginary trans- 

 verse and sagittal midlines intersect on the dorsum of the cephalic 

 plate (fig. i). Nerves innervating the mandibles and maxillae emerge 

 lateroventrally (fig. 2). 



It is conventional (following the terminology of Viallanes, 1887) 

 to divide the supraesophageal ganglion into three portions : proto- 

 cerebrum, deutocerebrum, and tritocerebrum. Since these neuromeres 

 are fused into a single mass in the brain of bipitncticeps, it is impos- 

 sible to delimit their extent in whole mounts. The protocerebrum is 

 poorly developed as contrasted with the brains of higher chilopods. 

 In centipedes with well-developed eyes, i.e., the Scutigeromorpha and 

 some of the Lithobiomorpha, this region is large and distinctly de- 

 limited from the underlying deutocerebrum and tritocerebrum (fig, 

 4, A and B). A typical arthropod protocerebrum, however, is not 

 present in the Geophilomorpha (fig. 4, D). It is represented only by 

 the frontal lobes. The term "protocerebrum," nevertheless, will be 

 employed to refer to the posterodorsal portion of the brain. 



The deutocerebrum constitutes the major bulk of the brain of bi- 

 puncticeps. The two antennary lobes are distinctly separated in the 

 brains of the other orders of centipedes (fig. 4, A, B, C) but in the 

 geophilomorphs are fused at the midline (fig. 4, D). The deutocere- 

 brum is mainly sensory and associational in function, as indicated by 

 its connection with the antennae, but several motor nerves to the an- 

 tennal musculature emerge from this region. 



The poorly developed tritocerebral lobes begin as a fused central 

 mass anteriorly but diverge at their posterior extensions as continua- 

 tions of the circumesophageal connectives. The median recurrent 

 nerve emerges between these two pyriform lobes and is flanked on 

 either side by two smaller nerves which innervate the labrum (fig. 2, 

 N14). No free tritocerebral commissure was observed in this species. 



CEPHALIC NERVES 



The sites where the cephalic nerves emerge from the central neural 

 mass were observed, but it was not always possible to trace the fibers 

 to their terminations. The delicacy of the fibers, the thickness of the 

 sections, and the varying degree of staining intensity of the surround- 

 ing tissue obscure the terminal endings. They seem to end in more 

 than one structure. In a coordination center as complex as the ce- 

 phalic nervous system this is not surprising. Rather than attempt 



