258 COMPAEATIVE ANATOMY. 



the ventral cliordj wMcli is primitively comjDosed of tliree, remains 

 embedded in tlie liead (sub-oesopliageal ganglion), and takes no 

 sliare in the concrescence which affects the other ganglia ; it is only 

 in a few cases — in Insecta degraded by parasitism — that a union of 

 this ganglion v/ith the rest of the ventral chord takes place. 



The cerebral ganglion (Fig. 131, A B G, r/s) is almost always 

 distinctly divided into two halves, each of which is again composed 

 of several smaller masses of ganglia, which are often complicated in 

 structure. The ganglia of the ventral chord are primitively paired, 

 and often become closely united. On the other hand, the longitu- 

 dinal commissures remain double, even when they are closely applied 

 to one another. There is also a separation of the ventral chord into a 

 superior and an inferior portion which corresponds to a physiological 

 diffei-entiation. 



The first ganglion of the ventral chord (G. iufraoesophageum) 

 sands off fibres for the organs of the mouth. The three succeeding 

 t'loracic ganglia principally give off nerves to the appendages, feet, 

 and wings ; they are consequently of some size. On the other hand, 

 the succeeding ganglia are, as a rule, small, the last alone being 

 an exception to this, for it is of a lai'ger size in correspondence 

 with its relations to the generative system. 



Even in the Aptera there is a fair amount of vai'iation, for 

 11 ventral ganglia (Lepisma) can be made out in the Thysanura, 

 while there are only 3-4 in the Collembola. The last portion of the 

 ventral chord seems to form a complex of ganglia in many (Orchesella, 

 Achorutes) . 



As to the Pterygota, the chief point is that of all the orders the 

 least amount of metamorphosis is seen in the Pseudoneuroptera. 

 In them the ventral medulla traverses the whole length of the body, 

 and there are 5-9 abdominal ganglia in addition to the three thoracic 

 (Fig. 131, A). The Orthoptera, which have 5-7 abdominal ganglia, 

 resemble them in this. 



There are great variations in the Coleoptera. In some the 

 ventral chord extends as far as the end of the abdomen, sometimes 

 possessing 8 separate ganglia (e.g. in Cerambycidte, Carabida?, 

 etc.); in others again the 3 ganglia of the thoracic portion are 

 only represented by two, the second and third being fused, whilst 

 the abdominal ganglia are also connected into one mass, which im- 

 mediately succeeds the preceding ganglion (Curculionida and Lamel- 

 licorniae). In other families there are connecting links of various 

 kinds between these coj^ditions which represent the extremes. In 

 the Hymenoptera we generally find the thoracic ganglia reduced to 

 two, while the abdominal part of the ventral chord has frequently 5 or 

 6 separate ganglia. These are in many, however, reduced to 4 or 3, 

 and even to one. The abdominal part of the ventral chord is, in the 

 Hemiptera, placed in the thorax, and is represented by a ganglionic 

 mass connected with the thoracic ganglia, which are also simple, by 

 a commissure of varying length. The nerves for the abdomen con- 

 sequently take a longer course, and form two longitudinal trunks 



