944 



INSECTA. 



diverge a little from each other, and include 

 between them the insertions of the first set of 

 diagonal muscles, and at the posterior part 

 of the third segment again approach each 

 other, and form the third ganglion (3). 



Fig. 406. 



Nenotu system of the larva of Sphinx liguitri. 

 (Newport, Phil. Tram.) 



They then again diverge, and continuing their 

 course into the next segment, pass on each side 

 of the insertion of a second set of muscles, and 

 approaching at the hinder part of the fourth 



segment form the fourth ganglion, (4,) from 

 which they continue their course side by side, 

 and in the next segment form the fifth and last 

 ganglion, (5,) that enters into the composition 

 of the thoracic portion of the nervous cord in 

 the perfect insect. From the fifth ganglion the 

 cords are continued in a direct line, into the 

 sixth, seventh, and succeeding segments, form- 

 ing in each a double ganglion, to the eleventh, 

 where they form the terminal ganglion (11,12). 

 This is considerably larger than any of the pre- 

 ceding, being in reality composed of two dis- 

 tinct pairs, which originally were separated 

 from each other by intervening cords, and be- 

 longed to the eleventh and twelfth segments, 

 but which seem to approach and become closely 

 approximated to each other during the earlier 

 period of the larva state, as suggested by Dr. 

 Grant, and supported by the fact that these 

 ganglia are found more or less approximated 

 together in different individuals, the terminal 

 ganglion in some being distinctly formed of two 

 pairs, scarcely united, and in others so com- 

 pletely coalesced as hardly to be distinguished. 

 In other Lepidoptera, as in Odonestis potutoria 

 and Lasiocampa neustria, and as represented 

 also by Lyonet, in Cossus ligniperclu,* the 

 ganglia continue distinct, and are separated by 

 a very short portion of the cords. But in the 

 Timurcha, the eleventh and twelfth ganglia 

 have completely coalesced, and it is remark- 

 able that they are also united even in the 

 rudimentary form of the nervous system in the 

 aculeate Hymenoptera, as in the larva of the 

 bee, and even in its still more rudimentary state 

 in the larva of Ichneumon Atropos, in both 

 which there are originally thirteen distinct pairs 

 of ganglia, including the supra-cesophageal ones, 

 although Burmeister has imagined that the 

 apodal larvae of Hymenoptera have a nervous 

 system without ganglia,f similar to what he has 

 observed and figured in the larvae of some Dip- 

 tera. It was shown by Swammerdam,} that in 

 the Lamellicornes, as in Oryctet nusicomis, 

 the cords are united laterally, and do not extend 

 beyond the fourth segment, from whence the 

 nerves radiate into the abdomen. In Dyticits, 

 according to Burmeister, the two approxi- 

 mated cords are very short, and the pairs of 

 ganglia are contiguous to each other, and he 

 has found a similar form of the nervous system 

 in the Hog-beetles, Calandra sommcri, in which 

 there are twelve pairs of closely approximated 

 subcesophageal ganglia. The supra-cesophageal 

 pair in this species are distinct, as in Timarcha, 

 but each pair of the subresophageal has co- 

 alesced into a single mass, as in that insect, and 

 the whole do not extend beyond the fifth seg- 

 ment, from whence the nerves radiate into the 

 abdomen, as in the Lamellicornes. These are 

 the conditions which the double cord presents 

 in the different classes. In describing the 

 nerves that proceed from it, we shall divide them 

 into those of the head, the thorax, the abdomen, 

 and the organic functions. 



Plate ix. 



t Manual, (translat.) p. 279. 



J Biblia Natura, Tab. xxviii. fig. 1. 



Op. cit. 



