946 



INSECTA. 



muscles of this segment, and the second is 

 given directly to the third pair of legs. The 

 fifth ganglion (5), which is situated at the an- 

 terior part of the fifth, or thoracico-abdominal 

 segment, belongs, as we shall hereafter see, to 

 the thorax. Like the abdominal ganglia it 

 gives off two distinct pairs of nerves, the an- 

 terior of which crosses the smaller, and de- 

 scends beneath the larger rectus, and is dis- 

 tributed to the muscles which afterwards con- 

 nect the thorax and abdomen of the perfect 

 insect; and the second, a smaller pair, passes 

 diagonally backwards below the third rectus, 

 to the triangular and transverse abdominal 

 muscles. These are the nerves of the thorax 

 in the larva of the Sphinx. In other species 

 there are some marked differences in their 

 mode of distribution. Thus, in the Cossus 

 ligniperda, as shown by Lyonet, the first and 

 second suboesophageal ganglia are closely ap- 

 proximated together, and have no intervening 

 cords, and their nerves, consequently, pass di- 

 agonally backwards, and not transversely, as 

 in the Sphinx. The nerves for the future wings 

 are not derived, as in the Sphinx, from the 

 ganglionless parts of the cord, but from the 

 ganglionated portion alone, and the distance 

 between the fourth and fifth ganglia is con- 

 siderably shortened. In the larva of the nettle 

 butterfly, Vanessa urtic<e, as formerly shown 

 by us,* the alary nerves are derived directly 

 from the cord itself, between the second and 

 third and third and fourth ganglia, but they do 

 not unite, as in the Sphinx, with a nerve from 

 the next ganglion, but only with the transverse 

 nerves. But in some of the Bombycid<z, as in 

 Odonestis potatoria, we have found the same 

 connexion to exist between the alary nerves 

 and those of the ganglia, as in the Sphinx, and 

 a similar union also between them and the trans- 

 verse nerves. This is particularly interesting from 

 its proving that three distinct branches enter 

 into the formation of the nerves for the future 

 wings. We have found a similar double ori- 

 gination of the alary nerves in the vermiform 

 larvae of Hymenoptera, as in Athalia centifolia; 

 and Burmeister has detected a similar condition 

 of the same nerves in the larva of one of the 

 Coleoptera, Calosoma sycophanta, and, as we 

 shall presently see, a similar condition exists 

 even in some perfect insects. Burmeister, who 

 observed these connexions of the nerves in 

 Calosoma, and called them auxiliary connect- 

 ing nerves, has somewhat curiously remarked 

 that he believes they have riot before been ob- 

 served in any insect, particularly in the Lepi- 

 doptera, in proof of which he adduces Lyonet's 

 description and delineation of the nerves in 

 Cossus, in which, as we ourselves have found, 

 they certainly do not exist. The reason for this 

 difference of manner in which nerves that are 

 given to similar parts in insects of the same 

 order and family originate, is a matter worthy 

 of much consideration. 



Nerves (if the abdomen. All the nerves from 

 the sixth to the terminal ganglion belong to the 

 abdomen, and are nearly uniform both in num- 



* Phil. Trans. 1834. 



ber and distribution In the segments. Each 

 ganglion produces one pair of large, and one 

 of small nerves, entirely distinct from the se- 

 ries of transverse nerves (o) that lie loosely 

 upon the cord. 



It will be remembered that, according to 

 our view of the structure of the cord and 

 nerves, each nerve from a gangliated portion 

 of this cord is formed of one set of fibres from 

 the external or gangliated part, and one from 

 the aganglionic or motor column (Jig- 400), 

 which passes over the ganglion (a), but so closely 

 attached to it as to appear as if it formed a part 

 of that structure. These, therefore, are quite 

 distinct from the transverse nerves (c). The an- 

 terior pair of these nerves from the gangliated 

 cord pass laterally across the smaller rectus, 

 having first received a minute branch from the 

 transverse nerves, and, while passing beneath 

 the larger rectus, each one gives off its first 

 branch (p), and when passing between the 

 second and third oblique muscles its second 

 branch (q\ which is directed forwards, and a 

 little farther onwards its third (r), and its 

 fourth (s), which are directed backwards. The 

 main trunk (t) then crosses the great lateral tra- 

 chea, and having received another filament 

 from the transverse nerve (?i), divides into two 

 branches (), which pass upwards between 

 the dorsal oblique and recti muscles, and 

 are divided into numerous ramifications. 

 About midway across the dorsal recti some of 

 the branches form a small plexus (w), before 

 they are ultimately distributed to the muscles 

 and tegument. The two first divisions of this 

 nerve merit particular attention. The first (p) 

 passes backwards beneath the greater rectus, 

 and divides into two branches. The anterior 

 one (v} is distributed to the four oblique mus- 

 cles, and to the external or under surface of 

 the rectus, which, as we shall presently show, 

 is supplied on its internal surface from the 

 transverse nerves (/). The second division 

 passes backwards and is given, one portion to 

 the under surface of the smaller rectus, and 

 the other to the great oblique, while the ter- 

 mination of this portion (w} is continuous with 

 a filament of the second branch of the trans- 

 verse nerves (/). Some branches from this 

 nerve pass between the triangular and second 

 oblique muscles (x), and others are given to 

 the latero-abdominal. The second branch (q) 

 of the great moto-sensitive nerve passes be- 

 neath the great oblique, and gives off branches 

 to the transverse abdominal muscles (y\ and 

 the latero-abdominal (2), while another sup- 

 plies the latero-abdominal (31) and the oblique 

 constrictor of the spiracle (25), and then di- 

 vides into two portions, one of which is given 

 to the retractor valvula (27), and the other 

 to the transverse lateral muscles (24). The 

 divisions of this branch of the moto-sensitive 

 nerve are particularly interesting. Before dis- 

 secting these nerves we had supposed that 

 the constrictor of the spiracle (25) and the 

 retractor of the valve (21) were supplied by 

 the transverse nerves, and hence were surprised 

 on finding that their nerves were derived from 

 the great moto-sensitive of the gangliated cord, 



