ON THE ANATOMY OF AMPHIOXUS LANCEOLAT 379 



downwards and backwards on the surface of the fibrous sheath 

 of the vertebral column ; attach themselves to the antero- 

 posterior aspect of each of the inferior muscular bundles, and 

 may be distinctly traced beyond the extremity of each bundle. 

 When an entire animal is examined by transmitted light, and 

 a sufficient magnifying power, the anterior extremity of the 

 spinal cord is observed, as before mentioned, to terminate in a 

 minute filament above the anterior extremity of the vertebral 

 column. The first pair of nerves is excessively minute, and 

 passes into the membranous parts at the anterior superior 

 angle of the mouth. The second pair is considerably larger, 

 and, like the first pair, passes out of the canal in front of the* 

 anterior muscular bundle. The second pair immediately sends 

 a considerable branch (corresponding to the dorsal branches 

 of the other nerves) upwards and backwards, along the anterior 

 edge of the first dorsal muscular bundle. This branch joins 

 the dorsal branch of the third pair, and, passing on, joins a 

 considerable number of these in succession, and at last becomes 

 too minute to be traced farther. After sending off this dorsal 

 branch, the second pair passes downwards, and backwards on 

 each side above the hyoid apparatus and joins all the ventral 

 branches of the other spinal nerves in succession, as its dorsal 

 branch did along the back. This ventral branch of the second 

 pair is very conspicuous, and may be easily traced along the 

 Line formed by the inferior extremities of the ventral divisions 

 of tlic muscular bundles, the ventral branches of the other 

 nerves joining it at acute angles between each bundle. It 

 may lie traced beyond the anus, but is lost sight of near the 

 extremity of the tail. Twigs undoubtedly pass from the 

 spinal and lateral nerves towards the abdominal surface of the 

 body, but, on account of their minuteness, and the difficulty 

 of detecting fchem in detached portions of the abdominal 

 membrane, they could nol be satisfactorily seen. 



When a portion of the spinal cord is examined under a 



