KUTCHIN. — PERIPHERAL NERVOUS SYSTEM OF AMPHIOXUS. 599 



with (iifficulties, since it is necessary to lift the pharynx wall to disclose 

 the nerves on the internal surface of the side muscles, and in the 

 ligamentum denticulatum, and this proceedure is more than likely to 

 break the slender nerve tlireads passing to the ligamentum denticula- 

 tum and the branchial basket. The accompanying series of figures 

 (PI. 2, Fig. 10; PI. 3, Figs. 13, 14; PI. 4, Ings. 15-21, PI. 5, Fig. 30j 

 aims to illustrate the distribution of numerous branches of the rami 

 viscerales ascendentes with as much clearness as these technical 

 difficulties permit. 



As has been noted, a visceral branch may bend around the ventral 

 border of the side muscles, or penetrate a myoseptum further dorsal- 

 ward. A dorsal nerve may possess two or more visceral branches, 

 instead of one. Figures 13 and 15 (Pis. 3 and 4) illustrate a condition 

 of frequent occurrence, in which a branch of the ventral ramus pene- 

 trates to the interior through the myoseptum just dorsal to the most 

 ventral myomere. In either case, di\ision into a descending and an 

 ascending visceral ramus takes place immediately after z'eaching the 

 internal surface of the side muscles, the ascending branch running 

 dorsalward, and the descending branch taking a general ventrad 

 direction. In Figure 15 the descending ramus is small, but in Figure 

 13 its size is equal to that of the ascending ramus. Such variations 

 in size are frequent, and when a branch is unusually small, compensa- 

 tion may ordinarily be noted in an adjoining ner^•e. The ramus 

 visceralis ascendens gives off liranches from either side as far as a main 

 stem can be observed, but the origin of the more ventral of these 

 branches is usually obscured by a gonadic pouch. When one of these 

 pouches is empty (Fig. 13), the nerves which innervate its entire 

 surface with an elaborate network may be readily seen arising either 

 from the ascending \-isceral ramus, or its first branches, or from 

 branches of the descending visceral ramus, or, finally, from meshes of 

 the inner abdominal plexus. These nerve branches ramify to form 

 a plexus which covers not only the more interior surface of the 

 gonadial pouch, but also the surface adjoining the muscles. This 

 plexus is extremely delicate and fine-meshed. The more exterior 

 portion is continuous throughout the region of the gonadic pouches, 

 the network arising from one ner\e uniting with that which is in 

 connection with an adjoining ner^•e. This network also joins the more 

 dorsal fan-like branches of the rami Aiscerales ascendentes. This 

 plexus is probably also continuous with that which is in connection 

 with the intestinal canal, and it possibly extends beyond the anus. 

 When the gonadic pouches are large and press against each other, the 



