582 PROCEEDINGS OF THE AMERICAN ACADEMY. 



plexus of the same region is formed symmetrically from branches of 

 nerve III, one the right nerve, the other the left. He\Tnans et van der 

 Stricht ('98) also note this condition in Branchiostoma lanceolatum. 

 Dogiel ( : 02) describes in detail the structure of the plexuses of the 

 mouth border. These plexuses Avill be briefly considered later in this 

 paper. I found no branches of left nerve II joining the inner mouth 

 plexus, as described by Hatschek; on the contrary, the specimens 

 observed indicate that the second nerve pair is not concerned in the 

 innervation of the deep-lying structures of the mouth border. 



Interesting ^•ariations occur in the distribution of the ])ranches of 

 nerve III in both species. In one case (B. lanceolatum) a branch of 

 right nerve III anastomosed with the branch of left nerve III on the 

 right side, thus actually joining the inner mouth plexus. Ordinarily 

 the ramifications of left nerve III in the inner mouth plexus on either 

 side join each other near the median line, but in one specimen (B. 

 caribaeum) no anastomosis occurred, thus presenting the curious 

 condition of an incomplete "nerve ring" in the inner mouth plexus. 



Ncrrc IV. — This nerve usually impregnates with methylene blue 

 coincidentally with nerve III, but the portion extending over the 

 trunk muscles often shows a better impregnation after longer immer- 

 sion. Bilateral asymmetry exists in the fourth nerve pair, as in the 

 third pair, and many variations occur in its manner of division, and 

 the distribution of its branches. As shown in Plate 2, Figure G (B. 

 caribaeum), nerve IV arches slightly more than nerve III immediately 

 after its exit from the neural tube. Near its place of exit, this nerve 

 divides into a dorsal and a ventral ramus. More than one dorsal 

 ramus may be present, and the ventral ramus may subdivide near its 

 origin; but I have never noted the entire absence of the ventral ramus, 

 or its equivalent, on either side of the body. Occasionally the dorsal 

 ramus of a sensory nerve is absent, or represented by a twig of practi- 

 cally no importance. In such cases its territory is supplied by branches 

 of an unusually well developed dorsal ramus of one of the two adjoin- 

 ing nerves. The place of division of nerve IV into a dorsal and a 

 ventral ramus may be readily noted in Figure 6. This place of 

 division is often concealed by the muscles, but frequently may be 

 ascertained by the aid of strong light, especially in methylene-blue 

 preparations preserved in glycerine and ammonium picrate. Such 

 preparations often become very transparent. Heymans et van der 

 Stricht ('98) say that the bifurcation of the dorsal roots into dorsal 

 and ventral rami is always made before exit from the myoseptum, 

 except in the case of the last caudal root. x\ccording to my observa- 



