196 EDWARD PHELPS ALLIS, JUN. 



as a part of the ligament and not a nerve. Still later I 

 found that the slight discoloration of the mesial edge of the 

 ligament, noticed in the section next anterior to the one in 

 which the nerve so suddenly and distinctly appeared, could 

 be traced upward and forward to a point where the ligament 

 crossed, and lay closely against that branch of the ramus 

 oticus that innervates organs 87 to 91 inf. Here, then, was 

 a rational explanation of it, and as I could not definitely 

 trace the nerve in sections, I had Mr. Nomura try to find it 

 by dissection. He succeeded, in two embryos, in finding the 

 dorsal end of the nerve, where it turned downward along the 

 ligament, but he could not trace it to its termination in the 

 diverticulum of the spiracle. As I readily find its ventral 

 end in sections, I am fully convinced that the nerve is a 

 branch of the oticus facialis that has this course and origin — 

 a conclusion practically confirmed by Hoffmann's (36) work 

 on Acanthias. Such being the case, and there being no 

 special sensory tissue in the auditory diverticulum of Mus- 

 telus, it is plainly evident that the sensory diverticula of 

 Amia and Lepidosteus cannot be the homologues of the 

 auditory diverticulum alone, of Mustelus, as Wright was 

 inclined to believe, unless it be assumed that the spiracular 

 sensory organs in the three fishes are not homologous. It 

 seems much more reasonable to assume either that the three 

 diverticula of Mustelus have fused to form the single diverti- 

 culum of Amia, and of Lepidosteus, or that the auditory 

 diverticulum of Mustelus has entirely disappeared in Amia, 

 but is still represented in Lepidosteus in that dorso-lateral 

 branch of the diverticulum of the fish that Wright considered 

 as the blind outer end of the spiracular cleft itself. In 

 Wright's descriptions of the cleft in Lepidosteus he says, 

 " The cleft itself is directed upwards, outwards, and slightly 

 backwards, but it will be readily observed from fig. 5 that 

 it is separated from the most anterior filaments of the pseudo- 

 branch by the whole thickness of the hyomandibular adductor 

 muscles." In my copy of Wright's paper the plates are 

 lacking, so I cannot compare this statement with the figure 



