The Lateral Sensory System in the Muraenidae. 155 



mediately anterior to, and contig-uous with the anterior one of the 

 two similar processes that support the dorsal end of the postorbital 

 part of the main infraorbital canal. Between the bases of these 

 supraorbital processes, and also immediately posterior to them, the 

 outer surface of the frontal is slightly gTOOved, the groove running 

 dorso-niesial to both the infraorbital processes and extending but a 

 short distance beyond them. The posterior part of this groove marks 

 the position of what is either the posterior end of the supraorbital canal 

 or the base of the fifth primary tube of the line, this depending on 

 whether the canal anastomoses with the infraorbital canal by its fifth 

 or sixth primary tubes, a point I am unable to definitely determine. The 

 anastomosis is quite probably by the sixth, or posterior terminal tube 

 of the line, the canal, under this assumption, bending sharply down- 

 ward and laterally, near its hind end, and passing between the infra- 

 orbital processes to join and anastomose with tube No. 9 of the infra- 

 orbital canal. The fourth sense organ of the supraorbital canal lies 

 between the bases of the supraorbital processes, the fifth organ Ij-ing 

 immediately posterior to it, transverse in position to it, and immedia- 

 tely dorsal to or even between the bases of the infraorbital processes. 

 It is this position of this organ that seems to indicate that the 

 supraorbital canal anastomoses with the infraorbital by its terminal 

 tube and not by the fifth one. 



Between organs 3 and 4 supraorbital, there is no primary tube 

 or pore. The canal is however here always considerably enlarged, 

 this enlargement undoubtedly representing an aborted primary tube. 

 Between organs 4 and 5, at the angle formed where the canal bends 

 sharply downward and laterally, and on the dorso-mesial aspect of the 

 canal, a large blind pouch has its origin. This pouch extends a short 

 distance forward of the orifice by which it communicates with the 

 supraorbital canal, and a considerable distance backward of it. Its 

 mesial edge lies close to the middle line of the head and close to the 

 corresponding edge of the pouch of the opposite side, but it does not 

 communicate with that pouch. The pouch is blind, as already stated, 

 and is either the fifth or sixth primary tube of the supraorbital line, 

 probably the former. 



