The Lateral Sensory System in the Muraenidae. 151 



seem to be either the fused premaxillaries of either side, or a maxillary 

 breathing'-valve bone. If the premaxillaries are not here represented 

 they must form part of the so-called superior maxillary bone of the 

 fish, a probability to which I have already once made reference 

 (No. 6. p. 278). Although I have not yet been able to form a de- 

 finite opinion on this subject, I am strongly inclined to think that 

 true piscine maxillary bones are wholly wanting in Conger, and that 

 the so-called maxillary bones are palato-premaxillarj^ ones. The so- 

 called premaxillaries, currently assumed to be completely fused with 

 the ethmoid and vomer (No. 16), would then be maxillary breathing- 

 valve bones. As favouring this assumption it is to be noted that 

 Conger has no maxillary breathing-valve, this fish thus being in mar- 

 ked exception to Dahlgren's (No. 13) statement that he found no single 

 teleost mthout it. 



Whatever the components that form the end of the skull of 

 Conger may be, it would seem, as stated above, that the median 

 sensory chamber must lie in a dermal ethmoid element. From this 

 chamber a central cavity, or canal, leads backward in the skull, and 

 soon separates into two portions, a dorsal and a ventral one. The 

 dorsal portion transmits the four buccal branches that innervate the 

 ethmoidal organs, tAvo on each side, the two nerves on each side 

 sometimes uniting while still in the canal to form a single branch, 

 and issuing, whether united or independent, by a single foramen which 

 lies at the hind end of the canal. Anterior to this foramen, a foramen, 

 on each side, gives passage to a vein coming from the central canal. 



The ventral part of the central canal is separated from the dorsal 

 part by a transverse layer of bone, is much longer than it, and 

 transmits bloodvessels coming from the median sensory chamber. In 

 the adult it soon becomes a large central cavity, and is traversed by 

 delicate partitions which give it a honeycombed appearance. It varies 

 irregularly in every specimen, and extends backward in the skull to 

 the anterior end of the cavum cranii, from which it may be separated 

 by only a thin bony partition. It passes dorsal to the interorbital 

 opening of the skull, and may be cut across transversely, about the 

 middle of its length, by a bony partition, the partition indicating the 



