THE OSTEOLOGY OF CAULARCHUS M/EAN- 

 DRICUS (GIRARD). 



EDWIN CHAPIN STARKS. 



The family Gobieosocidae to which Caularchus belongs was at 

 one time associated with the family Liparidas on account of a 

 ventral sucking disk possessed by both of them. Dr. Gunther x 

 in 1 86 1 worked out the anatomy of the disks, and showed that 

 they were very different in structure. There being no other sim- 

 ilarities between these two families the Gobieosocidse has since 

 that time been variously placed in the suborder Acanthopteri by 

 different authors, as near the Blenniidae, the Gobiidae or the 

 Batrachoididae. Apparently the most rational disposition of the 

 family has been made by Dr. Gill, who created for it the order 

 Xenopterygii, which he placed with other orders of doubtful 

 relationship with which it had nothing in common. 



In undertaking the following work my interest was (i) to work 

 over and describe the many osteological peculiarities of Cait- 

 larclms, and (2) to attempt to find some indication of its relation- 

 ship and systematic position. In the latter I have had small 

 success, and I present this paper with reference to what anatom- 

 ical value it may have. 



The form I have chosen is Caularchus mceandricus, a species 

 taken in abundance on the California coast. I have also examined 

 Gobicso.v sanguincus from the coast of Peru, and have added notes 

 on its characters where they differ from those of Caularchus. 



CRANIUM. 



The cranium is much compressed and as wide as it is long. 

 The vomerine region is broadly notched in front. On the supe- 

 rior surface and extending back to between the posterior margins 

 of the orbits is a broad depression for the reception of the flat 

 premaxillary processes. There is no supraoccipital crest and 

 only very low temporal crests, which are connected across the 



'Catalogue of Fishes of the British Museum, III., p. 495. 



292 



