OSTEOLOGY OP SCOMBROID FISHES 



THE COMPARATIVE OSTEOLOGY OF GERRES AND 

 LEIOGNATHUS. 



My material is Gerres embryx Jordan and Starks, from Maryland, 

 and Leiognathus fasciatus (Lacepede), from Samoa. 



The crania in profile and general shape are very similar, but in de- 

 tail they are not. In Leiognathus the supraoccipital extends backwards 

 in a high triangular crest just as it does in the Chsetodonts, without ex- 

 tending at all forward over the top of the head. Its upper edge is thick 

 and divided into three longitudinal ridges. On each frontal a high, 

 thick, crest converges backwards to the supraoccipital. In Gerres the 

 supraoccipital is simple and extends forward over the frontals. Beneath 

 its anterior end a large conical cavity is formed between the frontals, 

 by a high wing on each frontal curving upward and inward and meeting 

 its opposite fellow at the median line. The long ascending processes of 

 the maxillaries projects into this cavity, while in Leiognathus they lie 

 wholly on the upper surface of the cranium. 



Both genera have a lateral process projecting back from the middle 

 of the premaxillary, which, however, is common to many forms. This 

 process in Leiognatlius closes inside of the maxillary, while in Gerres 

 it closes outside. The maxillary elements are very similar. 



In Gerres the suborbital chain of bones is complete, and there is a 

 suborbital shelf present. In Leiognatlius the suborbital chain is incom- 

 plete,* and no suborbital shelf is present. The absence of the suborbital 

 shelf was one of the characters by which Doctor Boulenger defined the 

 family GerridaB (-j-Leiognathidae). That the character is valueless in 

 this connection is proved by the fact that it is not found in Gerres cine- 

 reus. In Gerres bra-silianus it is present, but not so well developed as 

 in Gerres embryx. 



In Gerres the articulation of the hyoid arch to the suspensory is as 

 usual the interhyal attached between the lower end of the hyomandib- 

 ular, and the upper end of the symplectic, which is long and runs 

 along the inner surface of the quadrate in a channel. In Leiognathus 

 the articulation is unique so far as known, the interhyal being articu- 



*I have long had in mind the question as to whether the suborbital canal was 

 ever present in the adult and the bones undeveloped for its protection. Until now 

 I have always found the bones accompanying the canal ; when the canal was ab- 

 sent the bones were. Leiognathus, however, proves to have a suborbital canal and 

 the bones not developed around it. 



