CABANGIDAE 43 



DecapteriM, Megalaspis, Caranx, Trachurus, Trachurops and Chloro- 

 scombrus they are longer, while in the rest, Gnathanodon, Alectis, Cit- 

 ula, Vomer and Selene, they are very long. 



The nasals are attached to the f rentals and ethmoid and project 

 forward. They are never nonprojecting and attached for their full 

 length to the frontal and ethmoid, as in some genera of the family 

 Scombridag. They are close together or wide apart, according to 

 whether the ethmoid is wide or narrow, so their variation in this 

 respect is indicated by the description of the ethmoid. In Trachurus, 

 which has the thinnest ethmoid, the nasals are almost in contact at 

 their bases. 



The suborbital chain of bones is always complete and carries a 

 sensory tube. In Caranx hippos a wide process is sent from the last 

 suborbital that is on the lower level of the eye (apparently the third 

 suborbital) upward and backward across the cheek to the upper end of 

 the hyomandibular, where it is attached to a broad, thin wing of bone, 

 which bends forward to receive it from just in front of the upper end 

 of the preopercle. The upper suborbitals send back a wide, thin plate, 

 which partly covers this stay. In Caranx chrysos the stay is only a 

 little developed. It ends in a free point, and there is no process devel- 

 oped from the hyomandibular to meet it. In Oligoplites the lower and 

 posterior suborbitals are spread out into thin laminas, which, running 

 backwards, covers a large part of the cheek. The lower edge of this is 

 developed from the lower level of the eye, and extends downward and 

 backward along the maxillary nearly to the ridge of the quadrate, 

 which is a continuation of the preopercular ridge. There is no indica- 

 tion of a stay in any of the other forms.* 



The suborbital shelf is in various conditions (see footnote under 

 diagnosis of characters). In some forms it is well developed, in others 

 it is slightly indicated or absent. In Oligoplites and some species of 

 Trachynotus it is absent. In Vomer it is only slightly indicated, and 

 between this condition and a well developed shelf there are all gradia- 

 tions. 



The eye has a bony sclerotic case, never so complete or never so 

 dense and thick as in the Scombrida?, and often no better developed 

 than in many Percoid fishes. 



*The condition of the suborbital bones in Caranx and Oligoplites suggests the 

 suborbital stay of the mail-cheek fishes. Though in one genus it extends upward 

 across the cheek and in the other downward it springs in each of them from the 

 last suborbital on the lower level of the eye just as 5t does in the mail-cheek fishes. 

 In this connection it of course signifies nothing, being doubtless a case of parallel 

 development. 



