442 CALIFORNIA ACADEMY OF SCIENCES. 70 



The two anal spines united with rest of the fin. No free 

 anal spines. Ventral ray really I, 5, the inner ray very 

 wide, made up of four branches so that the rays seem 

 more numerous : ventral spine obscure. Anal fin short. 

 Pectoral fin falcate. Both dorsal and ventral with sheath. 

 Soft dorsal and anal low, the last ray slightly lengthened. 



Dr. Gill is probably right in regarding Ncinatistius as 

 type of a family distinct from the CarangidcB. 



Family STROMATEID.F:. 



101. Rhombus medius (Peters). 



Originally described by Dr. Peters from Mazatlan ; not 

 seen by Dr. Gilbert or by us. Only the original type in 

 the museum at Berlin seems to be yet definitely known. 



Family CHEILODIPTERID^. 



102. Apogon dovii Giinther. 



This species was found hy Dr. Gilbert at Mazatlan, but 

 was not seen by us. 



103. Apogon retrosella Gill. Cardenal. Plate xxxvii. 

 Two specimens of this most beautiful little fish were 



obtained by us with dynamite off the Isla Blanca and 

 Creston Islands. Only the very young, found by Mr. John 

 Xantus, at Cape San Lucas have been hitherto known. 



Head 2-f; depth 2^3: scales 3—26-9; dorsal VI— 1, 10: 

 anal II, 9: eye 2]/% in head; maxillary i3/(; snout 4^ ; 

 interorbital 4; fust dorsal 2*4' ; second dorsal i| ; caudal 

 1 5 : pectoral !]/> ; ventral 1;'. 



Body rather plump, not much compressed, the profile 

 rising steeply from snout to first dorsal. Caudal peduncle 

 long and strong; eye very large; mouth large, oblique, 

 the maxillary opposite posterior margin of pupil. Teeth 

 small, the outer scarcely enlarged. Premaxillary pro- 

 tractile; no supplemental maxillary. 



