AMERICAN GOBIID^ AND CALLIONYMID^. 59 



Island, Marshall Group. We are not able to detect any speci- 

 fic difFerences. Some of the American specimens might 

 readily be identified as alhopunctatas, but, on account of 

 the perfect intergradation of these specimens with the 

 typical soporcdors, we prefer to refer all the specimens be- 

 fore us to the one species soporator, which seems therefore 

 to be universally distributed in tropical seas. 



23. Gobius andreiSauvage. Ecuador. 



Bull. Soc. Philom., Paris. Ser. 7, IV, 44; 1880. 



Depth 6 in total length; head, 4. D. YI-I, 9; A. I, 7. 

 Scales, 35 in a median, 12 in transverse series. 



Head much wider than high; snout equal to diameter of 

 eye, 4i; in head; interorbital space narrower than eye; max- 

 illary reaching below centre of eye. Teeth of outer series 

 larger than the others; no canines. First dorsal not quite 

 as high as body; caudal rounded; upper rays of pectoral 

 silky; ventrals not to vent. Scales ciliated. 



Uniform brown; lower part of spinous dorsal tinted with 

 black; soft dorsal and anal with large brown spots; anal 

 blackish. Length 0.160. 



(Guayas, Ecuador.) 



An examination of the type of this species will doubtless 

 prove it to be a soporator. 



24. Gobius nicholsi Beau. Califoruiau fauna. 



25. Gobius glaucofraenum (Gill)- West luilian fauna. 



There are four specimens of a Gohius in tlie collection 

 labelled "No. 13,212, Tortugas, Lyman." They agree with 

 Dr. Gill's description of glanco/rcenain, which is said to be 

 found on the coast of Washington Territory. Evidently a 

 mistake has been made in the locality of the type or of the 

 specimens in the Museum of Comparative Zoology, as it is 

 not probable that this species occurs in the waters of Wash- 



