NO. 7 ginsburg: new species of gobioid fishes 115 



other, becoming more or less diffuse on fleshy base of fin ; five similar 

 bands on top of head and nape, nearly spaced the same as bands on 

 body, the anteriormost one on interorbital space, these bands on dorsal 

 aspect forming broad curves, convex posteriorly, from a lateral view^ 

 appearing to run obliquely downward and forward at the back, becom- 

 ing nearly vertical on side of head, one on opercle, one just behind eye, 

 one under middle of eye and two hardly perceptible ones from anterior 

 lower quadrant of eye running forward to maxillary. 



Measurements. — Male, standard length 17.5. Depth 25, depth of 

 caudal peduncle 15, head 30, postorbital part of head 17.5, maxillary 

 12.5, snout 8.5, eye 11, antedorsal distance 36.5. 



Holotype. — In Bingham Coll. Espiritu Santo Island, Gulf of Cali- 

 fornia; male 17.5 mm in standard length. 



Paratypes. — All three in Hancock Coll. San Gabriel Bay, Espiritu 

 Santo Island, Gulf of California; March 7, 1937; 2 males 18-19 mm. 

 Puerto Refugio, Angel de la Guarda, Gulf of California; March 20, 

 1937; female 25 mm. 



Distinctive characters and relationship. — This species is very closely 

 related to (Gobius) Lythrypnus zebra (Gilbert) from the Pacific 

 coast of Southern California and Lower California. A most striking 

 difference shown by the specimens compared, is found in the color pat- 

 tern. The narrow cross bands in pulchellus are divided by a median, 

 narrow, clear-cut, dark streak, which under the microscope shows defini- 

 tive chromatophores, while the areas on either side of the streak are 

 either uniformly dusky or of a plain light color; in zebra the definitive 

 chromatophores spread out and what represents the clear cut dark 

 streak of pulchellus becomes diffuse to cover nearly the entire cross- 

 band. Again, the interspaces in pulchellus are uniformly colored, except 

 some of the posterior ones, while in zebra the chromatophores in the 

 interspaces are more concentrated along a transverse median area. In 

 gross effect, the bands in pulchellus are lighter than or of approximately 

 the same intensity as the interspaces, while in zebra the bands are darker 

 than the interspaces. 



Of structural characters the only difference that I can point out 

 now is the number of pectoral rays, but they intergrade in this character, 

 18-19 in pulchellus, 19-20 in zebra. 



It is evident that the corresponding populations on the Pacific coast 

 proper and in the Gulf of California, respectively, should be recognized 

 at least as distinct subspecies. However, the specimens examined are not 



