458 MR. J. Y. JOHNSON ON SOME NEW [June 3, 



adipose dorsal. The pectorals are placed low down, are narrow and 

 pointed ; they are much damaged in the specimen, biit they are 

 not rudimentary (as in JVannobrachiiim), as they reach back to the 

 root of the ventrals. The ventrals have five or six rays and reach 

 back to the vent, but not so far as the anal ; they are inserted 

 nearly under the commencement of the dorsal, 12 mm. from the 

 snout and 5 mm. behind the root of the pectorals. The anal begins 

 behind the end of the dorsal and has about 7 rays ; it is about as 

 high as the dorsal, but its base is rather longer. Caudal forked. 



There are no spines on the tail above or below. The lateral 

 line could not be made out. 



I am much indebted to the Rev. Padre Ernesto Schmitz for the 

 single specimen of this little fish that has occurred. It was obtained 

 from a fisherman. It was only ly^^ inch long, with a height of | 

 inch. "When it came into my hands it had been much injured, and 

 it has therefore been impossible to speak positively as to some of the 

 details. Perhaps it had been found in the stomach of another fish. 

 However that may be, it had a deep-sea aspect. The following 

 are the dimensions : — 



millim. 



Total length of the fish 39 



Length to base of the caiidal fin 34 



Height 6-5 



Head, length 8 ; thickness 4 



Eye, longer axis 2--5. Rictus nearly 7 



Dorsal fin, length of base ca. 5 ; height in front ca. . . .j 

 Ventral fins, distance from snout 12; from root of 

 pectorals 5 



Anal fin, height ca. 5 : length of base ca 6 



... 7 



Anal fin, distance from end to base of caudal 



Family SxEENorTTCHiD^ii:. 



5. GOUOSTOMA 31ADEEEXSE, Sp. U. 



B. 11. P. 10. V. 8. A. 33. C. III. + 19 + III. 



Elongate, compressed : the height compared with the length minus 

 the caudal fin being as 1 to 6i. Blackish, with two rows of silvery 

 or pale steel-blue spots along each side of the belly. The specimen 

 seems to have been clothed with scales, but they have disappeared 

 except from the head. The ridge of the back is rugosely warted, and 

 apparently there have been no scales in that part. 



The head is to <ihe length without the caudal as 1 to 54. The 

 top of the head is scaleless, and t wo low converging ridges meet in 

 front of the orbits. The cheeks bear rather large scales. The 

 opercular pieces are very thin and the gill-openings very wide. The 

 profile is rather steep and the snout short. The round eye does not 

 reach to the profile ; its diameter is contained in the head about 

 five times : it is distant from the snout rather more than one dia- 

 meter, and from the jaw rather less. The upper border of the month 

 is formed partly of the prcmaxillary and partly of the maxUlary, 

 both being armed with teeth ; the latter is dilated posteriorly. The 



