92 CALIFORNIA ACADEMY OF SCIENCES 



164. Centropomus armatus GUI. 



Plaik XIII, Flu. •Jli. 



Abiimlaiil in Panama Bay; distinguishable at sij;iil by the excessively 

 ilevelopeil second anal spine with the black interspinous membrane, and by the 

 strong, inflexible dorsal spines. 



The s])ccies is placed by Bonlenger (1805, p. 370), in the synonomy of C. 

 en^ifcrus, but this is wholly inadmissible, as has been pointed out by Jordan and 

 Evermann (189(5, p. 1123)/ 



C. ensifcrus agrees with armdtiaf in the strongly spinous bones of the head, 

 and in its greatly developed second anal spine and black membrane; but it differs 

 conspicuously in the slender, flexible dorsal spines, and in the course of the cranial 

 ridges and the shape of the included areas. In the three specimens of ensifcrus 

 before us from the Atlantic, there are but 5 or G scales in a vertical series between 

 lateral line and middle of second dorsal fin. The gill-rakers in these specimens are 

 as follows: O+l/, with 3 additional rudiments on each limb; 9+17 with 4 rudiments 

 above, below; 8+16, with 4 rudiments on each limb. The distance from the 

 middle of the vent to the base of the first anal spine equals, or a little exceeds, that 

 from tip of snout to hinder edge of pupil. The second anal spine extends to opposite 

 the edge of the scaly area on base of caudal, and is contained 3,"^ to 3g times in the 

 length. The character and direction of the ridges on to]) of head are very charac- 

 teristic, and are well shown by the accompanying figure (PI. XIII, fig. 24). 



In C. armatus, the second anal spine is longer than in any other Pacific 

 species, to be compared in this respect with C. ensiferus only. The fifteen specimens 

 secured by us vary little in the length of this spine, which extends, when depressed, 

 to opposite the base of the caudal in young specimens, well beyond that point in 

 adults. The third anal spine usually exceeds slightly the longest soft ray. The 

 membrane between the second and third anal spines is jet-black in the young, and is 

 usually blackish in varying degree in adults. In none of our specimens has the 

 pigment entirely disappeared, but it is not improbable that it occasionally does so. 

 The dorsal spines are much stronger than in any other known species, and are wholly 

 constant in this regard. The third spine is the strongest, and is inflexible; it is but 

 little longer than the fourth, which projects beyond a line joining the third and fifth, 

 giving a rounded contour to the fin. There is considerable variation in the height 

 of the spines, the fourth spine, when depressed, reaching occasionally to or beyond 

 the origin of the second dorsal, but usually falling short. The second spine is pro- 

 portionally long, 24 to 3| in the third spine. The dorsal formula is VIII-I, 10 (11 

 articulated rays in one specimen); the anal III, 6. The vent is posteriorly placed, 

 the distance from its center to the base of the first anal spine about equaling the dis- 

 tance from the tip of the snout to the middle of the eye. The preorbital has 5 to 7 

 strong retrorse spines; an equal number of still stronger spines are in the " humeral " 

 process; preopercular spines much larger than in rohalito or unionensis. 



The median pair of ridges on top of head are very strong. From the middle 

 of the occiput they diverge a little, both anteriorly and posteriorly, and are separated 



