238 BULLETIN 10 0, UNITED STATES NATIONAL MUSEUM 



jaw depressible. Canines few, large, in front of jaws. Nostrils well 

 separated. Preopercle moderately serrated on hind edge, lower limb 

 entire, without distinct antrorse spine. Opercle with 2 strong spines. 

 Gill rakers short, rather few. Pyloric appendages 10 to 20. Scales 

 small, ctenoid, often somewhat imbedded in skin. Maxillary usually 

 with small scales. Soft vertical fins usually more or less scaly. 

 Lateral line complete, scales triangular, cycloid. Dorsal spines about 

 11, seldom 10, not filamentous, last ones somewhat shorter than 

 median, rays 14 to 18. Anal spines 3, second usually largest, rays 7 

 to 9. Caudal rounded or lunate. Pectoral rays 15 to 20, rounded, 

 short, rarely symmetrical. Ventral moderate, inserted below pecto- 

 ral, each with strong spine, fins close together. 



A large genus, the principal one of the family and widely distributed 

 in most tropical seas. Very many species have been described, 

 though many of them extremely variable in color, pleasing and bril- 

 liant. Many are highly valued as food fishes and some reach a large 

 size. 



The present genus seems to us perhaps most typical of all fishes. 

 In every way its members are the most perfected or typical develop- 

 ment of true bony fishes. From it and its allies all the various 

 offshoots of the percoid series seem to graduate or become variously 

 modified. We have nothing to advance in phylogeny, having 

 adopted in the main the distinctions used by Boulenger in his excel- 

 lent catalogue of the British Museum materials. We hardly think 

 that the emphasis placed on osteological or internal characters, in both 

 this genus and others of the family, are sufficiently complete or 

 practical to be of much service at present. 



Although our series of individuals are extensive the species repre- 

 sented are comparatively few of those generally distributed in the 

 Indo-Pacific as a whole. Frequently the species are difficult to distin- 

 guish; so great are the variations in both color and structure. Even 

 after close comparisons have been made it is often not easy to assign 

 very young specimens to their respective species. In some cases this 

 is hardly possible without completed graduated series of materials 

 well toward adult conditions. It therefore seems desirable for us to 

 recommend to future investigators the collecting of all young tropical 

 serranids whenever possible. Even many of the more frequently 

 collected species have color phases rarely or seldom secured. 



Homalogrystes Alleyne and Macleay is correctly a synonym of Ser- 

 ranus. Its association with Acanthochromis in volume 7, page 27 of 

 the present bulletin, is wrong, as its genotype proves to be a synonym 

 of Serranus tauving. 



