340 NATURAL SCIENCE. May, 



the identification was correct for the family at least, although not for 

 the genus, and it could only have been possible for one having access 

 to the type to identify it as Mr. Boulenger has done. This is simply an 

 example of the results of the facilities enjoyed by the author and the 

 good use he has made of them. Undoubtedly, however, there will be 

 dissentients from him in some of his reunions. No one, for instance, 

 who has had the opportunity of examining carefully large numbers of 

 the Pomoxis annularis and P. sparoides will concur with Mr. Boulenger in 

 combining them in one species (Pomoxys sparoides, p. 7) ; Dr. Jordan 

 did indeed hastily anticipate him in such a conjunction, but he very 

 soon repented. 



Notwithstanding the reduction of species, genera not represented 

 by any species in the corresponding portions of the Catalogue of 

 Acanthopterygian Fishes are numerous. The additions are of Cen- 

 trarchidae, Elassoma (p. 34); of Percidae, Ulocentra (96), Ammocrypta 

 (101), Crystellaria (103); of Serranidae, Percalates (132), Sinipevca (136), 

 Parascorpis (145), Dinoperca (153), Liopropoma (154), Odontanthias (319), 

 Plectranthias (331), Dactylanthias (333). These, of course, are independent 

 of new genera based on previously known species. 



The system Mr. Boulenger adopts is excellent, so far as his 

 appreciation of the groups is concerned. The fact that the first 

 volume of the new Catalogue is devoted to specialised acan- 

 thopterygians is due to the circumstance that he " was desired to 

 begin the work with the Perciformes." 



The "division Perciformes " is much more restricted than is 

 Giinther's homonymous division, the Squamipennes, Teuthididae, and 

 Scorpaenidae being the principal types eliminated. The families, 

 also, are very differently limited, and the author has expressed the 

 facts of osteology in his diagnoses of the families, and not simply 

 given perfunctory descriptions of skeletons as addenda of curious but 

 useless information. He has, too, for the first time, utilised the 

 development or want of an ophthalmophorous shelf to the second 

 suborbital as a family character, and has also availed himself of the 

 mode of insertion of the ribs and the development of transverse 

 processes for diagnostic purposes. The results are most satisfactory. 



The Percidae are identical with the family as understood by 

 American naturalists, the Centrarchidae include the typical Cen- 

 trarchidae, Elassomidae, and Kuhlia, and the Serranidae form a very 

 large family with numerous subfamilies. 



The relationship of the peculiar Pacific genus Kuhlia to the 

 Centrarchidae was entirely unsuspected, and Mr. Boulenger is the 

 first to recognise the distinctness of its characters from those of forms 

 with which it was formerly associated. Probably some will be 

 inclined to consider it the representative of a distinct family — 

 Kuhliidae. Indeed, doubt must for the present be entertained whether 

 the similarity of the vertebral characters of Kuhlia and the Cen- 

 trarchidae is not the result of independent acquisition from different 

 sources rather than indicative of close genetic relationship. 



The Serranidae are those members of the old Percidae which 

 have the " second suborbital with an internal lamina supporting the 

 globe of the eye," and include as subfamilies the Serraninae, 

 Grammistinae, Priacanthinae, and Centropominae, besides others to 

 be hereafter described. The last two subfamilies named are 

 equivalent to the families Priacanthidae and Centropomidae of some 

 authors, and are very natural groups, whatever may be their value. 

 (We prefer to consider them of family rank.) The Serraninae are 

 extended by the addition of the Plesiopinae, which previously had 



