PROCEEDINGS OF UNITED STATES NATIONAL MUSEUM. 359 



in the Epinepheli is generally eleven (in one genus nine and in a single 

 bpecies ten). There can, we are sure, be no possible question of the 

 i:)ropriety of separating Epinephelus at least as a whole from Serranus 

 (typified by Sofanus scriba). The relations of Epinephelus and Serranus 

 are indeed not very close, and only in an artificial grouping could they 

 be confounded. 



Whether it is desirable to subdivide Epinephelus into genera depends 

 somewhat on the value which we wish to give to a genus. As a whole, 

 the species certainly form a natural group. It is also true that they 

 divide readily into several smaller groups, several of which are well 

 defined, easily recognized, and apparently natural. In the present 

 paper six of these are regarded as distinct genera, though we should 

 not seriously object to regarding them as subgenera of a single genus 

 Epinephelus^ as in some other publications we have already done. Four 

 of these groups {Mycteroperca^ Alphes'tes, Promicrops, Dermatolepis) are 

 characteristically American. The others {Epinephelus^ Enneacentrus) 

 are cosmopolitan. Epinephelus is the central genus of the group. Epi- 

 nephelus and Enneacentrus are also much less homogeneous than the 

 other genera, and perhaps may admit of further subdivision. 



We cannot, however, regard the several groups {Schistorus, Hypor- 

 thodus^ Lahroperca^ Petrometopon, Menephorus, &c.) as being worthy of 

 consideration as genera on the basis of the definitions which they have 

 thus far received. While some of these may be possessed of "cranial 

 characters" sufficiently distinctive, it remains to be shown what these 

 cranial characters are, and that they are not, like other characters, sub- 

 ject to intergradation, so their existence becomes merely a question of 

 more or less. 



As the purpose of this paper is, however, to facilitate the identifica- 

 tion of. species, we proceed at once to an analysis of the chief external 

 characters which distinguish the six genera admitted by us. 



ANALYSIS OF GENEKA ALLIED TO LPINEPHELUS. 



a. Scales or some of them more or less ctenoid ; canines distinct in front of each jaw; 

 body oblong, elongate ; preopercle more or less serrate, 

 b. Dorsal spines eleven (ten in Epinephelus analoyus). 

 c. Anal fin elongate, its rays III, 11 or III, 12: caudal tin lunate or truncate; 

 spines slender, those of the anal fin graduated; lower jaw strongly pro- 

 jecting; cranium rather broad between the eyes, posteriorly with three 

 suhequal crests; scales small, largely cycloid, those of the lateral line 

 simple; pyloric cceca few (12 to 20); soft dorsal with 16 to 18 rays. 



Mycteroperca, 1. 

 cc. Anal fin short, its rays III, 8 or III, 9; spines rather robust: posterior part of 

 cranium with the lateral crests little developed; scales cteuoid. 

 d. Scales of lateral line each with 4 to 6 strong radiating ridges; cranium 

 extremely broad and depressed between the eyes; the anterior profile a 

 little concave; lower jaw projecting; pyloric cojca excessively numerous; 

 second dorsal with 16 rays ; caudal much rounded; size very large. 



Promicrops, 2. 



