[3] 



REVIEW OF THE LABROID FISHES. 



601 



Labrus viridis, 20+21 = 41. 

 Labrus biuiaculatus, 18+21 — 39. 

 Labrus livens, 18+20 = 38. 

 Labrus berggylta, 19+19 = 38. 

 Ctenolabrus adspersus, 17+19 =36. 

 Acantholabrus palloni, 18+18 = 30. 

 Odax balteatus (Antarctic), 19+17 =36. 

 Hiatula onitis, 16+18 = 34. 

 Ctenolabrus suillus, 15+18 = 33. 

 Symphodus melops, 15+18 = 33. 

 Symphodus tinea, 15+18 = 33. 

 Symphodus ocellaris, 14+18 = 32. 

 Symphodus cinereus, 14+17 = 31. 

 Symphodus sciua, 13+18 = 31. 

 Symphodus mediterraneus, 13+17 = 30. 

 Lachnolaimus maximus, 12+17 = 29. 

 Lepidaplois scrofa, 11+17 =28. 

 Lepidaplois hirsutus, 11+17 = 28. 

 Decodon puellaris, 12+16 = 28. * 

 Clepticus genizara, 10+17 = 27.* 



Anampsescceruleopunctatns, 11 + 15 = 26. 

 Duymferia aurigaria, 10+15 =25. 

 Halichceres nigrescens, 10+15 = 25. 

 Hemitautoga hortulanus, 10+15 = 25. 

 Thalassoma pavo, 11+14 = 25. 

 Thalassomalunare, 11+14=: 25. 

 Thalassoma dorsale, 11+14 = 25. 

 Julis julis, 11+14 = 25. 

 Julis atlantica, 11+14 = 25. 

 Coris aygula, 11+14 = 25. 

 Giintheria trimaculata, 10+15 =25. 

 Pseudolabrus psittaculus, 9+16=25. 

 Pseudolabrus laticlavius, 9+16 = 25. 

 Gomphosus tricolor, 9+15 = 24. 

 Pseudolabrus celidotus, 10+13 = 23. 

 Xyrichthys macrolepidotus, 10+15 = 25. 

 Xyrichthys novacula,9+16 = 25. 

 Cheilinus fasciatus, 10+13=23. 

 Cheilinus trilobatus, 10+13 = 23. 

 Sparisoma cretense, 11+14 = 25. 



As in most other large groups there has been shown considerable 

 difference of opinion as to the characters which should be used 

 in dividing the Labroids into genera. The tendency with all recent 

 writers has been towards a rather minute subdivision. The numbers 

 of vertebrae seem to us to yield characters of the highest importance. 

 Other characters not to be neglected can be drawn from the size of the 

 scales, the numbers of the dorsal spines, and the dentition. The de- 

 gree of squamation of the head seems to us to have an importance lower 

 than that attributed to it by Bleeker and Gunther. The number of 

 genera as given in this paper could bear reduction only in the Rarpince 

 and Julidince. The genera of the Labrince seem to us natural enough, 

 and most of those of the Scarince are most trenchantly separated. 



ANALYSIS OF SUBFAMILIES OF LABEIDM. 



a. Lower pharyngeals T-shaped or Y-shaped, their teeth conical or tubercular ; teeth 

 in jaws more or less distinct. Carnivorous species, the sexes usu- 

 ally not colored alike. 

 b. Dorsal spines 8 or more, usually well separated from the soft rays ; anal spines 

 3 to 6 (rarely 2). 

 c. Vertebrae about 36 (15 + 18 to 20 + 21); dorsal spines 14 to 21 ; anal spines 3 to 6 ; 

 all the spines pungent; anterior canines f; no posterior canines ; 

 lateral line complete; caudal tin never forked. Species of northern 



waters, none of them tropical Labrince, A. 



cc. Vertebrae 22 to 29 ; dorsal spines 8 to 13 ; (species of tropical or subtropical seas.) 

 e. Vertebrae 27 to 29 (so far as known) ; dorsal spines usually 12 (11 to 14) ; 

 sides of head more or less scaly ; preopercle serrulate or entire. 

 /. Anterior canines strong ; lower pharyngeals large, with large, tubercular 

 teeth ; spinous dorsal not enveloped in scales ; lower jaws naked. 



Harping, B. 



In specimens examined by us. 



