376 EEJ'ISIOX OF SEBASTlNJi—ElGENMANN AND BEESOK. vol. xvii. 



given depth are fitted for it or not. To this canse is to be atti'ibuted, in part, the 

 fact that so many northern forms have but hitely been added to the fanna of San 

 Diego, and that a given species may be canght for several days in succession, and 

 then not appear again for some timi". As the ditfereut rock-cod boats have found 

 new conditions, even within a few hundred yards of their usual fishing grounds, 

 they invariably liave brought novelties. Tlius on one day, S. proriger [^ macdonaldi], 

 rufus. COS and viclanosfomas * « * were all brought by one boat which had acci- 

 dentally found new conditions. [All were new to science.] « * * ,S'. /■(</»(/• and 

 Icvis are frequently associated, while ruhrivinctus, <loii(/atHS, chlorosticius, consiellatu.s, 

 rosaccus, vexiUaris, cliiu/sonielas and serrlceps form another group. 



The Sebastin:e are seemingly as abuudaut on the coast of Japan as 

 they are on the coast of the United States. Fe^y species extend 

 further south than the boundary of the United States and they are 

 entirely absent from Mexico and other tropical coasts, but reappear 

 on the coast of Chile in Sehastomus oculatus, which is synonymous 

 with S. capensis of the Cape seas. 



We have examined most of the American species, but none of the 

 Japanese forms. We are fnlly aware of the hazardous nature of 

 attemi^ting a generic subdivision of a large number of species when a 

 good percentage of the whole number is not available for study, and 

 especially when the absent members practically all belong to a partic- 

 ular region; but an examination of tlie skeletons of a large number of 

 species warrants us, in the absence of other evidence,* to considerably 

 increase the number of genera heretofore admitted. The condition of 

 the parietals has been taken as the primary character for generic 

 division and the constant presence or absence of certain cranial spines, 

 associated with a number of minor characters, have been drawn upon 

 to furnish definitions for the genera heretofore united under the names 

 Sebastodes, Sehastomus, Sehnstosomus, and Sebastichthi/s. 



The cranial spines used in generic definitions are located as follows: 

 (1) The preocular is the continuation of the upper posterior angle of 

 the prefrontal into a spine. It is usually present. (2) The supraocu- 

 lar, (3) the postocular, and (4) the tympanic are always near the outer 

 border of the frontal. The last of this series of spines always over- 

 arches a nnicous pore and is present and homologous throughout the 

 group. The postocular, on the other hand, is absent in several genera. 

 (5) The coronals are also on the frontals, but nearer their middle and 

 directly in front of the parietal ridges. They are developed in but few 

 species. There seems to have been a confusion of this name in the 

 Scorpfenina3 and the spine called coronal in Soorpcena does not seem 

 homologous with the spine called coronal in Auctospiim, which has just 

 been described. (0) The parietals (occipital of Jordan and Gilbert), as 

 their name implies, are on the parietal bones and form the spine at the 

 end of the ridge running near the middle of these bones. (7) The 

 nuchals are much less coustantand their taxonomic value consequently 

 much less than that of the other spines. They are formed by the 

 transverse division of the parietal ridges. 



