32 FOSSIL FISHES OF LOMPOC 



soft dorsal, its rays II, 9 to III, 12. In SEBASTODES the anal spines are 

 quite small, with small second interhsemal, and the anal fin has more rays 

 than in related forms. 



This genus, RIXATOR (no cranial characters being available) seems 

 to differ from ROSICOLA Jordan and Evermann in the very long second 

 anal spine and weaker skeleton. From the less nearly related SEBASTO- 

 MUS Gill (ROSACEUS), with which it agrees in these regards, it is sep- 

 arated by the smaller number of vertebrae, 24 (25) instead of 27. Those 

 who choose may regard ROSICOLA and RIXATOR alike as sections of SE- 

 BASTODES. 



The genus RHOMARCHUS (ENSIGER), described by us in 1919, seems 

 related rather to SCORP^ENA than to these SEBASTINE forms. It has much 

 stronger dorsal spines and the second anal spine still longer. The count 

 of 18 vertebrae in the type is evidently an error, some six of the body ver- 

 tebrae being crushed. 



The group of Rock-fish or Rock Cod (SEBASTIN^E), allies of the 

 Rose-fish (SEBASTES MARINUS) of the Atlantic, forms a most conspicuous 

 feature of the living fish fauna of both shores of the North Pacific. 

 About fifty-three species are known from the American shores and 

 twenty-two others from Japan. 



These fish differ enormously among themselves, especially in cranial 

 characters; but these features are subject to a variety of intergradations 

 defying all attempts at sharp division. All possess thirteen dorsal spines, 

 three anal spines, a broad, divided hypural plate, and small rough scales. 

 These facts have led Jordan and his associates, Gilbert, Evermann, and 

 Cramer, to refer the whole assemblage to the single genus SEBASTODES, 

 characterized by the presence of 13 dorsal spines and 27 vertebrae. Mean- 

 while these authors have recognized the groups defined by Gill and by 

 Eigenmann as subgenera. 



A recent study of these fishes made by Dr. Jordan at Monterey in- 

 clines him toward a revival of some of these suppressed genera, especially 

 of the groups called SEBASTODES, ROSICOLA, SEBASTOSOMUS, SEBASTOMUS, 

 HISPANISCUS, PTEROPODUS and SEBASTICHTHYS. 



The number of vertebrae is not uniformly 27 as supposed. In most 

 of the species, but not in all, the number is 27, counting the smaller one, 

 which forms the base of the hypural plate. He finds 27 (26+1) verte- 

 brae in SEBASTOSOMUS (MYSTINUS, FLAVIDUS) in PTEROPODUS (VEXILLARIS, 



ATROVIRENS, NEBULOSUS, CARNATUS and CHRYSOMELAS), 3S well as in 



SEBASTOMUS (ROSACEUS and CONSTELLATUS). 



But in the species called SEBASTODES (PAUCISPINIS, GOODEI) and 

 ROSICOLA (PINNIGER, MINIATUS) the vertebrae number but 25, including 

 the hypural one. The elongate form of the body, the low, straight outline 



