4 PEOCEEDINGS OF THE NATIONAL MUSEUM vol.82 



frayed and silky lower rays. From those two genera Crossochir 

 differs further, though not very trenchantly, in having the gill 

 rakers somewhat longer and more numerous, approaching those of 

 the remaining genera of the Amphistichinae. In addition to having 

 shorter and fewer gill rakers, Grossochii' also differs from these 

 other genera in the dorsal spine and interneural characters, as well 

 as in other respects. The color of the type species gives Crossochir 

 a distinctiveness of appearance sufficient for quick identification. 



The type and only known species is C. koelzi^ herein described. 



The name Crossochir (fringed hand) refers to the frayed and 

 silky lower pectoral rays. 



CROSSOCHIR KOELZI. new species 

 Plate 1 



Holconotus rhodotcrus Yabkow and Henshaw. 1878, p. 205 (Santa Barbara 

 record not checked) ; Bean, 1880, p. 88 (San Diego record not checked; 

 Santa Cruz Island and Santa Barbara records checked) ; Jordan and 

 Gilbert, 1881a, p. 456 (Monterey Bay and Santa Barbara records checked) ; 

 Jordan and Joxjy, 1881, p. 10 (Monterey and Santa Barbara records 

 checked) ; Jordan and Gilbert, 1881b, p. 50 (Santa Barbara record 

 checked; Tomales and Soqiiel records may also apply to this species); 

 Eigenmann, 1892, p. 156 (San Diego, in surf; record not checked); 

 EiGENMANN and Eigenmann, 1892, p. 354 (San Diego record not checked; 

 Santa Barbara and Monterey records checked) ; Gilbert, 1895, p. 466 

 (San Simeon Bay record, checked by Myers) ; Hubbs, 1918, p. 12 (diag- 

 nosis, in key). 



Amphistichus rhodoterus Eigenmann and Eigenmann, 1890, p. 9 (San Diego 

 record not checked; life colors). 



In recording this species repeatedly under the name of rhodoterus, 

 none of the authors quoted indicated any doubt in the identification. 

 An examination of the synonymy of Holconotus rhodoterus (and 

 of other embiotocids), as given by Eigenmann and Ulrey (1894, p. 

 388) and by Jordan and Evermann (1898, p. 1502) gives no indi- 

 cation that a name has ever been proposed for the present species. 

 Cymatogaster yulchellus and C . larhinsri^ as nearly as can be told 

 from the original diagnoses by Gibbons (1854), are based on the 

 true Holconotus rhodoterus Agassiz. All three nominal species 

 were described from San Francisco, where rhodoterus is common. 



Another quoted synonym of Holconotus rhodot^rus^ Cymatogaster 

 elHpticus (Gibbons, 1854), is clearly the species later called 

 Hyperprosopon or Todchthys agassizii^ as I have already indicated 

 (1928, p. 12). On the other hand, Emhiotoca or Ennichthys heer- 

 manni Girard, from Cape Flattery, should certainly be restored to 

 the synonymy of Holconotus rhodoto^us. It was properly so placed 

 until 1890, when Eigenmann and Eigenmann transferred the name 

 to the synonymy of Amphistwhus argenteus^ presumably because 

 their '"''Holconotus rhodoterus'''' was another species {Crossochir 



