FISHES. — MCCULLOCH. 97 



study of their variation. Some of the differences indicated in 

 the following key may therefore prove to be invalid. 



a. Body elongate, height about 3 in length ; 4 dorsal 



spines elongatus 1 . 



act. Body deeper, height 2-2| in length ; 5-7 dorsal spines. 



b. 17 dorsal rays atlanticus. 



bb. 13-15 dorsal rays. 



c. Superficial bones of head delicate and narrow. 



d. 5 dorsal spines ; 15-19 abdominal scutes 



pacificus. 



eld. 6 dorsal spines; 9-13 abdominal scutes 



mediterraneus fy intermedins, 

 cc. Superficial bones of head coarse and broad.... gig as. 



Hoplostethus mediterraneus, Cuvier §> Valenciennes, 



var, latus, var. nor. 



(Fig 5.) 



Hoplostethus mediterraneus, Cuvier & Valenciennes, Hist. Nat. 

 Poiss., iv., 1829, p. 469, pi, xcvii.. Id., Goode & Bean, 

 Oceanic Ichthyol., 1895, p. 189, fig. 208. Id., Alcock, 

 Cat. Indian Deep-sea Fish., 1899, p. 34, and Illustr, Zool. 

 Investig., Fishes, 1895, pi. xiv., fig. 3. 



D.vi./13 ; A.iii./9 ; V.L/6 ; P.12-14; C.v-vi/21/v-vi. L.Lat. c.28. 

 Abdominal scutes 11-12. Height of body at origin of dorsal 

 1.90-2.05 in the length from the snout to the hypural. Length 

 of head, to end of bony operculum, 2.65-2.85 in the same. Orbit 

 2.33-2.63 in the head, and wider than the iuterorbital space 

 which is 2,85-3.05 in the same. Sixth dorsal spine 1.87-2.05, 

 longest ray 1.45-1.61, and third anal spine 2.41-2.26 in the head. 

 Base of anal fin 1.94-1.90 in that of the dorsal. 



The above are the proportions of three specimens, 100-126 mm. 

 long, which are evidently very similar to H. mediterraneus, as 

 originally described and figured by Cuvier and Valenciennes. 

 They differ however in being deeper and rounder, and in having 

 only nine instead of ten anal rays. According to Goode and Bean, 

 the number of rays in the dorsal and anal fins is subject to con- 

 siderable variation (DJ2-15; A.8-9), but in forty-five specimens, 

 representing three species, I find them to vary very little. In the 



1 Trachichthys elongatus, Giinther (" Challenger" Rept., Zool., xxii., 

 1887, p. 22) agrees in all its important structural details with 

 Hoplostethus. Ogilby has placed it in Paratrachichthys , Waite (Ogilby, 

 Mem. Qld. Mus., i., 1912, p. 43) but it has no affinity with that genus. 



