286 



Mr. C. T. Eeffan on the 



It will be seen that I am able to give osteological charac- 

 ters for only 3 of the 6 families, but these are exceptionally 

 well defined. Liitken has given valuable descriptions, 

 accompanied by excellent figures, of the osteology of 

 Ceratias and Himantoloj^hus (Dan. Vidensk. Selsk. Skr. (5) 

 xi. 1878, pp. 307-348, c. figg., and (6) iv. 1887, pp. 323- 

 334, c. pL). Garman's figure of the skeleton of Dolopichthys 

 (Mem. Mus. Comp. Zool. xxiv. ])\. xiv.) does not reveal any 

 essential differences from Himajitolophus. As most of the 

 species of tliis group in the British Museum collection are 

 represented by unique types, a thorough osteological study 

 was out of the question ; on the other hand, the skin of these 

 fishes is so loose, that once an incision has been made 

 (which has generally been done to ascertain the branchial 

 formula) it is a very simple matter to examine the bones of 

 the upper surface of the head. Thus I have been able to 

 ascertain that Diceratias ^ho^s, mix,ny important resemblances 

 to llimantolophus, but Melanocetus differs in the presence of 

 well-developed parietals. 



Y\g. 6. 



soc epo 



B. 



C. 



Skulls of A. Diceratias hispinosm, B. Ceratias hoJboUi and C. Melanocetus 

 johnso7iii, seen from above (somewhat diagrammatic). B is based 

 on Liitken 's figure. 



Lettering as before. 



Family 1. Ceratiidae. 



Strongly compre.ssed ; eyes small ; gill-openings below 

 and immediately behind pectorals ; gills 2h Mouth pro- 



