Fishes 199 



but are pelagic' The exact records of captures of Ceratiids by the 

 Valdivia are given by Dr. Brauer (1906), and although these records 

 do not include Himantolophus, we regard his statement as probably 

 referable to that genus also on account of the anatomical peculiarities 

 below noted, which Himantolophus shares with most Ceratiids, and 

 as being perfectly consistent with all that is certainly known of 

 the members of this obscure genus. Accordingly, with the greatest 

 respect for many earlier authors, we are" unable to regard Himantolo- 

 phus as a bathybial form, and prefer to consider it as a pelagic fish 

 normally living, perhaps not actually at the surface, but at any rate in 

 the upper strata of the ocean. 



Our reasons for advancing this view are briefly as follows : 



1. The apparently healthy condition of Dr. Wolfenden's specimen 

 when taken in shallow water. 



2. The somewhat compressed form, normally - formed pectorals, 

 and absence of ventrals, which contrast strongly with the depressed 

 form and geniculated pectorals of the allied but bottom-living 

 Lophiidce. The loose skin of the abdomen strongly suggests a cor- 

 relation with a belly capable of being distended with air and acting 

 as a float, as in some pelagic Antennariidce. 



3. The known records, which are perfectly consistent with the 

 view that we have to deal with a pelagic Atlantic fish occasionally 

 carried shorewards by the easterly and northerly Atlantic drift. 

 Such an eastward or northward wandering of a pelagic species of the 

 Western Atlantic may be paralleled by the record from the Norwegian 

 coast of a specimen of Antennarius histrio, and by the arrival, on at 

 least two occasions, of shoals of Lirus perciformis following drifting 

 timber or wreckage on the Irish coast. 



The impossibility of preserving Rheinha.rdt's original specimen 

 makes the task of discussing the question of its specific identity 

 with the subsequently described specimens difficult. Girard (1893) 

 seems to have regarded all recorded specimens as referable to the 

 same species, and there is much to be said in favour of this 

 view. 



