338 Dr. Chr. Liitken on Onelrodes Eschrichtii, 



the right, and afterwards forms several smaller convolutions, 

 would, if fully extended, be more than half longer than the 

 total length of the fish (from the snout to the tip of the tail), 

 but far from twice as long ; at the commencement it is very 

 wide (diameter about 14 millims.), afterwards considerably 

 narrower (about 4 millims.), but wider again in its last portion 

 (9 millims.). There are no pyloric cseca {Ceratias has two 

 short cgeca pylorica, according to Kroyer) ; a swimming- 

 bladder is also wanting. The hindmost part of the ventral 

 cavity is occupied by two large, oval, somewhat flat ovaries • 

 when the outer membrane of these is removed, masses of ova 

 are seen, forming as it were a chaplet in each of them, com- 

 posed of a plate contorted into close folds. The ova are small 

 and excessively numerous. 



The essential differences between Melanocetus and Oneirodes 

 which have come out under the preceding comparative exami- 

 nation of the new arctic Lophioid will be as follows : — 



1. The mouth in Oneirodes is not vertical, but horizontal, 

 and proportionately less than in Melanocetus ; the length of 

 the jaws is in it at the utmost one fourth of the total length ; 

 and the teeth are comparatively smaller. 



2. T1\\Q frontal ray is clavate, and its thick "head" fur- 

 nished with various delicate tentacular filaments ; its shaft is 

 articulated to a horizontal " interspinal," resembling it in 

 form, inserted under the skin of the forehead. 



3. The thick, isolated, soft (second) dorsal ray is wanting 

 in Melanocetus^ which has fourteen, and Oneirodes only six, 

 rays in the true dorsal fin. 



To these we may also perhaps add a small difference in the 

 number of the branchiostegal rays ; and, finally, we might 

 say that the belly in Oneirodes does not forift a large pendent 

 sac, if it did not seem probable that this peculiarity was only 

 due to the fact that the specimen upon which the genus Mela- 

 nocetus is established had accidentally, a little before it was 

 captured, furnished an exceptionally strong proof of its vora- 

 city. Perhaps at another time it would not have presented a 

 belly more remarkably pendent than it is in our Oneirodes ; 

 nor, perhaps, should we have any more ground for surprise if 

 the latter should at some other time make its appearance with 

 its belly not much less distended than that of Melanocetus. 

 But even if we attach but little importance to this peculiarity 

 (which Giinther has, however, and, in my opinion, rightly, 

 included in the generic characters oi Melanocetus) j there still 

 remain sufficient characters to justify the opinion from which 

 I have started here, namely, that Oneirodes and Melanocetus 

 belong to two different genera. Whether we adhere to tlie 



